Aristotle Transformed

i
 Ancient Commentators on Aristotle

GENERAL EDITORS : Richard Sorabji, Honorary Fellow, Wolfson College,

University of Oxford, and Emeritus Professor, King’s College London, UK ; and
Michael Griffin, Assistant Professor, Departments of Philosophy and Classics,
University of British Columbia, Canada.

Aristotle Transformed (edited by Richard Sorabji) is a companion volume to the

series, which translates the extant ancient Greek philosophical commentaries
on Aristotle. A webpage for the Ancient Commentators Project is maintained at
ancientcommentators.org.uk and readers are encouraged to consult the site for
details about the series as well as for an updated Bibliography.

‘Aristotle and Alexander of Aphrodisias’, bronze plaquette by Ulocrino (see Donald R.

Morrison’s note on pp. 521–4). Reproduced by courtesy of the Victoria and Albert
Museum, London.

ii
 Aristotle Transformed
The Ancient Commentators
and Their Influence

Second Edition

edited by
Richard Sorabji

Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc

LON DON • OX F O R D • N E W YO R K • N E W D E L H I • SY DN EY

iii
 Bloomsbury Academic

An imprint of Bloomsbury Publishing Plc

50 Bedford Square 1385 Broadway

London New York
WC 1B 3DP NY 10018
UK USA

www.bloomsbury.com

BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc

First edition published by Gerald Duckworth & Co. in 1990

This second edition published in 2016

© Richard Sorabji and the contributors 1990, 2016

Richard Sorabji has asserted his rights under the Copyright, Designs and Patents Act, 1988,
to be identified as Editor of this work.

All rights reserved. No part of this publication may be reproduced or transmitted

in any form or by any means, electronic or mechanical, including photocopying,
recording, or any information storage or retrieval system, without prior permission
in writing from the publishers.

No responsibility for loss caused to any individual or organization acting on or

refraining from action as a result of the material in this publication can be
accepted by Bloomsbury or the authors.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN : HB : 978-1-47258-907-1
ePDF : 978-1-47258-909-5
ePub: 978-1-47258-908-8

Library of Congress Cataloging-in-Publication Data

Names: Sorabji, Richard, editor.
Title: Aristotle transformed : the ancient commentators and their
influence / edited by Richard Sorabji.
Description: Second Edition. | New York : Bloomsbury, 2016. | Includes
bibliographical references and index.
Identifiers: LCCN 2016016470 (print) | LCCN 2016017198 (ebook) | ISBN
9781472589071 (hardback : alk. paper) | ISBN 9781472589095 (epdf) | ISBN
9781472589088 (epub)
Subjects: LCSH: Aristotle—Influence. | Philosophy, Ancient.
Classification: LCC B485 .A655 2016 (print) | LCC B485 (ebook) | DDC 185—dc23
LC record available at https://lccn.loc.gov/2016016470

Cover image © bpk/Skulpturensammlung und Museum für Byzantinische Kunst,

SMB/Antje Voigt.

Typeset by RefineCatch Limited, Bungay, Suffolk

iv
 Contents

Preface to the First Edition vii

Acknowledgments ix
List of Contributors x
Introduction to the Second Edition xii

1 The ancient commentators on Aristotle 1

Richard Sorabji

2 Review of the Commentaria in Aristotelem Graeca 35

Karl Praechter

3 The earliest Aristotelian commentators 61

Hans B. Gottschalk

4 The school of Alexander? 89

Robert W. Sharples

5 Themistius: the last Peripatetic commentator on Aristotle? 119

Henry J. Blumenthal

6 The harmony of Plotinus and Aristotle according to Porphyry 133

Pierre Hadot

7 Porphyry’s legacy to logic: a reconstruction 151

Sten Ebbesen

8 How did Syrianus regard Aristotle? 187

H.D. Saffrey

9 Infinite power impressed: the transformation of Aristotle’s

physics and theology 195
Richard Sorabji

10 The metaphysics of Ammonius son of Hermeias 215

Koenraad Verrycken

v
vi Contents

11 The development of Philoponus’ thought and its chronology 251

Koenraad Verrycken

12 The life and work of Simplicius in Greek and Arabic sources 295
Ilsetraut Hadot

13 Neoplatonic elements in the de Anima commentaries 327

Henry J. Blumenthal

14 The Alexandrian commentators and the introductions to

their commentaries 349
L.G. Westerink

15 Boethius’ commentaries on Aristotle 377

James Shiel

16 Boethius as an Aristotelian commentator 403

Sten Ebbesen

17 An unpublished funeral oration on Anna Comnena 423

Robert Browning

18 The Greek commentators on Aristotle’s Ethics 439

H.P.F. Mercken

19 Philoponus, ‘Alexander’ and the origins of medieval logic 481

Sten Ebbesen

20 Aristotle’s doctrine of abstraction in the commentators 501

Ian Mueller

Note on the frontispiece: ‘Aristotle and Alexander of Aphrodisias’

by Ulocrino 521
Donald R. Morrison

Select Bibliography 525

Index Locorum 573
General Index 585
 Preface to the First Edition

The story of the ancient commentators on Aristotle has not previously been told
at book length. Here it is assembled for the first time by drawing both on some
of the classic articles translated into English or revised and on the very latest
research. Some of the chapters will be making revisionary suggestions unfamiliar
even to specialists in the field. The philosophical interest of the commentators
has been illustrated elsewhere.1 The aim here is not so much to do this again as
to set out the background of the commentary tradition against which further
philosophical discussion and discussions of other kinds can take place.
The importance of the commentators lies partly in their representing the
thought and classroom teaching of the Aristotelian and Neoplatonist schools,
partly in the panorama they provide of the 1100 years of Ancient Greek
philosophy, preserving as they do many original quotations from lost philosophical
works. Still more significant is their profound influence, uncovered in some of the
chapters below, on subsequent philosophy, Islamic and European. This was due
partly to their preserving anti-Aristotelian material which helped to inspire
medieval and Renaissance science, but still more to their presenting an Aristotle
transformed in ways which happened to make him acceptable to the Christian
Church. It is not just Aristotle, but this Aristotle transformed and embedded in
the philosophy of the commentators, that lies behind the views of later thinkers.
Many of the commentaries have been translated in the series ‘The Ancient
Commentators on Aristotle’, published by Duckworth and Cornell University
Press from 1987 onwards (general editor: Richard Sorabji). The present book
will also serve as an introduction to them.
Chapters 1, 4, 10, 11, 19 and 20 are new; 2, 6, 8 and 12 are translated; 5, 9, 14,
15 and 18 are substantially revised. Others are revised in more minor ways;
Greek and Latin passages are translated throughout. The original articles
appeared as follows:
Chapter 2: Karl Praechter, ‘Die griechischen Aristoteleskommentare’,
Byzantinische Zeitschrift 18, 1909, 516–38.

1
For example, Time, Creation and the Continuum and Matter, Space and Motion, and in a related
volume Philoponus and the Rejection of Aristotelian Science, London and Ithaca N.Y. 1983, 1988,
1987.

vii
viii Preface to the First Edition

Chapter 3: Hans B. Gottschalk, pp. 1089–112 and 1150–1 of ‘Aristotelian

philosophy in the Roman world from the time of Cicero to the end of the second
century ad ’, in W. Haase (ed.), Aufstieg und Niedergang der römischen Welt II ,
36.2, Berlin 1987.
Chapter 5: H.J. Blumenthal, ‘Themistius, the last Peripatetic commentator on
Aristotle?’, in Arktouros, Hellenic Studies presented to Bernard M.W. Knox on the
occasion of his 65th birthday, Berlin 1979.
Chapter 6: Pierre Hadot, ‘L’harmonie des philosophies de Plotin et d’Aristote
selon Porphyre dans le commentaire de Dexippe sur les Catégories’, in Plotino e
il Neoplatonismo in Oriente e in Occidente, Problemi attuali di scienza e di cultura
198, Accademia Nazionale dei Lincei, Rome 1974, 31–47.
Chapter 7: Sten Ebbesen, Commentators and Commentaries on Aristotle’s
Sophistici Elenchi, 3 vols, vol. 1 = Corpus Latinum Commentariorum in Aristotelem
Graecorum (CLCAG), vol. 7, part 1, Brill, Leiden 1981, 133–70.
Chapter 8: H.D. Saffrey, ‘Comment Syrianus, le maître de l’école
néoplatonicienne d’Athènes, considerait-il Aristote?’, in Jürgen Wiesner (ed.),
Aristoteles: Werk und Wirkung, vol. 2, Berlin 1987.
Chapter 9: Richard Sorabji, ‘Infinite power impressed: the transformation of
Aristotle’s physics and theology’, in Sarah Hutton and John Henry (eds), New
Perspectives in Renaissance Thought: Essays in the History of Science, Education
and Philosophy in memory of Charles B. Schmitt, London 1989.
Chapter 12: Ilsetraut Hadot, ‘La vie et l’oeuvre de Simplicius d’après des
sources grecques et arabes’, in Simplicius – sa vie, son oeuvre, sa survie, Peripatoi
15, Berlin 1987, 3–39.
Chapter 13: H.J. Blumenthal, ‘Neoplatonic elements in the de Anima
commentaries’, Phronesis 21, 1976, 64–87.
Chapter 14: L.G. Westerink, Anonymous prolegomena to Platonic Philosophy,
Amsterdam 1962, Introduction, x-xxxii.
Chapter 15: James Shiel, ‘Boethius’ commentaries on Aristotle’, in R. Hunt, R.
Klibansky, L. Labowsky (eds), Medieval and Renaissance Studies 4, 1958, 217–44.
Chapter 16: Sten Ebbesen, ‘Boethius as an Aristotelian scholar’, in Jürgen
Wiesner (ed.), Aristoteles: Werk und Wirkung, vol. 2, Berlin 1987.
Chapter 17: Robert Browning, ‘An unpublished funeral oration on Anna
Comnena’, Proceedings of the Cambridge Philological Society n.s. 8, 1962, 1–12.
Chapter 18: H.P.F. Mercken, The Greek Commentaries on the Nicomachean
Ethics of Aristotle in the Latin Translation of Robert Grosseteste, vol. 1 = Corpus
Latinum Commentariorum in Aristotelem Graecorum (CLCAG), vol. 6, part 1,
Brill, Leiden 1973, Introduction, *3–*29.
 Acknowledgments

I wish to thank the following for their generous permission to print, and in some
cases to translate, earlier work:

Chapter 2: BSB B.G. Teubner Verlagsgesellschaft, Leipzig.

Chapter 3: the editors of Aufstieg und Niedergang der römischen Welt.
Chapters 4, 8, 12, 16: Walter de Gruyter and Co., Berlin.
Chapter 6: Accademia Nazionale dei Lincei, Rome.
Chapters 7 and 18: Professor G. Verbeke, director of the Corpus Latinum
Commentariorum in Aristotelem Graecorum.
Chapter 9: the organisers of the Charles B. Schmitt Memorial Symposium held
at the Warburg Institute on 20–1 February 1987.
Chapter 13: van Gorcum, Assen – The Netherlands.
Chapter 14: North Holland Publishing Co., Amsterdam.
Chapter 15: the editors of Medieval and Renaissance Studies.
Chapter 17: the Cambridge Philological Society.

The work on this book, like that on the translations in the series ‘The Ancient
Commentators on Aristotle’, has been generously funded by the following
sources: the National Endowment for the Humanities, an independent federal
agency of the USA , the Leverhulme Trust, the British Academy, the Jowett
Copyright Trustees, the Royal Society (UK ), Centro Internazionale A. Beltrame
di Storia dello Spazio e del Tempo (Padua), Mario Mignucci (Padua) and
Liverpool University.
For an enormous investment of work on the typescripts, including electronic
wizardry, I am indebted to John Ellis, Duran Dodson and Eric Lewis, and for the
indexing to John Ellis, Duran Dodson, Lars Mortensen and Jean-Pierre
Schneider. For his role in the preparation of the King’s College bibliography on
the commentators, I am grateful to Titos Christodoulou. The three translations
from French and one from German were all executed by Victor Caston. The
typing kindly undertaken by Gertrud Watson and D. Woods was exceptionally
difficult because of the extensive revision of some chapters. Thanks are due to
them and not least to Deborah Blake at Duckworth, whose patience and editorial
skill has brought this book and the related volumes of translation to light.

ix
 List of Contributors

Henry J. Blumenthal (1936–1998) was at the time of writing Chapter 5 Senior

Lecturer and at the time of writing Chapter 13 Gladstone Reader in Greek at the
University of Liverpool, UK . From 1997 he was Professor of Greek.

Robert Browning (1914–1997) was at the time of writing Lecturer at University

College London. From 1965 he was Professor of Classics and Ancient History at
Birkbeck, University of London, UK , and later advisor at Dumbarton Oaks,
Washington DC , USA .

Sten Ebbesen is Professor of the Aristotelian Tradition at the University of

Copenhagen, Denmark.

Hans B. Gottschalk (1930–2004) was at the time of writing, and to his retirement,
Reader in Classics at Leeds University, UK .

Ilsetraut Hadot is retired. She was Directeur de Recherches, CNRS , Paris,

France.

Pierre Hadot (1922–2010) was at the time of writing Director of Studies at the
École Pratique des Hautes Études. From 1982 he was Professor of the History of
Hellenistic and Roman Thought at the Collège de France.

H. Paul F. Mercken was at the time of writing Associate Professor of Philosophy

at Florida State University, USA . From 1978 to 1991 he was Professor in the
Faculty of Philosophy, University of Utrecht, Netherlands.

Donald R. Morrison is Professor of Philosophy and Classical Studies, Rice

University, USA .

Ian Mueller (1938–2010) was at the time of writing, and to his retirement,
Professor of Philosophy at the University of Chicago, USA .

x
 List of Contributors xi

Karl Praechter (1858–1933) was at the time of writing Professor of Classical

Philology at the University of Bern, Switzerland. From 1907 to his retirement he
was Professor of Classical Philology at the University of Halle, Germany.

Perè H. D. Saffrey is a scholar at the Couvent St Jacques, Paris.

Robert W. Sharples (1949–2010) was at the time of writing Reader in Classics at

University College London. From 1994 he was Professor of Classics.

James Shiel (1921–2010) was at the time of writing Lecturer in Classics at

University College Dublin, Ireland. From 1963 he was Lecturer then Reader at
the University of Sussex, UK .

Richard Sorabji is Emeritus Professor of Philosophy, King’s College London,

and Honorary Fellow, Wolfson College, University of Oxford, UK .

Koenraad Verrycken is Emeritus Professor of Philosophy at the University of

Antwerp, Belgium.

L. G. Westerink (1913–1990) was at the time of writing teaching in Emmen, the

Netherlands. From 1965 he was Professor of Classics at the State University of
New York at Buffalo, USA .
 Introduction to the Second Edition
Richard Sorabji

Although Aristotle Re-Interpreted (London 2016) will now provide a sequel to

Aristotle Transformed, showing how much has been learned about the subject in
the more than 25-year gap since the first book was published by Duckworth in
1990, the original articles still provide an essential background to understanding
what has been discovered more recently. Many of them need little or no updating.
But if in the introduction to this new edition of Aristotle Transformed I offer a
few updates, sometimes indicating in cases of doubt that I believe no update is
needed, I believe that this will preserve the value of the original articles by
keeping readers abreast of subsequent discoveries. Many of the discoveries will
be described in detail in Aristotle Re-Interpreted.
Two relevant sourcebooks have been compiled in the intervening period, one
explicitly on the commentators up to 600 ad and one on Aristotelian philosophy
as far as 200 ad.1 I now turn to the original chapters of Aristotle Transformed.

In my own Chapter 1, I start by treating Andronicus in the first century bc and

Porphyry in the third century ad as providing two major impulses, the first
making commentary on Aristotle possible and the second putting Aristotle
forever on the curriculum in Europe. If I were writing now, I would add how
much Andronicus’ pupil Boethus (not the famous Latin writer Boethius)
contributed, both by inventing the idea of word by word commentary on
Aristotle and by introducing numerous controversial interpretations of him,
some only now coming to attention. These interpretations stimulated replies
especially from Aristotle’s leading defender, Alexander of Aphrodisias in the
early third century and from the first Neoplatonist commentator on Aristotle,

1
Richard Sorabji, The Philosophy of the Commentators 200–600 AD, A Sourcebook, 3 vols, London
2004; Robert W. Sharples, Peripatetic Philosophy 200 BC to AD 200, Cambridge 2010.

xii
 Introduction to the Second Edition xiii

Porphyry, later in that century. I would also emphasise how Alexander

reconstructed Aristotelianism to enable it to meet advances in the rival Stoic and
Platonist schools.
I offered an interpretation of the mosaics installed by Julian, the pagan
emperor of 362–3 ad, in the Neoplatonist school of the late Iamblichus at
Apamea in Syria. Since then, twenty classrooms of the sixth century have been
excavated of the famous school of Alexandria in Egypt, and I believe the structure
of the rooms reveals a lot about the conception of philosophy teaching.2
There has been much re-thinking about the relations between the
commentators and Christianity. It is not clear that Porphyry’s questioning of
Christianity was really any more hostile than the sincere doubts he pressed about
the Egyptian religion beloved by Iamblichus, the man who both attended to and
‘heard’ Porphyry, whether or not that means that he was a pupil. It has been
questioned whether Porphyry’s Against the Christians was a single book, or a
series of discourses, and whether he was endorsing the pagan oracles in his
Philosophy from Oracles. To correct what I said, he doubted that priestly ritual or
theurgy could bring one all the way to the gods, preferring philosophy. Moreover,
even the pupils of the arch champion of priestly power, Iamblichus, were divided.
The emperor Julian was a devotee of Iamblichus’ theurgy – hence his installation
of the mosaics at Apamea. But this cost his former tutor Themistius, another
commentator on Aristotle, much embarrassment, when Julian openly opposed
his teacher’s views, proclaiming a close connexion between philosophy, priestly
ritual, and kingly rule.
I made a suggestion about the supposedly sordid deal that the philosopher
and commentator Ammonius struck with the Christian bishop of Alexandria.
The deal may have been necessary for keeping the pagan philosophy school
open more than a hundred further years after religious riots around 485–6, and
the consequent persecution of pagan philosophers in 488. But the deal, I thought,
was no more than staying off the inculcation of theurgy, and that has been
regarded by Ilsetraut Hadot as the most plausible interpretation,3 although some
evidence that I tried to offer later has not found favour, as will emerge in one of
the contributions to Aristotle Re-Interpreted. All I would say here is that I did not

2
Richard Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science, Second Edition, Bulletin
of the Institute of Classical Studies Suppl. 103, 2010, 1–10, updated in ‘The Alexandrian classrooms
excavated and sixth-century philosophy teaching’, in Pauliina Remes, Svetla Slaveva Griffin (eds),
The Routledge Handbook of Neoplatonism, London 2014, 30–39.
3
Ilsetraut Hadot, Athenian and Alexandrian Neoplatonism and the Harmonization of Aristotle and
Plato, Leiden 2015, 29.
xiv Aristotle Transformed

mean that Ammonius agreed to hide some devout enthusiasm for theurgy.4 I
take it that Neoplatonists were not agreed on how much theurgy could achieve,
and that Ammonius may have been like Porphyry in doubting that it could
achieve all that much.
As regards my treatment of individual commentators, an important dictionary
of ancient philosophers has been compiled since 1989 by Richard Goulet in
French, which, proceeding in alphabetical order, has so far in seven volumes
reached Rutilius Rufus.5 In addition, there has been work since 1990 on early
commentators up to the second century ad, including one book on Aspasius in
Athens and another on commentaries on Aristotle’s Categories up to 200 ad.6
About Alexander I need only add that an inscription has been transcribed from
a stone he erected in his home town to honour his father, and it confirms the
conjecture that he did hold the Aristotelian chair in Athens.7 About Themistius,
I have already mentioned that he was a public figure. Indeed, he was a very
prominent civil servant who believed that the philosophy of Plato and Aristotle
was something important to promulgate in public life. But he came into conflict
with rival views about the public role of philosophy, including, as mentioned, the
views of his ex-pupil, the emperor Julian. To get a rounded picture of his
philosophy, it is therefore necessary to read not only his influential commentaries
on Aristotle, but also his public letters and speeches as public orator. These will
be relevant, when we reach Henry Blumenthal’s Chapter 5 below, and its
suggestion that Themistius was more Aristotelian than Platonist, but I shall say
much more about it in Aristotle Re-Interpreted.
Porphyry was a philosopher wide-ranging in a different way, not in public life,
but in the huge spectrum of philosophical topics he discussed, including ethics
and religion, and going far beyond his commentaries on Aristotle, although
these too had a transformative and enduring impact. The study of him has been
enhanced since 1990 partly by Andrew Smith’s edition of the fragments of his
work and partly by the decipherment and translation since then of a further

4
I am sorry to have given a contrary impression to my friend David Blank, in his ‘Ammonius
Hermeiou and his school’, in Lloyd Gerson (ed.), The Cambridge History of Philosophy in Late
Antiquity, Cambridge 2010, vol. 2, p. 659, and ‘Ammonius’, in Edward N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Fall 2014 Edition), URL = <http://plato.stanford.edu/archives/fall2014/
entries/ammonius/>.
5
Richard Goulet (ed.), Dictionnaire des philosophes antiques, Paris 1989–.
6
Antonina Alberti, Robert W. Sharples (eds), Aspasius: The Earliest Extant Commentary on Aristotle’s
Ethics, Berlin, 1999; Michael J. Griffin, Aristotle’s Categories in the Early Roman Empire, Oxford
2015. See also chs 26 to 35 of Richard Sorabji and Robert W. Sharples (eds), Greek and Roman
Philosophy 100 BC–200 AD, Bulletin of the Institute of Classical Studies Suppl. 94.2, 2007, 502–637.
7
Angelos Chaniotis, ‘Epigraphic evidence for the philosopher Alexander of Aphrodisias’, Bulletin of
the Institute of Classical Studies 47, 2004, 79–81.
 Introduction to the Second Edition xv

substantial fragment of the most important and massive of his commentaries on

Aristotle, his longer commentary on Aristotle’s Categories, invisible to the naked
eye, but made accessible through the use of multi-spectral lighting,8 which will
be reprinted in a shorter version in Aristotle Re-Interpreted.
A reconstruction will be offered in Aristotle Re-Interpreted of the identity of
the authors, some of them ancient, used in the twelfth century and again in the
thirteenth to compile the composite commentary on Aristotle Metaphysics
Books 6–13 which has previously been known as that of pseudo-Alexander. It
will also reveal the identity of the author of what has previously been treated as
an alternative version of parts of Alexander’s commentary on Books 1 and 2.
Some other updates to my Chapter 1 will emerge in the discussion of other
chapters.

Karl Praechter’s Chapter 2 needs very little updating. It was no mean feat to
survey in 1909 the huge corpus of texts of the ancient commentaries on Aristotle,
so recently completed by the Prussian Academy under the editorship of Hermann
Diels. Praechter explains the undertaking and its importance. He enumerates the
main commentators and explains their purposes and their idiosyncracies of
style, revealing the background of recording classroom teaching. He discusses
when and why they developed a standard set of opening questions to be
addressed. The text has none of the daring speculations of his 1910 article of the
following year,‘Richtungen und Schulen im Neuplatonismus’, whose controversial
reception is described by Koenraad Verrycken in Chapter 10 below. Among
useful postscripts it should be mentioned that it is no longer unclear that
Olympiodorus was a pagan, that the excessively late date for Stephanus of 610
ad will be very strongly challenged in Aristotle Re-Interpreted, while in Chapter
17 below Robert Browning re-dates Michael of Ephesus to a century later. We
cannot be quite sure that lecturers in Alexandria had a text of Aristotle and that
there was in Alexandria a school library which had another, but we do know that
in Rome Plotinus had texts read aloud at the beginning of his classes.

In Chapter 3, Hans Gottschalk surveys the commentators on Aristotle from the

first century bc to late in the second century ad, and some of their Platonist
opponents. He gives the most space to the first of them, Andronicus, persuasively

8
See Andrew Smith, Porphyrii Philosophi Fragmenta, Stuttgart 1993, and Riccardo Chiaradonna,
Marwan Rashed, David Sedley, ‘A rediscovered Categories commentary’, Oxford Studies in Ancient
Philosophy 44, 2013, 129–94.
xvi Aristotle Transformed

arguing that he worked in Athens without going to Rome, and telling something
of Andronicus’ philosophical comments on Aristotle and of his editorial work
on Aristotle’s school writings (as opposed to his works then better known, but
now largely lost, for publication outside the school). He rightly says that
Andronicus presented Aristotle as a system. As I indicated in commenting on
Chapter 1 above, his younger contemporary in Athens, Boethus, stimulated
enormous reaction from later commentators by his detailed and idiosyncratic
interpretation of Aristotle, fragments of which they recorded. So the description
‘scholasticism’, insofar as it suggests to us something rather dry, is not a
description we should now be likely to use, especially after the recent discovery
of new fragments of Boethus. But Aristotle Re-Interpreted will include a
contribution on some of Boethus’ achievement and further detail on the
commentators after him is supplied in other recent works listed above in note 6.
The only big matter of controversy concerns the two words ‘critical edition’ at the
opening of Gottschalk’s chapter, which could be taken for granted in 1990. It was
challenged by Jonathan Barnes in 1997.9 A critical edition is produced by
comparing different copies of the original in order to discover more closely what
the original may have said. Barnes argued powerfully that this is not what
Andronicus did. Indeed, if he did not go to Rome to examine the manuscript
there, it is even less likely that he did. One reaction was to think that this greatly
reduced the importance of Andronicus. But a contribution in Aristotle Re-
Interpreted will take up the other editorial activity including the presentation of
Aristotle’s school writings as a system. It was far more valuable, according to this
argument, to create a coherent canon of Aristotle’s voluminous school writings,
by joining or separating pieces and arranging them in a coherent order for
reading, than to seek the original wording in a critical edition.

In Chapter 4, Robert W. Sharples discusses Alexander of Aphrodisias, Aristotle’s

greatest defender, most famous as the commentator on his works. But Alexander
also wrote many other works, including among others three books of Questions,
one book of Ethical Questions, parts of a Mantissa or lightweight supplement on
the soul, On Fate, On Time which survives in a Latin translation from Arabic as well
as in a later version of the Arabic, all translated by Sharples. He finds comparatively
few signs of a teaching context in the commentaries on Aristotle. Only some have
introductory explanations suitable for students. The non-commentary works

9
Jonathan Barnes, ‘Roman Aristotle’, in Jonathan Barnes, Miriam Griffin (eds), Philosophia Togata II:
Plato and Aristotle at Rome, Oxford 1997, 1–69.
 Introduction to the Second Edition xvii

provide more evidence of teaching, and perhaps the clearest is something that is
found occasionally in the commentaries, but much more often in the other works:
expressions such as ‘and then it was said’, ‘One of those who were present said’.
Evidently, the orally expressed views of others are included in these passages. So we
may see in them evidence of Alexander’s classroom.
Sharples refers to Paul Moraux’s view that one of Alexander’s teachers, besides
Herminus and Sosigenes, was Aristotle of Mytilene. But since then in 2000, he
has co-authored with Jan Opsomer a different explanation of Alexander’s
remark, ‘I heard about the intellect-from-without from Aristotle things which I
preserved’.10 Alexander’s reference could have been to Aristotle, the founder of
his school more than five hundred years earlier, since it can mean: I heard an
interpretation of Aristotle’s ‘intellect from without’ which I preserved.
One of Alexander’s Questions discussed here by Sharples was about the nature
of universals and Sharples considered certain apparent discrepancies which
might lead one to believe that other people’s views had been interpolated.
Sharples made another contribution to that subject in 2005,11 while a different
view from his again will be published in Aristotle Re-Interpreted. But it is not that
one discussion supersedes another; rather each offers a different way of seeking
consistency so far as possible between statements made by Alexander himself at
different times, and readers need to know the alternatives.
I believe that this chapter requires no caveats other than that supplied by
Sharples himself, given his comprehensive familiarity with the details of
Alexander’s work.12

In Chapter 5, Henry Blumenthal’s paper needs to be studied since it offers the

fullest case yet for classifying Themistius as an Aristotelian thinker, rather than a
Neoplatonist, and that by a scholar who knew Neoplatonism very well indeed.
Themistius is taken out of strict chronological order, in order not to insist on
placing him among the Neoplatonists. However, he did in fact flourish in the
fourth century, some decades after the death of the first two Neoplatonists,
Plotinus and Porphyry, who are the subject of the next two chapters. Blumenthal’s

10
J. Opsomer, Robert W. Sharples, ‘Alexander of Aphrodisias, De Intellectu 110,4: ‘I heard this from
Aristotle’. A modest proposal’, Classical Quarterly 50, 2000, 252–6.
11
Robert W. Sharples, ‘Alexander of Aphrodisias on universals: two problematic texts’, Phronesis 50,
2005, 43–55.
12
His memorial volume edited by Peter Adamson, Bulletin of the Institute of Classical Studies 55, 2012
cites at pp. 11–18, besides his 6 volumes of translation, 23 articles on Alexander or pseudo-
Alexander, including a survey article: ‘Alexander of Aphrodisias: scholasticism and innovation’, in
W. Haase (ed.), Aufstieg und Niedergang der römischen Welt II , 36.2, Berlin 1987, 1176–1243.
xviii Aristotle Transformed

argument for his Aristotelianism is nuanced, because he allows, for example, that
Themistius prefers Plato’s view of the motion of elements up and down, and I
think this variability is very like Themistius. Ilsetraut Hadot has in 2015 put the
fullest rival case for Themistius being a Neoplatonist, and that equally needs to
be studied. She refers to me as having detached Themistius from the Aristotelian
school by citing the view that he requires Platonic Ideas to explain the apparent
generation of grubs without parents (‘spontaneous generation’).13 But
interpretation of that point in Themistius will be superseded in Aristotle Re-
Interpreted at least to the satisfaction of the original propounder whom I cited.
And in fact there are three views to consider, because I shall there argue in the
Introduction that Themistius is a member neither of the Neoplatonist nor of the
Aristotelian school. He was not only a commentator on Aristotle, but a leading
civil servant with dominant influence as President of the Senate of the imperial
capital of Constantinople, serving under six emperors, five of them Christian,
though Themistius was not Christian himself. He insisted, moreover, on
knowledge of both Plato and Aristotle in public life, and more particularly for
civil servants, for whom he ran a school. It was probably for the school that he
wrote the simplified commentaries on Aristotle, which he called paraphrases,
but which actually exercised enormous influence over subsequent commentators.
It is important to bring in, as Ilsetraut Hadot does, his speeches and letters on
public matters,14 in which he tells us, with full sympathy, how his father loved
both Plato and Aristotle and also how civil servants must inculcate in public the
ethics of both those philosophers. On the supposed parentless generation of
grubs, it seems that he does not after all require Platonic Ideas, but his commentary
tries to do justice to both Aristotle and Plato, so far as possible, with limited
success. He responds to the first two thinkers whom we classify as Neoplatonists,
Plotinus and Porphyry, but he would have thought of them simply as Platonists,
it being too early to foresee that they represented a new school. His discussion of
spontaneous generation draws on some Platonic thoughts of Plotinus. But when
it is difficult to harmonise Plato and Aristotle, I suspect he does not think it the
business of a teacher of civil servants to dwell on problems of harmonisation.
The development he did dislike was that of his former pupil Julian, the emperor
called by Christians ‘the Apostate’ for trying to reimpose pagan religion,
disagreeing openly with Themistius’ conception of the role of philosophy in

13
Ilsetraut Hadot, Athenian and Alexandrian Neoplatonism, 74.
14
Another relevant work of Themistius translated from Syriac will appear in the series Ancient
Commentators on Aristotle.
 Introduction to the Second Edition xix

public life, and starting to direct public life with reliance on what Themistius
called a ‘new song’. This was a version of Iamblichus’ Neoplatonism, which
included tremendous resort to priestly ritual, called ‘theurgy’ and the animal
sacrifice, which Porphyry had rejected for philosophers.
I myself find it hard to think of Themistius as a member of the Neoplatonist
school, a term he would not have understood, (any more than of the Aristotelian),
when he never took the opportunity of making his life under Julian easier by
indicating that he believed in the great powers of theurgy up to a point. He says
he had no philosophy teachers other than his father and father-in-law,15 which
would be unusual for a Neoplatonist. He also seems never to have recognised the
Neoplatonists’ supreme divinity, the One above their Craftsman-God (the
Demiurge) who fashioned the physical universe. Ilsetraut Hadot argues that he
could have been referring to the One when he comments on Aristotle’s application
to his God of Homer’s line: ‘let there be one ruler’. But it would be hard for anyone
to understand that as an attempt to impose on Aristotle a ruler above Aristotle’s
God, when Themistius had not mentioned such a higher ruler anywhere else.
What Themistius does have in common with the philosophers whom we call
Neoplatonists is that he wishes to harmonise Plato and Aristotle wherever
possible, as Ilsetraut Hadot makes very clear. Aristotle never lightly disagreed
with Plato, he says.16 But while her case seems very reasonable that that is a
necessary prerequisite for being a Neoplatonist, it is not sufficient: there were
harmonisers before Neoplatonism, and Themistius prefers harmony but does
not think it his business to argue for harmony at length.
None of this is to suggest that the case for Themistius’ Aristotelianism or
Neoplatonism is superseded. On the contrary, this is an ongoing controversy,
and it is important that readers should have access to all three views, including
the strong case made by Blumenthal in this chapter for a qualified Aristotelianism.

In Chapter 6, Pierre Hadot draws attention to a very striking passage of Plotinus,

which suggests that Plotinus, writing in Rome, qualified his attack in Enneads
6.1–3, on Aristotle’s treatment of substance (ousia) in his Categories for excluding
the true substances of the intelligible realm, Plato’s Ideas. At any rate, at 6.1.3.1–5,
Plotinus suggests there might be a solution available. Substance could be called
a single category, including not only matter, form, and the compound of the two
in the sensible realm accessed by sense perception, but also intelligible substance.

15
Themistius, Oration 21 (‘The examiner or philosopher’), 244.
16
Themistius, Oration 20 (‘In honour of his father’), 236.
xx Aristotle Transformed

This, he says, would be like saying that the genus (or clan) of Heraclids was a
unity, although they do not have anything in common except that they are
descended from one thing – Heracles. Then the intelligible substance (parallel to
Heracles) would be substance primarily and the other substances (parallel to his
descendants) secondarily and less so.
I am not sure that the talk of ‘from one thing’ (aph’ henos) here is a reference
to Aristotle’s idea of being derived from one thing and directed towards one
thing (aph’ henos and pros hen), because Aristotle explains that in terms of
definition. What a medical man and a medical instrument and a medical book
have in common is that they are all defined by reference to the medical art: one
is trained in, one used in, and one instructional in the medical art. Thus the thing
they have in common is a variety of relations to a single thing, the medical art.
Aristotle infers from this in Metaphysics 4.2, that, just as there can be one science
of the medical (or of the healthy) so there can be one science of substance, but
that science deals chiefly with the primary substance on which the other
substances depend (êrtêtai). Plotinus does not seem on the face of it to be talking
about definition. Possibly his idea is instead closer to what Aristotle says in
Metaphysics 12.10, 1075a11–19, that the highest good in the universe, as in an
army, lies in its order, but more in its commander to whom the order is due, so
that all things are ordered together (prostetaktai) towards one thing (pros hen).
At the beginning of 6.3.5, Plotinus offers another idea. If Aristotle’s account of
substance applies to intelligible substance, he says, that will perhaps be by
analogy and not in the same sense. These two ideas could be compatible, if the
first idea is that the applicability of substance and other categories to perceptibles
depends on their relation to an intelligible, while the second idea is that we can
know how substance or another category applies to an intelligible only by analogy
with its application to a perceptible. The dependence relation runs in one
direction; the knowledge by analogy in the other. Aristotle had been thinking of
the use of analogy in mathematics when he described analogy as saying: as A is
to B, so C is to D. If Plotinus is thinking that way here, he might mean that as
certain descriptions apply to sensible substance, so those descriptions taken in a
different sense apply to intelligible substance.
So far, Hadot has pointed to two solutions proffered by Plotinus to Aristotle,
but he also draws on a commentary by a later philosopher, Dexippus. What
Dexippus says in his commentary on Aristotle’s Categories at 41,1–3 is presented
as an interpretation of Plotinus. He is there talking about Aristotle’s account of
substance as not present in nor said of anything else, because it is a subject of
which other things are said or in which they inhere. Dexippus says, ‘If everything
 Introduction to the Second Edition xxi

leads up to this principle of substance, as being dependent (êrtêmena) on it [this

does use the word that Aristotle applied to definitional priority], [Aristotle’s]
sketch of substance [as not said of nor present in] can also reflect the first
principle of substance, from which [sensible] substance has fallen to the lowest
level’.
Dexippus moves from discussing Plotinus to discussing Aristotle in 41,12–
42,3, and continues to talk about definitional matters, when he says that sensible
substance will be homonymous with (lack the same definition as) intelligible
substance. But, as Hadot notes, he confuses the issue, by adding that Aristotle
applies the word ‘substance’ only metaphorically and analogically (as if these
were the same) to intelligible substances. The talk of analogy reminds us of
Plotinus’ second passage which does not mention any priority of intelligible
substance, but Dexippus seems actually to reverse the priority of definition that
he was earlier associating with Plotinus and to award that priority to sensible,
not to intelligible substance.
The brief solutions that Hadot locates in Plotinus and finds later reflected
with distortion in Dexippus show us Plotinus himself offering to Aristotle
possible solutions to the powerful objection he had raised against Aristotle’s
Categories, that it ignores intelligible substance. Hadot, as his title suggests, holds
that these solutions were more fully worked out in the interim in the lost
commentary on Aristotle’s Categories by Plotinus’ pupil in Rome, Porphyry.
Certainly Porphyry wanted to harmonise Plato and Aristotle, so he would have
been fully motivated to seize on any compromise between his teacher Plotinus
and Aristotle. So Hadot offers a number of arguments that Dexippus was merely
following Porphyry, drawing on Porphyry’s lost major commentary on Aristotle’s
Categories, albeit a Porphyry distorted by Iamblichus.
Initially, Hadot’s identification of Porphyry as the main source of Dexippus
might seem surprising for two reasons. First, Porphyry in his surviving short
commentary on the Categories, had offered a defence of Aristotle for ignoring
intelligible substances very different from that sketched by Plotinus and reflected
by Dexippus. Porphyry’s defence of Aristotle is that Aristotle’s Categories, which
he regards as for beginners,17 discusses the signification by words of things and
the original signification of words when they were first imposed was to perceptible
things. That is why Aristotle does not talk about Plato’s intelligible substances.18
Dexippus and Simplicius do record, but much more briefly, this answer by

17
Dexippus, in Cat. 40,21–2.
18
Porphyry, in Cat. 19,13–27.
xxii Aristotle Transformed

Porphyry.19 Hadot’s suggestion, however, rightly implies that Porphyry could

have offered a different solution in his much longer lost commentary on the
Categories for more advanced students, elaborating what Plotinus said. The
second surprise comes from a report in the later commentary of Simplicius on
Aristotle’s Categories at 2,9–25, that Iamblichus closely copied Porphyry’s lost
major commentary on the Categories, but that he inserted two kinds of addition
in almost every section. One addition was Pythagoreanism; the other was
Iamblichus’ so-called ‘intellective theory’, which concerned our present subject,
the insistence that Aristotle’s categories, including substance, apply to Plato’s
Ideas. Ought it not then to have been Iamblichus rather than Porphyry from
whom Dexippus, Iamblichus’ pupil, drew his further reflections on intelligible
substances? It may be thought that there is an answer to this in Simplicius’ next
words at 2,25–9, which concern Dexippus. Dexippus addressed only the
problems raised on the subject by Plotinus and added virtually nothing to
Porphyry and Iamblichus. Perhaps, then, Porphyry’s lost commentary had taken
up the subject that Dexippus garbles of analogy and metaphor. But had not the
‘intellective theory’, which Simplicius ascribes to Iamblichus, added its own
views on such subjects as analogy and metaphor?
Pierre Hadot’s important contribution is his detective work on Plotinus’
proffered solutions and his stress on their subsequent influence. As regards
Dexippus, the subject has been taken forward by Concetta Luna.20 She prioritises
Iamblichus’ commentary as the source of Dexippus and Simplicius, though it
draws in turn on Porphyry’s lost commentary, but with its own distinctive
additions. But she makes Simplicius the superior evidence, as giving far more
detail than Dexippus, who telescoped the arguments, and as being more faithful
to Iamblichus. Simplicius explains the contribution of the Pythagorean Archytas,
who was brought in by Iamblichus, which is what makes Iamblichus look like
Simplicius’ source. But given the overlaps with Dexippus’ very reduced report, it
is more economical, she argues, to suppose that Dexippus and Simplicius both
drew on Iamblichus.21 The problem of applicability to intelligibles, she adds, was
discussed by Dexippus and Simplicius under at least two heads, one concerning
intelligible substance and one concerning all ten of Aristotle’s categories.
Simplicius’ account favours a version of what Pierre Hadot showed to be Plotinus’

19
e.g. Dexippus, in Cat. 42,17–18; 44,32–45,2; Simplicius in Cat. 73,28–74,22; 76,17–23; 82,1–6.
20
Concetta Luna, Commentaire sur Les Catégories d’Aristote, Chapitres 2–4, Paris 2001, 750–89.
21
She cites as coming from Porphyry in the passage she discusses only Dexippus, in Cat. 75,5–8. John
Dillon, in his translation of Dexippus, p. 11, n. 12, finds dependence on Porphyry outside that
passage at 26,9ff.; 43,10ff.; 65,14ff.
 Introduction to the Second Edition xxiii

argument that sensibles are derived from and directed to (aph’ henos, pros hen)
an intelligible, on which they depend (exêrtêsthai).22 This intelligible substance is
not merely a genus, as Plotinus thought, but a source (arkhê) of being, as Aristotle
and Archytas are said to have recognised,23 although Dexippus presents this last
as Plotinus’ own view, and Aristotle’s.24 In the reverse direction, it is appropriate
for Aristotle to think that we can pass from sensibles to intelligibles by analogy.25
Simplicius does not here, like Dexippus, conflate analogy with metaphor, and
indeed he had pointed out that Porphyry distinguished these.26 Much later in his
commentary, however, Simplicius takes a view that Plotinus and Dexippus had
mentioned,27 but which is not that of his source for interpreting Aristotle, that
not all categories are applicable to intelligibles in the same way. The category of
Being-in-a-position is not applicable, except metaphorically,28 and the categories
of Posture, Passivity, and Relative are not applicable, except by analogy.29
In the light of this, we can see how much fruit was harvested from the first
steps which Hadot uncovered in Plotinus, and indeed Plotinus turns out to have
anticipated even more of the options later used. The next chapter is complementary
to Pierre Hadot’s work in this chapter.

Sten Ebbesen in Chapter 7 is concerned with Porphyry’s influence on medieval

philosophy of language, and he sees him as the force behind its concern with
concepts. In a tradition which is described as starting as early as the commentator
Boethus (not Boethius) in the first century bc , Simplicius says that many
commentators took Aristotle’s Categories to be concerned with three matters: (1)
simple, primary, and generic spoken sounds, (2) insofar as they signify (3) things,
and Porphyry is listed as one who agreed. Indeed Porphyry does say in his
surviving shorter commentary on Aristotle’s Categories that the book is about
spoken sounds insofar as they signify things.30 Moreover, the infinity of beings
and of expressions that signify beings, or vocal sounds that indicate them, are
included under ten genera.31 Only some commentators treated the signification
as introducing a third entity, concepts, into the account besides spoken sounds

22
Simplicius, in Cat. 75,3.
23
Simplicius, in Cat. 76,25–77,11.
24
Dexippus, in Cat. 40,25–41,16.
25
Simplicius, in Cat. 73,35–74,3; 74,22–8.
26
Simplicius, in Cat. 32,19–33,21, referring to Porphyry, in Cat. 66,29–67,22.
27
Plotinus, Enneads 6.1.1.21–2; Dexippus, in Cat. 42,18–31.
28
Simplicius, in Cat. 337,15–17.
29
Simplicius, in Cat. 205,28.
30
Porphyry, in Cat. 58,4–7.
31
Porphyry, in Cat. 58,12–15.
xxiv Aristotle Transformed

and things. Simplicius was one, for, he says in the same breath that, insofar as
things are signified by spoken sounds, the Categories is giving instruction not
only about the things signified, but also about concepts (noêmata).32 Aristotle
himself mentioned concepts in a semantic context at the beginning of his On
Interpretation, 16a10, after first calling them effects in the soul. He says at 16a3–8
that spoken sounds are symbols of effects in the soul, and written marks are
symbols of spoken sounds, while the effects in the soul (concepts) are likenesses
of actual things. But Porphyry foregoes the opportunity to connect signification
explicitly with concepts.
Ebbesen finds concepts playing a role in three commentators earlier than
Simplicius: first Porphyry, then Dexippus, followed by Ammonius. For Porphyry,
Ebbesen finds only two possible explicit invocations of concepts in semantics,
but Dexippus may have been drawing on Porphyry, since we have seen in
connexion with Chapter 6 that Dexippus is likely to have learnt a lot about
Porphyry from his teacher Iamblichus, who, according to Simplicius closely
copied Porphyry’s lost major commentary on Aristotle’s Categories, albeit with
his own additions.
Both references to Porphyry are in reports by Boethius. Ebbesen finds
Porphyry talking of mental discourse, which is probably a reference to Porphyry’s
associating the effects in the soul of Aristotle On Interpretation 16a3–4 with
names and verbs composed in the mind as something different in kind from
those spoken or written.33 Porphyry, however, again foregoes the opportunity of
calling the effects in the soul ‘concepts’, although the later commentator
Ammonius does in a similar passage, as Aristotle had originally.34 In a second
report by Boethius, Porphyry is interested in the hearer waiting for a concept
(intellectus in Latin, standing for Greek noêma), when the speaker attends to one
thing (res), and hearer and speaker eventually agree on the same concept.35 But
here reference to concepts and things alternates, so Porphyry is not consciously
stressing concepts as opposed to things.
The evidence in Dexippus is much more extensive and this may throw light
on Porphyry, because we have seen that, according to Simplicius, on this subject
Dexippus mostly addresses problems raised by Plotinus and that he added little
to Iamblichus (his teacher) and Porphyry. This was taken above to suggest that

32
Simplicius, in Cat. 13,11–18.
33
Boethius, in Int.2 29,29–30,10 Meiser, apparently alluded to in Sten Ebbesen, ‘Porphyry’s legacy to
logic: a reconstruction’, Chapter 7 below, p. 153.
34
Ammonius, in Int. 22,14.
35
Boethius, in Int.2 40,9–28 Meiser.
 Introduction to the Second Edition xxv

Porphyry himself may in his lost work have discussed Plotinus on the applicability
of Aristotle’s categories to intelligible substances. Certainly, Porphyry repeats in
his Isagôgê Plotinus’ example of the clan of Heraclids forming a genus with
Heracles as the primary Heraclid explaining the membership of the others.36
From Dexippus we learn that one of Alexander’s teachers, Sosigenes, had put
up a case, without deciding on it, that the objects of speech or what is signified
(ta legomena, ta sêmainomena) are things.37 But Dexippus thinks that concepts
(noêmata) are signified by speech, and he says that it was the view of the ancients
that the only things signified are concepts. In his own view, they are at least what
is primarily signified. Truth and falsity also come from the combination of
concepts and ambiguity depends on them.38 Ammonius later concurs about the
source of truth and falsity.39
Although Ebbesen’s own detective work has not, I believe, found a very strong
emphasis on concepts in surviving texts about or by Porphyry, even where one
might expect it, the evidence from Dexippus does seem to me significant.
Dexippus could have found his somewhat distorted information in Iamblichus’
reprise of an expansion on Plotinus written by Porphyry in his lost major
commentary on the Categories.

In Chapter 8, H.D. Saffrey brings us to the Neoplatonist Syrianus, the second

head in the fifth century of the re-established Platonist school in Athens, and
examines his attitude to Aristotle. Syrianus expresses the greatest admiration for
Aristotle and calls him the benefactor of humanity. But in his commentary
which survives on four books of Aristotle’s Metaphysics, he takes a different view
about Aristotle’s Books 13 and 14. There Aristotle offered an onslaught on the
mathematical interpretation in Pythagoras and Plato of the principles of the
universe. Syrianus thought it his duty to protect inexperienced students from
being carried away by contempt for divine realities and divinely inspired
philosophy. Syrianus’ reply was hard-hitting and sharp. An example which I have
praised elsewhere comes from Syrianus’ commentary on Book 13 and is
devastating, though quite respectfully put. Aristotle thought that we gain our
geometrical concepts from the imperfect shapes we perceive by sense perception.
But how then, Syrianus asks, do we know how to correct the imperfect shapes
into the perfect ones of geometrical thinking? Aristotle’s sense perception cannot

36
Porphyry, Isagôgê 1,18.
37
Dexippus, in Cat. 7,4–8,24.
38
Dexippus, in Cat. 7,1–2; 9,22–10,10.
39
Ammonius, in Int. 18,2–12; 24,29–30.
xxvi Aristotle Transformed

do that, and the best answer is the Platonist one. The human mind contains the
rational principles (logoi) of perfect geometrical shapes,40 and, Syrianus says,
these logoi are not identical with Plato’s Ideas.41 As his pupil Proclus makes clear,
they are rather reflections (emphaseis), copies (eikones), and projections
(probolai) of Ideas.42 How could Aristotle’s sense perception provide the
necessary correction?
Syrianus’ attitude of respect, undiminished by the need for sharp criticism, is
very interesting to compare with his pupil Proclus’ very much more extensive
criticism of Aristotle, which will be the subject of a masterly study by Carlos
Steel in Aristotle Re-Interpreted. Ilsetraut Hadot has reviewed five studies of
Syrianus’ polemics, including Saffrey’s and that of John Dillon with Dominic
O’Meara in the Ancient Commentators on Aristotle translation of Syrianus’
commentary on Books 13–14 (pp. 11–20 of the Introduction), as well as another
treatment by O’Meara and treatments by Michael Frede and Loredana Cardullo.43

My Chapter 9 on infinite power starts from the criticisms of Aristotle by Syrianus

and Proclus, of which Chapter 8 discussed a selection. One criticism made by
both objectors concerned Aristotle’s argument about the need for infinite power.
A universe of finite size, such as was accepted by both Plato and Aristotle, cannot
contain the infinite power needed to explain its infinitely enduring motion. So
God’s infinite power must be invoked to explain it. In Proclus’ version, the reply
complains that Aristotle failed to see that, by the same token, God’s infinite
power is needed to explain the universe’s infinitely enduring existence. In other
words, God is not just the mover, as Aristotle thought, but the creator of the
universe, and beginninglessly its creator as beginningless. Proclus’ pupil,
Ammonius, was a harmoniser, and made just the opposite assessment of
Aristotle, who, he said, intended exactly the conclusion desiderated. Ammonius
wrote a whole book ascribing to Aristotle many arguments for God as
beginningless creator of a beginningless universe, including the argument based
on the need for infinite power to produce infinite existence.
Although few will now believe that this was Aristotle’s intention, Chapter 9
traces the enormous effect of Ammonius’ interpretation in the history of
philosophy. Two versions of the argument became available, because Ammonius’

40
Syrianus, in Metaph. 95,29–38
41
Syrianus, in Metaph. 105,36–106,2.
42
Proclus, Elements of Theology Prop. 194, On Plato’s Parmenides 4, 896,25; 897,15–16.22–3; 896,23–7;
982,4–9.
43
Ilsetraut Hadot, Athenian and Alexandrian Neoplatonism, 107–15.
 Introduction to the Second Edition xxvii

pupil Philoponus, being a Christian, adapted it to draw the conclusion, contrary

to Aristotle’s, that the very nature of the universe limits its power and therefore
its duration – unless indeed God overrides that nature, but God cannot override
its natural incapacity to be beginningless. Versions of the arguments about
infinite power passed through the Arabic-speaking world and on to Latin-
speaking Christians and to Jews, with Christians and Jews, but not all Muslims,
preferring some version requiring a beginning of the universe. The Aristotle that
was passed on was a rich compilation transformed by the different preoccupations
and allegiances of different commentators.
Some updates to what I said will be provided in Aristotle Re-Interpreted. In
brief, it will be argued by one contributor that the Muslim philosopher al-Farabi
was not the author of the Harmonisation of Plato and Aristotle, and so can be
seen as accepting Aristotle’s belief in the eternity of the world, as did the other
later so-called Muslim ‘falāsifa’ or philosophers, like Avicenna and Averroes, as
opposed to the so-called theologians, like the later Ghazali, although both sides
were in our sense philosophers.
I shall argue in Aristotle Re-Interpreted that there was no first edition of
Philoponus’ commentary on Aristotle’s Physics in 517 ad. That year is the date
only of composing the commentary on Physics Book 4, chapters 10–14, and I am
not aware of it having received a second edition.
Finally, in Aristotle Re-Interpreted it will be argued in one chapter that
Philoponus himself, not Stephanus, is the author of the commentary ascribed to
him on Aristotle On the Soul, Book 3, and that Stephanus was nonetheless the
pupil of Philoponus, whether or not he ever acted as his editor.

Chapter 10 is again about Ammonius of Alexandria and the rejection of Karl

Praechter’s daring thesis that the Alexandrian school did not recognise the
Neoplatonists’ highest divinity, the One. That was first rejected by A.C. Lloyd for
Simplicius, the Athenian who was trained in Alexandria, and for the commentary
by Asclepius from the voice of Ammonius, and then in more detail by Ilsetraut
Hadot for the Alexandrian Hierocles and again for Simplicius. Koenraad
Verrycken here extends the rejection by examining what Ammonius’ theology
actually was.
In a recent review of Verrycken’s contribution, Ilsetraut Hadot welcomes his
rejection in relation to Ammonius of Praechter’s thesis about Alexandrian
omission of the One.44 But Verrycken offers some qualification, which she

44
I. Hadot, Athenian and Alexandrian Neoplatonism, 26–8 and 143ff.
xxviii Aristotle Transformed

opposes. He thinks that Ammonius did simplify the higher theology of his
Athenian teacher Proclus with its extensive hierarchies. Ammonius hints that he
recognises these hierarchies, but is not inclined to pursue them, and this is ‘in
line with his preference for the study of Aristotle’ and with ‘a relative decrease of
interest in Plato’. Ammonius sought to harmonise Aristotle more with Plato by
‘Neoplatonising’ Aristotle’s theology and Aristotelising Plato’s, thus making the
two philosophers closer to each other – closer, I presume than had his teacher
Proclus, who emphasised, to Aristotle’s disadvantage, his disagreements with
Plato. Indeed, Ammonius Aristotelised the Neoplatonic system as a whole.
‘Probably Ammonius was just not the type of the true Platonic hierophant by
Athenian standards’. Alternatively, Ammonius ‘may have thought it was useless
to rival Proclus in Platonic exegesis’. In any event, ‘the study of Aristotle became
more important than the study of Plato’. Hence in this reduced sense, there was
a distinctive Alexandrian Neoplatonism different from that in Athens.
Much of this does not go beyond the correct statement of Ammonius’ pupil,
the future Athenian head, that Ammonius was more practised (mallon exêskêto)
in Aristotle’s thought.45 Verrycken may go further when he says that Ammonius
‘did not see Plato as the real thing and Aristotle merely as a propaedeutic’. Even
so, I think Verrycken’s position is not quite captured by Ilsetraut Hadot’s
formulation, which she rejects, that Ammonius did not believe in the superiority
of Plato to Aristotle, or that he denied the inferiority of Aristotle. What Ilsetraut
Hadot does rightly point out is that Asclepius’ commentary from the voice of
Ammonius on Aristotle’s Metaphysics says that the scope of Aristotle’s Metaphysics
is to speak of ‘all beings insofar as they are beings’,46 and this immediately
excludes discussion of, though not of course the reality or importance of, Proclus’
divine hierarchy above the level of the divine Craftsman, the Demiurge. That it
excludes discussion does not, however, I think address the question why
Ammonius did not write commentaries on Plato’s theological texts, which would
have required such discussion. Verrycken makes several useful suggestions
about that. He certainly rightly agrees and so, I believe, does Ilsetraut Hadot, that
this was not a reluctant concession to the Christian authorities to forego
something that Ammonius would have preferred to discuss. Nor, I think, does
Ammonius reveal the slightest reluctance to forego discussion of the Chaldaean
Oracles and the Orphics, which would have brought in the pagan priestly powers
of theurgy, and to which Plato was seen by Proclus as a propaedeutic, which is

45
Damascius, Philosophical History fr. 57c Athanassiadi.
46
Asclepius from the voice of Ammonius, in Metaph. 2,10–11.
 Introduction to the Second Edition xxix

not to say that Proclus would have wanted to use a word like ‘inferior’ about
Plato. I am therefore inclined to think that Verrycken has by and large made a
good case, even though it remains a controversial one.

Verrycken’s discussion in Chapter 11 of Ammonius’ pupil Philoponus was of the

highest originality. It was based on a scrutiny of Philoponus’ texts more intensive
than any earlier one that I know of. Verrycken found apparently contradictory
statements within Philoponus’ commentaries, often within a single commentary,
and, to explain that, he offered an extremely daring and comprehensive thesis,
although with all due modesty recognising that his thesis had not dealt with all
the details. What this meant was that anyone who offers a different account
needs to know the passages which Verrycken highlighted in this chapter, in order
to secure their alternative interpretation.
Verrycken’s hypothesis was that Philoponus started as a Christian, but was
converted by Ammonius to pagan Neoplatonism, as revealed in his earlier
commentaries on Aristotle. After Ammonius’ death, he reverted to Christianity
and used the intensive knowledge he had gained of Ammonius’ Neoplatonism to
reject certain Neoplatonist views and re-edited his earlier commentaries in such a
way as to leave contradictory interpretations standing within the same commentary.
Verrycken upheld his view in 2010 in his contribution to The Cambridge
History of Philosophy in Late Antiquity, edited by Lloyd Gerson.47 I recorded, also
in 2010, with one important omission, the fairly extensive disagreement with the
original chapter that had arisen in the interim after due pause for thought.48 An
alternative chronology for Philoponus was offered by Pantelis Golitsis in 2008,49
and in Aristotle Re-Interpreted I shall discuss these two major hypotheses. In
doing so, I found that I had repeatedly to consider whether the new hypothesis
could take into account the findings that Verrycken’s earlier hypothesis had
uncovered. So whether or not one adheres to Verrycken’s conclusion, it is
necessary to keep available and consult the findings on which he based it.

47
Koenraad Verrycken, ‘John Philoponus’, in Lloyd Gerson (ed.), The Cambridge History of Philosophy
in Late Antiquity, Cambridge 2010, vol. 2, pp. 733–55.
48
Richard Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science, Second Edition, Bulletin
of the Institute of Classical Studies Suppl. 103, 2010, 14–18. I apologise again for having misattributed
to his reviewer the important queries expressed by Clemens Scholten in his Antike Naturphilosophie
und christliche Kosmologie in der Schrift “De Opificio Mundi” des Johannes Philoponos, Berlin 1996,
and I had not yet encountered Scholten’s further queries in his Johannes Philoponos, De Aeternitate
Mundi, Über die Ewigkeit der Welt I, Turnhout, 2009.
49
Pantelis Golitsis, Les commentaires de Simplicius et de Jean Philopon à la Physique d’Aristote, Berlin
2008.
xxx Aristotle Transformed

In Chapter 12, Ilsetraut Hadot discusses three major questions about Simplicius,
having been in the forefront of bringing them, more than anyone else, to our
attention. All three questions are large, controversial, and difficult to settle. Are
the Alexandrian and Athenian schools of Philosophy united in their viewpoints?
Where did Simplicius, the commentator trained in Alexandria, but teaching in
Athens, go, after the Athenian philosophers in exile gave up their refuge with
King Khosroes I of Persia? Who wrote the commentary on Aristotle On the Soul
attributed to Simplicius? Not only is Ilsetraut Hadot’s knowledge of Simplicius
comprehensive, but she returns with new evidence, in response to challenges, so
that it is difficult to say when an issue has been finally decided, although different
readers may have reached different provisional decisions.
As regards the first question, she had in 1990 found no difference between the
Athenian and Neoplatonist schools in the first 40 pages of Simplicius’ commentary
on Aristotle’s Categories, after comparing seven other commentaries. Twenty five
years later in her 2015 book, she still finds no difference between the two
schools.50 Certainly both schools believed in the One as a divinity higher than
the divine Craftsman or Demiurge of Plato, who fashioned our universe. Further
examples that have emerged above are that both schools had some interest in the
priestly ritual called theurgy, and both believed in some harmony between Plato
and Aristotle. Ilsetraut Hadot speaks of a uniformity of tendency between the
two schools. She allows that doctrines may differ, but they differ only between
individuals, not between schools. Certainly, it is hard to find tendencies dividing
the schools, but there do seem to be sharp differences between individuals in the
degree to which they follow tendencies and I think this is particularly true of
differences between the Alexandrian Ammonius and his Athenian teacher
Proclus, even though Ammonius expresses the proper indebtedness to his
teacher, and expresses it even when, as in his commentary on On Interpretation,
his views on divine names differ dramatically from those of Proclus. I have
suggested above (and Ilsetraut Hadot might not disagree) that the Alexandrian
Ammonius’ silence about theurgy, in contrast with the enthusiasm of his
Athenian teacher Proclus, need not have been secured against Ammonius’ will.
He might, as I have said, have been like an earlier important Neoplatonist,
Porphyry, who thought that the powers of theurgy were limited and that it could
not bring us to the gods. Verrycken in Chapter 10 above hazarded that Ammonius
had no interest in commenting on Plato’s theology. In a paper of 2004 to be
reprinted in Aristotle Re-Interpreted, Robbert van den Berg argues that

50
I. Hadot, Athenian and Alexandrian Neoplatonism.
 Introduction to the Second Edition xxxi

Ammonius attached no theological significance to divine names, in extreme

contrast with Proclus who used their names to illustrate the priestly power of
theurgists. I think what is true of theurgy would also be true of the different
degrees of commitment as between Ammonius and Proclus concerning the
harmony of Plato and Aristotle. Ammonius was a harmoniser, who had to
contradict some of Proclus’ interpretations, for example the one on infinite
power discussed in Chapter 9, in order to make the case for harmony. But
precisely because individuals differed so much, one cannot say that a lowered
interest in theurgy or a heightened interest in harmony characterised the whole
Alexandrian school: a high interest in theurgy has been shown by Ilsetraut
Hadot herself for the earlier Alexandrian school51 and the later Alexandrian
Olympiodorus clearly had faith in theurgy. The Athenian school is not uniform
either, seeing that Damascius seems to want theurgy to be conducted only by
priests, not philosophers.52 Meanwhile, as for the harmony of Plato and Aristotle,
the Athenian Syrianus was surely more respectful of Aristotle than Proclus, and,
as Gerd van Riel has shown,53 Damascius was ready to prefer Aristotle’s definition
of pleasure to Plato’s. If there is a tendency in the Alexandrian school from
Ammonius onwards, it is an absence of commentary on Plato’s theological
dialogues, not to mention the theurgical Chaldaean Oracles. But while, as
Verrycken suggests, that may have been congenial to Ammonius, it had in any
case become a matter of safety in Alexandria, as Ilsetraut Hadot notes,54 since
even Olympiodorus only explains his faith to his Christian students in Christian
terms, so that they can understand it; he does not advocate it.
On a second issue, everyone is very indebted to Ilsetraut Hadot for drawing
attention to the important work of a scholar of easterly languages and regions,
Michel Tardieu, and strongly fortifying it. His hypothesis is that Simplicius in
exile moved to the town of Harrân, just inside the present border of Turkey with
˙
modern Syria, after the Athenians left the protection of King Khosroes’ palace in
Ctesiphon. She also takes Simplicius’ commentaries to have been written in this
post-Ctesiphon period and usefully argues for their relative dating. By 2007,55
she had assembled and strengthened a very large number of arguments in favour
of Harrân as the venue, although most of them had met detailed opposition. One
˙

51
I. Hadot, Athenian and Alexandrian Neoplatonism. 1–25.
52
Damascius, Life of Isidore or The Philosophical History fr. 88 Athanassiadi.
53
Gerd van Riel, Pleasure and the Good Life: Plato, Aristotle and the Neoplatonists, Leiden 2000.
54
I. Hadot, Athenian and Alexandrian Neoplatonism, 22–3.
55
I. Hadot, ‘Dans quel lieu le néoplatonicien Simplicius a-t-il fondé son école de mathématiques, et où
a pu avoir lieu son entretien avec un manichéen?’, The International Journal of the Platonic Tradition
1, 2007, 42–107.
xxxii Aristotle Transformed

of them, however, seems to me to be different from the other arguments and to

makeHarrân a stronger option than any other. Before the Athenian philosophers
˙
left the safety of Ctesiphon, King Khosroes in 532 included a clause in their
favour in a treaty with the Christian emperor Justinian in Constantinople, who
had stopped their pagan teaching in Athens in the first place. The clause required
that the Athenian philosophers were to be allowed to return to their abodes in
Byzantine territory and to live without harassment from then on. There were
several features making Harrân a suitable place for Simplicius to choose. It was
˙
near enough for safety to the then Persian border, and the absence of Christian
persecutors is suggested by Khosroes’ later decision in 549 when he was able to
exact tribute, to exempt it from paying because it had kept the old religion and
not adopted Christianity. But the resulting comparative safety was not on its own
unique to Harrân. What is more unusual is that Simplicius records his experience
˙
on the River Aboras (or Khaboras) of boats carrying heavy loads with the
support only of inflated animal skins. The word is ‘experienced’, not ‘saw’, for he
might have seen such boats at Ctesiphon, and why does he mention the much
smaller river of Aboras? Because it was only 40 kilometers from Harrân and he
˙
could very easily have reached that spot.
Ilsetraut Hadot’s third issue is the authorship of the commentary attributed to
Simplicius on Aristotle’s On the Soul. In three publications, Carlos Steel, the first
time in conjunction with Fernand Bossier, suggested that the true author was a
different Athenian philosopher, Priscian, who also recorded the answers, shortly
to be published in English translation, given to Khosroes’ philosophical and
scientific questions during the stay at Ctesiphon. Ilsetraut Hadot supported
Simplicius’ authorship of the commentary on On the Soul a number of times,
including in the present chapter, but most fully in 2002, with a further summary
in 2014.56 Something else also turns on this question because the disputed author
of the commentary makes three cross-references to his own commentary on the
Metaphysics, which we do not have, and to his epitome of Theophrastus’ Physics.
Arabic sources credit the first to Simplicius, but as regards the second, what we
do have is a paraphrase (metaphrasis) of Theophrastus on perceptual faculties by
Priscian, and this sounds very much as if it could be part of the epitome of
Theophrastus’ Physics. The question has also been asked whether the Arabic
ascription of the Metaphysics commentary to Simplicius is unreliable because

56
I. Hadot, ‘Simplicius or Priscianus? On the author of the commentary on Aristotle’s De Anima
(CAG XI ): a methodological study’, Mnemosyne 55, 2002, 159–99; Le néoplatonicien Simplicius à la
lumière des recherches contemporaines, Un bilan critique, Sankt Augustin, 2014, 187–218.
 Introduction to the Second Edition xxxiii

based simply on the three cross-references to a Metaphysics commentary in the

commentary On the Soul whose authorship is now disputed.
To confess my own changing opinions, I was quite early inclined to think that
the text was not by Simplicius, but I recognised that Ilsetraut Hadot had arguments
to explain the differences from his other works, and moreover I was not at first
convinced that the alternative must be Priscian. For me this changed in 2012, when
Steel published his case for Priscian in his translation of the last part, which we
agreed to ascribe to ‘Simplicius’, in inverted commas, so as not to pre-empt any
decision on the controversy. It appears that Ilsetraut Hadot's latest publications of
2014 and 2015 had not yet been able to take account of Steel’s latest case.57
I need hardly say how important it is to record her views on the three issues
she discusses, all the more because all three issues are still matters of controversy.

Henry Blumenthal’s Chapter 13 comes from the distant days of 1973, when, as he
says, if the commentators were read at all, it was only with the hope of throwing
light on Aristotle. Of course they do all throw light because of their extraordinarily
close attention to the text. But Blumenthal is warning that the Neoplatonist
interpretations of the text tend to see Aristotle’s philosophy as a manifestation of
Neoplatonic truth.
As in his 1979 article in Chapter 5 above, he strongly distinguishes Themistius
from this Neoplatonist tendency. The aims that Themistius states at the opening
of his commentary on Aristotle’s On the Soul are to follow Aristotle, to clarify,
and if necessary to expand him.58 There is no mention of truth, of Plato, or of
Platonists. Simplicius (rightly or wrongly) treats it as an exception, when
Themistius prefers Plato: ‘Although Themistius prefers the views of the Peripatos
in most things, in this he seemed to favour those of Plato’.59 Aristotle defines the
soul in On the Soul 2.1 as the form (and first actuality) of a natural body with
organs (organikon) that has life potentially. In the commentary on this passage
by ‘Simplicius’ (Blumenthal rejected its attribution to Priscian), organikon is
taken to mean something quite different from ‘having organs’, and more in line
with Plato. ‘Simplicius’ understands that the soul does not form a single entity
with the body, as Aristotle is nowadays taken to hold, but rather uses the body as
its instrument and makes it to be an instrument (organon). Moreover, ‘Simplicius’
distinguishes two layers of soul, one which uses the body, and a lower one which

57
I. Hadot, Le néoplatonicien Simplicius, 198–9; Athenian and Alexandrian Neoplatonism, p. 163,
n. 140. Cf. Introduction to Carlos Steel's Simplicus, on Artistotle On the Soul 3.6–13, London 2013.
58
Themistius, in DA 1,2–5.
59
Simplicius, in Cael. 69,9.
xxxiv Aristotle Transformed

gives it form and life. Philoponus’ commentary, by contrast, taken from the
seminars of Ammonius, preserves Aristotle’s sense: body with organs, even
though Philoponus states Neoplatonic views (we might add: derived from
Ammonius). Themistius also takes organikon in Aristotle’s sense, and denies the
sort of view later endorsed by ‘Simplicius’ (or Priscian) that there are two different
entities, one giving life and the other using the body.
Aristotle makes a puzzling statement about his reference to first actuality
(entelekheia). The human soul is not a second actuality, because it is not at all
times active, but it is a first actuality because it has already provided the
potentially living body with life and it is the person’s already implemented
capacity for activity. But why does Aristotle say ‘It is not yet clear whether the
soul is the actuality of the body in the same way as a sailor is of the ship’?60
‘Simplicius’ and Philoponus are puzzled for opposite reasons. To ‘Simplicius’ as a
Platonist, it seems obvious that the soul which uses the body is separable from
the body as is a sailor from a ship. But he construes Aristotle’s doubt as meaning
that the layer of soul which uses the body is inseparable from the body when
using it, but is separable because it can refrain from using the body. Philoponus,
by contrast, knows that Aristotle thinks the soul does not survive separation
from the body and finds the passage difficult to interpret for the opposite reason.
He suggests that the rational soul may be separable from the body and not the
body’s actuality when it no longer attends to it, and the point is called still unclear
only because rational soul is not discussed by Aristotle until Book 3. Themistius,
whom Philoponus (or his teacher Ammonius) may have copied, is somewhat
similar, except that he substitutes ‘intellect’ (nous) for ‘rational soul’.
Three updates are required for this discussion. One is that some of Philoponus’
commentaries on Aristotle, including this one, are originally based on the
seminars of his teacher Ammonius, so that some comments are those of
Ammonius and some those of Philoponus. Another update has been mentioned
above. It is that, although Blumenthal accepts in passing the view that Philoponus’
commentary on the third book of On the Soul is really by Stephanus, there will
be two chapters in Aristotle Re-Interpreted, one arguing the case for the
commentary on Book 3 really being by Philoponus, and the other on Stephanus,
suggesting that he was Philoponus’ pupil. A final comment is that, as emerged in
connexion with Chapter 12, there is controversy about Blumenthal’s view that
Simplicius, not Priscian, is the real author of the other commentary on On the
Soul which he discusses.

60
Aristotle, DA 2.1, 413a8–9.
 Introduction to the Second Edition xxxv

L.G. Westerink’s Chapter 14 comes from a classic work, and indeed from the
update of a classic work, since the chapter is his own 1990 English revision of an
excerpt from his Anonymous Prolegomena to Platonic Philosophy, Amsterdam
1962, a book which appeared in a second edition in French from Les Belles
Lettres also in 1990. The chapter provides an exhaustive comparison of the
different Introductions to Aristotle and to Porphyry’s Introduction (Isagôgê) to
Aristotle’s logic, as introduced by Ammonius and continued in his school.
Westerink was in the forefront of identifying two of the last Greek writers of
introductions in the school: Elias, and the one he re-named pseudo-Elias, while
explaining his relation to Elias and David.61 It was Ammonius’ achievement to
formalise these introductions to Aristotle and Porphyry, which were to be drawn
on by the Christian Greek fathers of the Church, Maximus the Confessor and
John of Damascus, by Syriac-speaking Christians, and through them eventually
in Arabic by Muslims and Christians, especially by the Muslim Ibn al-Tayyib.
The introductions by David were translated from Greek into Armenian, a
language in which they exerted enormous influence, and the Definitions and
Divisions of Philosophy have been translated into English from Armenian by
Bridget Kendall and Robert Thomson and will shortly be translated from the
original Greek version for comparison.
Ammonius seems to have been the one who introduced six types of
introduction for beginning students of Aristotle. He was the first to write a
commentary on Porphyry’s Introduction (Isagôgê) to Aristotle’s logic and to
precede that with a commentary on it and an introduction to the commentary
and an introduction to philosophy as a whole. These four introductions preceded
an introduction to Aristotle, and to his Categories (the first of his works in the
curriculum). But for the introduction to Aristotle, Ammonius would have been
following a version of the ten questions already laid down by his teacher Proclus
as needing to be asked first.62 The introductions to philosophy in Ammonius’
school were particularly interesting. They included six definitions of philosophy
attributed to earlier philosophers by Ammonius, followed by philosophy’s
divisions and sub-divisions. The definitions of philosophy included ‘preparation
for death’, a definition based on the idea ascribed to Socrates in Plato’s Phaedo,
that philosophy prepares you to separate the soul from attention to the body,
thus preparing you for its separation and separate continuation after death. But

61
L.G. Westerink, ‘Elias on the Prior Analytics’, Mnemosyne 14, 1961, 126–39; Pseudo-Elias (Pseudo-
David), Lectures on Porphyry’s Isagôgê, Amsterdam 1967.
62
Proclus is credited by Elias, in Cat. 107,24, as pointed out by James Shiel in Chapter 15.
xxxvi Aristotle Transformed

the story of Cleombrotus is told, who misunderstood separation as an invitation

to suicide. As a result there is a discussion of whether suicide is ever legitimate
for a Platonist, and in some later texts for a Stoic. The commentator David
precedes the definitions of philosophy with a question misguidedly raised by
sceptics, whether philosophy even exists, already mentioned by Ammonius.
Westerink points out that the introductions, especially to Aristotle, are more
uniform as between authors than the ensuing commentaries, although they
grow more elaborate as time goes on. The same is true, I think, of Philoponus’
commentary drawn from the seminars of Ammonius on Aristotle’s On the Soul,
that there are fewer signs of additions made by Philoponus to Ammonius in the
introductory Prooemium.
The descriptions of some of the writers Westerink picks out in the first half of
the chapter need a few updates, just as he himself modestly implies in his note 66
that some earlier biographical remarks may be doubted. Westerink offers his
own conjecture about what the deal was that Ammonius struck with the bishop
in Alexandria to keep the school running. But Verrycken raises a caveat in
Chapter 10 about Westerink’s explanation. I have mentioned above my view that
it was to avoid advocating the pagan priestly rituals of theurgy.
As regards Stephanus, it has been noted above that in Aristotle Re-Interpreted,
one chapter will re-date him from being the last in the series in 610 ad to being
earlier in the sixth century than David, or pseudo-Elias, but later than
Olympiodorus and Elias. Here I must update my own editorial footnote 97. At
that time, I took Mossman Roueché to have dated Stephanus earlier than the
other three successors of Olympiodorus but not yet to have put him back to the
sixth century and made him a pupil of Philoponus. Again, as already mentioned,
it will be argued in another chapter of Aristotle Re-Interpreted that the
commentary on Aristotle’s On the Soul Book 3 ascribed to Philoponus is indeed
by Philoponus, not by Stephanus, whether or not Stephanus had any hand in
editing it. If this bit of commentary was by Philoponus, that in turn explains why
it contains so many Christian references, and there will be no need to follow
Westerink in ascribing them to Stephanus.
Westerink’s biographical remarks are especially interesting on Ammonius’
pupil and next successor but one, Olympiodorus, who strongly influenced Elias,
David, and pseudo-Elias, having taught at least one of them, if not more. What is
interesting is the different Christian climate since Ammonius’ experience of riots
between Christian and pagan students and the resulting persecution of pagan
Neoplatonist teachers and the necessity for Ammonius to conclude a peace-
making deal. Now Olympiodorus speaks in his commentaries as if his audience
 Introduction to the Second Edition xxxvii

of students was Christian. He can safely explain to them his own pagan belief,
but he does so in Christian terms. As mentioned above, he acknowledges his
own faith in priestly theurgy. He explains the sense in which he could be seen as
a monotheist, like them. He explains that stones and idols are worshipped only
as a reminder of immaterial powers, a position not so far from Porphyry’s in the
third century. He argues that for Neoplatonists divine punishment after death is
not so severe as Christian eternal punishment. He allows that suicide may
occasionally be acceptable provided it is for a spiritual good. And he explains
Socrates’ guardian spirit as moral conscience.

James Shiel in Chapter 15 gives us a surprise after all the attention to Ammonius
in Chapter 14. Boethius, the younger contemporary of Ammonius, lived under
the Ostrogothic king at Ravenna and produced Latin commentaries on Greek
logic based on Greek ones, with their own introductions. Shiel argues against a
thesis of Courcelle, but on the same side as Minio-Paluello. Courcelle had argued
that Boethius studied with Ammonius and drew his commentaries from him.
Shiel argues that Boethius owed nothing to Ammonius, for a start not to the
anonymously edited commentary of Ammonius on the Categories. Indeed the
main source for Boethius’ commentary on Aristotle’s Categories was Porphyry
from two hundred years earlier. Boethius drew on Porphyry’s smaller
commentary on Aristotle’s Categories, not, in Shiel’s view, his much larger one,
now lost. Boethius does not even know the larger commentary, except for two
quotations, which he seems to have taken from others. Boethius quotes
Porphyry’s junior Iamblichus only twice, and again second hand, once through
Themistius’ criticism. He does not know Iamblichus’ pupil Dexippus. Boethius
does, however, have some affinities with later commentators, for example with
Syrianus, and he observes the six questions introductory to reading a text, which
Shiel suggests Proclus, rather than Ammonius, may have introduced along with
the ten questions introductory to Aristotle, which are the ones ascribed to
Proclus by Elias.
In his smaller and larger commentaries on Aristotle’s On Interpretation,
Boethius’ introductory headings and his division of the text into four are both
different from that of Ammonius’s commentary on On Interpretation, and
this commentary was one of two which he wrote himself and which were not
edited by his students. Boethius explicitly says that he has translated one
elucidation mainly from Porphyry’s (lost) commentary, but partly from the
others (mentioned by Porphyry), and there are four quotations from Syrianus.
The smaller commentary gives one of Proclus’ prescribed questions and the
xxxviii Aristotle Transformed

larger adds another three. It also explicitly mentions when it leaves out a bit of
Porphyry.
Ammonius has only one other commentary on Aristotle which has not been
edited by a student, that on Porphyry’s Introduction (Isagôgê). Once again,
Boethius’ two commentaries on Porphyry’s Isagôgê are quite different from that
of Ammonius, and do not follow Ammonius’ plan or headings. As for Boethius’
treatises on Aristotle’s logic, those on syllogism follow Porphyry and his De
differentiis topicis uses Themistius’ evaluation of Aristotle’s Topics.
In the two commentaries on On Interpretation, Shiel translates Boethius as
saying, ‘For there are scholia of numerous points heaped up all together and so I
have spent almost two years in a constant sweat of writing comments’. Boethius
remarks that his comments were mostly from Porphyry, but he ‘translates all the
other notes’. Elsewhere he says, ‘these too are in the margins’.
Shiel concludes from such comments and from the scarcity of sources used
that Boethius was using marginal comments in Greek in some later commentary
which also cited Porphyry extensively. His rejection of Ammonius as a source
has stood the test of time well. The case will be stronger still if there are no traces
of Ammonius’ extra introductions, and if it can be confirmed that Proclus, before
Ammonius, introduced the six questions introductory to reading a text. But
there has not been so much support for the thesis of Boethius drawing only on
marginalia. John Magee has argued that Boethius was using a full version of
Porphyry’s smaller commentary on the Categories and of his commentary
(surviving now only in 59 pages of Andrew Smith’s fragments) on On
Interpretation, however much he may have used marginalia as well.63

Sten Ebbesen in Chapter 16 also disagrees with Shiel’s view that Boethius read
Porphyry only in excerpts from marginal notes, and he also cites a reference by
Boethius which he takes to be to Porphyry’s larger commentary on the Categories
(Boethius, in Cat. 233B-D), which Shiel thought Boethius had not read. But
neither of these points are Ebbesen’s main interest. He admires Boethius partly
for being better than his predecessors at translating Greek philosophical terms
into Latin, and at keeping the translations consistently word for word. He also
admires Boethius for being conscious, like Porphyry, and like Aristotle as
interpreted by Porphyry, about the need for degrees of paedagogical simplicity,
according to the level of student or reader, even if there are more complex things

63
John Magee, ‘On the composition and sources of Boethius’ second Peri Hermeneias commentary’,
Vivarium 48, 2010, 7–54.
 Introduction to the Second Edition xxxix

that could be said. Having given Porphyry’s account of what Aristotle’s Categories
is about, Boethius says (in Cat. 160A-B) that he hopes to give an account different
from Porphyry’s using the Pythagoreans. This sounds like a plan, never carried
out, to offer an account based directly or indirectly on Iamblichus’ theory that
Aristotle’s Categories can also refer to Pythagorean levels of reality as well as to
the perceptible world.
Ebbesen thinks that Boethius also did well to move to the beginning of his
commentary on the Categories Porphyry’s theory of the origins of language
according to which there were two impositions of names. First there were names
of things and then, as it were, names of names. Boethius (in Cat. 176D–177A)
followed Dexippus (in Cat. 26,23–27,2), a commentator influenced by Porphyry
(via Iamblichus), to solve a particular problem which Porphyry had raised
without, however, using the solution available to him (in Cat. 81,18–20). The
problem, which Ebbesen discusses also in Chapter 19, is formulated in the
commentaries of all three philosophers on Aristotle’s Categories, and it runs as
follows. It would be a fallacy to argue: ‘Socrates is a man. Man is a species.
Therefore Socrates is a species’. Dexippus and Porphyry solve the problem by
invoking Porphyry’s second imposition of names. ‘Man’ is the name of a thing,
but ‘species’ is the name of a name, and that is why the argument is invalid.
Boethius could profitably have made use of this type of solution elsewhere,
Ebbesen claims, but he did so only when reproducing Dexippus who was
influenced by Porphyry. The idea of names of names does not appear in Boethius’
work on the Topics of Cicero and Aristotle, for the latter of whom his guide is
Themistius.

In Chapter 17, Robert Browning turns from the sixth century ad to the
renaissance of commentary on Aristotle, most of which he re-dates as starting in
the twelfth century. He made use of a funeral oration for Anna Comnena,
daughter of the former Byzantine emperor, Alexius I. She died some years after
1147, and before 1156, perhaps between 1153 and 1155. The oration was delivered
by the little known George Tornikes, Metropolitan Bishop of Ephesus. The end
of the Oration comments: ‘I have myself heard the wise man from Ephesus
attribute the cause of his blindness to her, because he spent sleepless nights over
commentaries on Aristotle at her command, whence came the damage done to
his eyes by candles through dessication’. By identifying the wise man as the
philosopher Michael of Ephesus, Browning brought forward the dates of
Michael’s commentaries by a century from the preceding century, where he had
been placed by Karl Praechter, to the period 1118–38. And not only did he bring
xl Aristotle Transformed

Michael forward to the twelfth century, but also another probable member of
Anna’s project, the philosopher Eustratius, about whom H.P.F Mercken has more
to say in Chapter 18. There had already been in the eleventh century a resumption,
after a five hundred year gap, of paraphrase of, and commentary on, Aristotle’s
logical works with Michael Psellos, but what was special about Anna Comnena’s
twelfth century project was her attempt to recover bits of missing commentary
from the first six centuries ad, and to fill in gaps in that corpus. After her father’s
death in 1118, she was sidelined for the next 35 years and retired to monastery
premises in Constantinople overlooking the inlet of the Golden Horn, but, the
Oration tells us, she assembled learned men to study Aristotle, Plato, Euclid, and
Ptolemy. Browning argues that she wanted the work not to conflict with Christian
belief in a beginning of the universe and just reward or punishment after death.
Browning points out that James of Venice was in Constantinople in the 1130s
and subsequently translated into Latin Aristotle’s De Sophisticis Elenchis on
fallacies, his Physics, and others of his works. This was at the very beginning of
the recovery by the Latin-speaking world of Aristotle and his commentaries in
the twelfth century beyond the point at which they had been left by Boethius
in the sixth. Michael himself commented on Aristotle’s De Sophisticis Elenchis
and it is possible that James became acquainted with these works through Anna’s
circle.
Eustratius, pupil of Psellos’ pupil John Italos, had, like his teacher, as Mercken
describes in Chapter 18, been condemned merely for applying Aristotelian logic
to Christian theology. Eustratius is praised by Anna in her own History of her
father’s reign. Eustratius had already written commentaries on a work of Aristotle
close to his own interest in demonstrative proof, the Posterior Analytics. But in
his commentaries on Nicomachean Ethics Books 1 and 6, he describes himself as
old and he addresses his unnamed patron as basilis, which can mean princess.
This is all evidence for Browning’s conjecture that he too came to be part of
Anna’s circle.
Browning’s important re-dating and contextualising of this renewed
commentary movement to the twelfth century has been widely accepted, and
only a few updates are called for. First, I believe that his description of the
Alexandrian commentators David and Elias as pupils, rather than predecessors,
of Stephanus represents an acceptance of Mossman Roueché’s initial revision,
but at a stage where Roueché had not yet brought Stephanus back into the sixth
century, nor made him later than Elias, as he has done and will do in Aristotle
Re-Interpreted. Secondly, Philoponus’ arguments for a beginning of the universe
represent not the Alexandrian school as a whole, but the idiosyncratic dissent of
 Introduction to the Second Edition xli

Philoponus, the Christian pupil and editor of Ammonius, expressed, I shall

argue in Aristotle Re-Interpreted, surprisingly early in his association with
Ammonius.

In Chapter 18, H.D.P. Mercken gives the most comprehensive treatment I know
of commentaries on Aristotle’s Nicomachean Ethics. These had not been included
in the Neoplatonist curriculum from the later third century to the sixth, so are
either of earlier date or among the gap-fillers started by Anna in the twelfth
century. A composite commentary by a number of hands was assembled in
twelfth century Constantinople, the commentary on Books 5, 9, and 10 being by
Michael, on 1 and 6 by Eustratius, on 8 by Aspasius, the Aristotelian of the first
half of the second century, while anonymous commentaries from the last quarter
of the second century are included on Books, 2 to 5, that on 5 duplicating the one
by Michael. Robert Grosseteste (c. 1168–1283), the first Chancellor of the
University of Oxford, translated the composite commentary into Latin, so, as
also pointed out by Browning in Chapter 17, twelfth century Constantinople
helped to make possible the Latin-speaking world’s recovery of the bulk of
Aristotle – only helped because another conduit was translations made in Spain
from Arabic translations and commentaries.
Mercken gives a fuller account of Eustratius. His commentaries on the
Posterior Analytics Books 1 and 2, Aristotle’s account of science as subject to
demonstrative proof, was an early work. His commentaries on Nicomachean
Ethics Books 1 and 6 often convert Aristotle’s arguments into syllogistic proofs.
This work is highly Christian in a way that Browning suggests Anna would like.
The work also supports against Aristotle Plato’s theory of Ideas, but Eustratius
includes a caution that he may be wrong, perhaps remembering that Psellus had
been accused of backing Plato against Christ. But Eustratius was to be involved
in other heretical charges. In 1082, he joined in anathematising his teacher John
Italos’s doctrines for discussing Christ dialectically. But after a period of success
as a court theologian to Alexius I Comnenus, he was himself excluded from
divine office for life on the charge of explaining Christ’s incarnation dialectically
and saying that Christ reasoned in syllogisms. It was during this final period of
disgrace that Anna got him to write the commentaries on Aristotle’s Ethics.
Michael started his philosophy at the Academy of an earlier emperor,
Constantine IX Monomachus. He produced commentaries on a huge range of
Aristotle’s texts. At least one of the two on the Nicomachean Ethics, we have seen,
belongs to his old age. That on the Sophistici Elenchi is the only one that survives
of his commentaries on Aristotle’s logic. Other surviving commentaries written
xlii Aristotle Transformed

before that include one on Aristotle’s psycho-physical works, the Parva Naturalia,
and apparently before that again on four of Aristotle’s zoological works. There
are only a few pages surviving from a commentary on Aristotle’s Politics. He had
a role in compiling and in part writing a commentary on Aristotle’s Metaphysics.
The commentary on the first five Books is by Alexander of Aphrodisias from the
early third century and the rest used to be called the work of pseudo-Alexander.
But a contribution to Aristotle Re-Interpreted will propose new identifications of
all the authors of this commentary from Metaphysics Book 6 onwards, and will
slightly shrink Michael’s role. According to these new proposals, Michael wrote
the commentary on books 7 to 10, drawing on the sixth century commentary of
Asclepius, but was supplemented after his time on Book 9 by George Pachymeres.
Lost are the commentaries he refers to on Aristotle’s Analytics, Topics, Rhetoric,
and Physics.
The anonymous commentaries, or rather skholia (notes), on Books 2–5 of the
Nicomachean Ethics seem quite possibly to have been compiled and written up
by one person and can be dated to the late second century because of a reference
to the earlier Platonist Atticus and because of material drawn from the earlier
commentator Aspasius. The compiler is thought to have drawn also on Adrastus,
who earlier in the century wrote On Historical and Literary Questions. Paul
Moraux collected fragments in the Ethics commentary whose subject matter
suggests that they were drawn from that book.
From Aspasius there survives a separate commentary on Books 1–4 of the
Nicomachean Ethics, part of 7, and, incorporated in the composite commentary,
most of 8. This is the subject of the 1999 study cited above, edited by Alberti and
Sharples.

In Chapter 19 Sten Ebbesen relates the Greek commentators to philosophy in

medieval Latin. He looks for the sources in the Greek commentators on Aristotle
of the shifting preferences among later medieval Latin philosophers for treating
universals as words, concepts, or things of a more robust kind. In Aristotle Re-
Interpreted, I shall try to give an account of where the Greek commentators stood
on this question. But the impression I derive from Ebbesen’s account is that
chance may have dictated what ancient texts caught the imagination of medievals
and encouraged their preference for one way of thinking rather than another.
Ebbesen’s point is interesting that Philoponus’ commentary on the Posterior
Analytics encouraged an interest in universals as things of a more robust kind,
because this is not what one would have particularly expected from his explicit
discussion of the status of universals in his commentary on Aristotle’s Categories.
 Introduction to the Second Edition xliii

Porphyry is one of many commentators in a six hundred year tradition who

took Aristotle’s Categories to be concerned with three matters: (1) spoken sounds
insofar as they (2) signify (3) things.64 Moreover, the infinity of beings and of
signifying expressions are included under ten supreme genera, the categories.
Not until Porphyry’s critic Iamblichus, who drew closely on Porphyry, was the
second term, ‘signify’ spelled out as implying concepts (noêmata), as reported by
Philoponus, Elias, and Olympiodorus, and this decision is approved by Simplicius.
This opened the way to a certain conceptualism about Aristotle’s ten categories
or supreme genera. I have noticed only one Greek commentator, Olympiodorus,
who shortened Porphyry’s formula so as to make him regard the Categories as
being about spoken sounds, as if he were a nominalist, but he is contradicted by
others. It was the critics, the Stoics Athenodorus and Cornutus and the Platonists
Lucius and Nicostratus, none of them commentators, who strangely thought
that the Categories was about parts of speech and criticised Aristotle for leaving
some out.
In the commentary tradition, Porphyry discusses the status of genera and
species more generally and it has recently been emphasised that it is in the
different context of commenting on Ptolemy’s Harmonics that he connects
universals with concepts abstracted from sensibles. His more general view is that
Aristotle’s Categories is not concerned with the intelligible level of Plato’s Ideas,
but only with the sensible world and that the sensible world includes three kinds
of genus and species at the level of spoken sounds, concepts, and things. I think
it was Ammonius, two hundred years later, who presented Porphyry as
connecting genera in his works on Aristotle’s logic with concepts, but his evidence
was not good. Ammonius’ pupil Philoponus on the whole agrees with Ammonius’
treatment of universals as it appears in the commentary on the Categories
anonymously edited from Ammonius’ seminars. But on the subject of predication,
discussed by Ebbesen, Philoponus disagrees with his teacher and says that what
is predicated of individuals is not the common items existing in the many
particulars but concepts and the conceptual which are after the particulars.
This enduring distinction of three levels of universal, Plato’s intelligibles
before the many, universals in the many, and universals after the many, or, as
Eustratius would say in the twelfth century, from the many, had been introduced
by Ammonius’ teacher, Proclus. Only universals at the third and lowest level
were concepts for Ammonius. As for the highest level of Plato’s intelligible Ideas,
Proclus had denied that Ideas shared the same characteristics as the particulars

64
Porphyry, in Cat. 58,4–7.
xliv Aristotle Transformed

that fell under them. Rather, the particulars were associated with each other by
their diverse relations to the Idea, in a relationship which Aristotle called pros
hen and aph’ henos – being directed to and dependent on one and the same thing.
Simplicius later got the same sort of conclusion by a different route. He said that
the Platonic Idea is common (i.e. universal) as a cause, but not as a common
nature (phusis).
The idea of common natures, which Simplicius here denies to Plato’s Ideas,
Ebbesen points out, was taken up in the Middle Ages. Much earlier the
Aristotelian commentator Alexander, who did not try to accommodate Plato’s
Ideas since he did not believe in them, had said that an enmattered nature
(phusis) is not in reality or in its own nature universal (katholou) or (equivalently)
common, but can become common by being thought or constructed by a
conceptual separation.65 The separation seems to be from the material
circumstances of the enmattered nature. Examples of enmattered natures really
existing in individual humans, or in individual animals, are rational mortal
animal, and animate being with sensation. Alexander says that when the nature is
thought of separately from the things with which it exists, it is thought of not as
it exists.66 This last caveat is expanded, it has been pointed out, by Boethius in his
second commentary on Porphyry’s Isagôgê.
Boethius also reproduced the distinction made by Porphyry in his
commentary on Aristotle’s Categories of two stages in the human development
of language. First there were names of sensible things and then names of names
(58,1–3). Boethius, as Ebbesen points out, makes this more prominent by
advancing it to the very beginning of his own commentary on Aristotle’s
Categories. But Porphyry, so far from implying there that the genera and species
of the Categories are only names, immediately goes on to give his threefold
formula. Aristotle’s Categories is about (1) simple vocal sounds that (2) signify,
insofar as they signify (3) things. This is stated in support (gar) of, not in contrast
with, the point that the project of Aristotle’s book is the first imposition of (1)
expressions (lexeis), an imposition which (2) manifests (prostatikê) (3) things
(58,3–6). Boethius is completely faithful to Porphyry’s threefold distinction. He
concludes his opening statement by saying: ‘Accordingly, the direction of this
work is to carry on a thorough discussion from (1) utterances (2) signifying (3)
things (res) insofar as they are signifying.’

65
See Alexander, DA 90,2–11 (cf. Quaestio 1.3, 8,3–4); Quaestio 1.3, 8,17–22; Quaestio 2.28, 79,16–18;
Quaestio 2.28, 78,18–20.
66
Alexander, Quaestio 1.3, 8,19–20.
 Introduction to the Second Edition xlv

In his Introduction (Isagôgê) to Aristotle’s logic, Porphyry declined to discuss

certain questions about universals as unsuitable for beginners. The first question
was whether universals have real subsistence or whether they reside only in bare
concepts (monai psilai epinoiai). Boethius translates this in Latin as: in solis nudis
purisque intellectibus. The third question is whether they are separated or exist in
sensibles. I suspect that it was also for the sake of beginners that Porphyry
described individuals not as Aristotle’s compounds of matter and form, notions
which are not yet introduced for beginners in Aristotle’s Categories, but, as
Ebbesen mentions, as bundles of properties. This idea is already in Plato’s
Theaetetus, from where, I suspect, the Stoics developed it, and Porphyry, seeking
in an introduction to avoid Aristotle’s more complex notions, could have drawn
it from Plato’s Theaetetus, even if he had not known about the Stoics.
If the Latin-speaking medievals had had access to all these discussions in the
Greek commentators on Aristotle, they could have had precedents for all these
views about universals. Ebbesen tells us of a rather haphazard selection of
commentators’ texts that they drew on. It is not for me to discuss whether in the
meantime more uses of the commentators by medievals have come to light, but
it would be very good to have Ebbesen’s own update.

In Chapter 20, Ian Mueller discusses the commentators’ views on the objects of
mathematical thought. He is not discussing how the mind forms non-
mathematical universals, even though some (not all) commentators might say
that abstraction is involved in both cases. Those who appeal to abstraction for
the mind’s formation of mathematical objects include Aristotle’s defender
Alexander, Porphyry (in his commentary on Ptolemy’s Harmonics), Ammonius,
and Philoponus. But Iamblichus introduced a rival Pythagorean theory of mental
projection, somewhat analogous to the projection of a movie onto a cinema
screen, and with this the Athenians Syrianus and Proclus agreed. Abstraction is
a thought process of selectively attending to some features and not others and
can be called epinoia (thought), no matter whether what is attended to is a
physical object or a mental image. What is ignored in mathematical abstraction
is the matter and motion of physical objects, and what is left is their mathematical
properties. Projection, however, centrally involves the screen of the imagination
(phantasia). But for geometry projection also involves humans having within
them rational principles (logoi) of perfect geometrical shapes, and as Proclus
explains, in geometers perfect images of these logoi are projected into the
imagination. He seems to be speaking of mathematicians generally, not of some
special Pythagorean ones. Syrianus uses the powerful argument cited above in
xlvi Aristotle Transformed

favour of this theory of mathematics, that one could never get to think about
perfect figures, if one had to rely, as abstractionism does, on sense perception,
which never shows us perfect figures. How could abstraction show one how to
correct the imperfect figures that perceptible objects supply, unless one had
within oneself the logoi of perfect geometrical figures, which the projectionist
theory postulates?
Ammonius and Philoponus are on the other side. The geometer, on seeing a
copper, wooden, or stone circle, is able to abstract and ignore the matter, thus
gaining the idea of a mathematical circle. Simplicius, however, though recognising
that Aristotle believes in abstraction, himself favours projection. But, Mueller
points out, he nonetheless draws a further distinction, with which Syrianus may
have agreed. For he allows that the method of projection may be used by
Pythagorean mathematics and ordinary mathematics in different ways. He says
that what Pythagorean mathematicians investigate is not the imprints (tupoi) of
logoi projected into the imagination, but the logoi themselves which do the
projecting. They take the objects of mathematics to be substantial causes and
substances, and this activity is philosophy (in DA 277,1–6). I do not know whether
Simplicius has in mind such interests of Syrianus as the investigation of the
principles of number, the One, and the Dyad, which in turn may be principles of
geometrical shapes and hence eventually of the triangles which in Plato’s Timaeus
make up physical bodies. For Aristotle, by contrast, according to Simplicius, the
objects of mathematics, which Aristotle erroneously thinks are reached by
abstraction, are in the imagined imprints. And it is also true, according to
Simplicius, of the projection used by ordinary mathematics that its objects are
the imprints.
One update has come to my attention. Marwan Rashed has considered
Mueller’s account of mathematical abstraction in Aristotle and Alexander too
mentalistic, not granting enough reality to the objects of mathematics.67 He
allows that this could be cured by a suggestion I made, that abstraction is meant
to be carried out by the mind (the word epinoia is in the dative), not in the
mind.68 In my example, we can by the mind think of the equator as a line round
the middle of the earth. But we do not think of it as in the mind; we think of it as
on the earth. In the same way, a geometer could by selective mental abstraction
think of a copper circle without its copper, and then by using the mind would be

67
Marwan Rashed, Alexandre d’Aphrodise, Commentaire perdu à la Physique d’Aristote (Livres IV–
VIII), Berlin 2001, 58–65.
68
For Aristotle, Matter, Space and Motion, London 1990, 16–17; for Alexander, Philosophy of the
Commentators, 200–600 AD, A Sourcebook, vol. 3, London 2004, ch. 12a, p. 293.
 Introduction to the Second Edition xlvii

thinking of the geometrical circle as a distinct entity in the copper. Rashed agrees
that that would save the geometrical circle from being a mere figment of the
mind, and even agrees that it is compatible with Alexander’s text. But he does
not think it is actually suggested by Alexander’ text at in Metaph. 52,13–21,
whatever might be inferred from the Aristotle passage I cited, Metaphysics 9.9,
1051a21–33.

My conclusion is that all the chapters of Aristotle Transformed are of continuing

importance, including those that have aroused controversy, or against which
alternative interpretations will need to be measured. Some chapters seem to
me to require no addition; others require a few brief updates, some of which I
have tried to supply. Frequently, those updates will be spelled out in fuller detail
in Aristotle Re-Interpreted, and repeatedly the two volumes will, I believe, be
complementary to each other.
xlviii
 1

The ancient commentators on Aristotle

Richard Sorabji

Two impulses: Andronicus and Porphyry

Cicero reports in the Topics (ch. 3), written in 44 bc , that Aristotle is ignored by
all but a few philosophers, and indeed his own knowledge of Aristotle does not
extend far beyond the early dialogue works. Yet even before the time he was
writing, as Gottschalk argues in Chapter 3 below, a new explosion of interest in
Aristotle was under way, which was to occupy the rest of the century. Whether in
Rome or more probably in Athens, Andronicus of Rhodes had begun his work
on the great edition of Aristotle which forms the basis of today’s editions, and he
had accompanied some of the treatises with commentaries. Altogether five
different commentaries were produced on Aristotle’s Categories by the end of the
century, along with a Doric version of the Categories purporting to be the work
of the old Pythagorean Archytas, and two compendia of the Philosophy of
Aristotle. The commentaries are lost, except for fragments, notably those
preserved in Simplicius’ commentary on the Categories, and the next comparable
boost to Aristotelian studies would not come until Porphyry in the third century
ad. But the tradition of commentary on Aristotle had begun.
The earliest surviving commentaries come from Aristotelians of the second
century ad, culminating at the end of that century and the beginning of the next
in the commentaries of the greatest expositor and elaborator of Aristotle’s
thought within the Aristotelian tradition, Alexander of Aphrodisias (appointed
to his Aristotelian Chair between 198 and 209).
Outside the Aristotelian schools, the chief interest in the first two centuries
ad still focused on Aristotle’s Categories. The work seems to have acted as a
catalyst, attracting commentaries from three schools, the Stoic, Platonist and
Aristotelian. Gottschalk suggests below that it was the Aristotelians who first
described the Stoic scheme as one of ‘categories’, but that the comparison forced

1
2 Aristotle Transformed

them in return to establish a correct order for their own scheme of categories,
since order was important to the Stoic scheme in a way that Aristotle never took
it to be important.
Arguments raged for and against the viability of Aristotle’s categorial scheme,
and the arguments against might well have won. For in Rome Plotinus (c. 205–
260), whom we tend to regard, following a modern classification, as the founder
of Neoplatonism, sided with the arguments against in Enneads 6.1–3. But on this
issue, Porphyry (232–309), his disciple, biographer and editor, resisted him (P.
Hadot, Chapter 6 below).1 Platonists should not complain with Plotinus that
Aristotle’s categories fail to take account of the Platonic Ideas, for the Categories
is not about things, but only about words insofar as they signify things,2 and
words get applied primarily to things in the sensible world, not to the Ideas in the
intelligible realm.3 Porphyry’s pupil Iamblichus (c. 240–c. 325) went further with
his ‘intellective theory’ (noera theôria). He tried to show how the categories,
properly understood, apply first and foremost to the intelligible realm. The
category of place, for example, understood as a boundary which embraces and
holds something together, applies most fully there.4 Iamblichus’ pupil Dexippus
(c. 330), perhaps drawing on Porphyry, adds that the reason why the Categories
is not directly about things is that it is designed for beginners.5 And he shows
how Aristotle’s account of substances may still apply in an analogous way to
intelligible as well as to sensible substances.6
Porphyry’s intervention was decisive. He is credited with a work in seven
books entitled On the School of Plato and Aristotle Being One and (if it is distinct)
one On the Difference between Plato and Aristotle. Some of the contents are
preserved in an Arabic treatise by al-’Almirî.7 From there on, the study of Aristotle
was assured. For the unity of Platonism and Aristotelianism meant that Aristotle’s
logic and a wide selection of his other texts became a standard prerequisite for
Platonic studies in the Neoplatonist schools, and Neoplatonism was now the
dominant philosophy. Porphyry himself wrote an introduction to the Categories

1
See also A.C. Lloyd, ‘Neoplatonic logic and Aristotelian logic’, Phronesis 1, 1955–6, 58–79 and 146–
60.
2
Porphyry in Cat. 57,7–8; 58,5–7; 91,19; ap. Simplicium in Cat. 10,22–3.
3
Porphyry in Cat. 91,19–27.
4
Iamblichus ap. Simplicium in Cat. 363,29–364,6; general programme at 2,13.
5
Dexippus in Cat. 40,19–25.
6
Dexippus in Cat. 41,18–42,3, discussed by P. Hadot in Chapter 6. However, if Dexippus (in Cat. 44)
rejects the view of Iamblichus (ap. Simplicium in Cat. 116,25) that intelligible substances also satisfy
the criterion of admitting opposites, this may lend support to Sten Ebbesen’s suggestion in ch. 20
that we can see (at 17,1) some impatience directed against Iamblichus by Dexippus.
7
As-Saâdah Wa’l-Isâd (On Seeking and Causing Happiness), M. Mînuvî (ed.), Wiesbaden 1957.
 The ancient commentators on Aristotle 3

called the Isagoge (Introduction) or Quinque Voces (Five Terms), two Categories
commentaries, of which one is extant, one fragmentary, and half a dozen or more
other commentaries on Aristotle, for some of which fragments survive.8

The harmony of Plato, Aristotle and other

Greek philosophers

The harmony of Plato and Aristotle was accepted to a larger or smaller extent by
all commentators in the Neoplatonist tradition, and the great bulk of the ancient
commentators, Christians included, are in that tradition. Among the major
commentators after Porphyry, there are only two exceptions. Themistius, so
Blumenthal argues in Chapter 5, remained more Aristotelian than Platonist.9 And
in the twelfth-century revival of commentary writing in the circle of Anna
Comnena, while Eustratius observes the traditional ‘harmony’,10 Michael of Ephesus
reverts to the Aristotelian tradition represented by Alexander of Aphrodisias.
On the harmony of Plato and Aristotle, Iamblichus of the Syrian school
(c. 240–c. 325) was accused of going too far and denying that Aristotle
contradicted Plato’s theory of Ideas.11 In fifth-century Alexandria, Hierocles also
espoused a thorough-going version of harmony and made Plato and Aristotle
agree on the subject of Creation.12 He credited the general thesis of harmony to
his teacher Plutarch of Athens (died 432) and to his Alexandrian forebear
Ammonius Saccas, who taught Plotinus in the third century. A more nuanced
view was taken in Athens by Plutarch’s other pupil, Syrianus (died c. 437), and
their common pupil Proclus (c. 411–485). They accepted harmony in many
areas,13 but could see that there was disagreement on the theory of Ideas (H.D.

8
See Francesco Romano, Porfirio e la fisica aristotelica, Catania 1985, for a collection of the Physics
fragments in Italian translation. The named fragments of the lost Categories commentary are being
assembled and edited by Andrew Smith. The tentative list of ten commentaries and related works
in Romano, p. 33, is augmented by Ebbesen’s suggestion of an in SE, in Chapter 7 below.
9
The chapter incorporates a response to the opposing view of E.P. Mahoney, ‘Themistius and the
agent intellect in James of Viterbo and other thirteenth-century philosophers (Saint Thomas
Aquinas, Siger of Brabant and Henry Bate)’, Augustiniana 23, 1973, 422–67, at 428–31; id.,
‘Neoplatonism, the Greek commentators and Renaissance Aristotelianism’, in D.J. O’Meara (ed.),
Neoplatonism and Christian Thought, Albany 1982, 169–77 and 264–82, esp. n. 1, 264–6. Some
Platonist influence is cited too in Robert Todd, introduction to translation of Themistius in DA 3,
4–8, in Frederick Schroeder and Robert Todd, Two Greek Aristotelian Commentators on the Active
Intellect, Toronto, forthcoming.
10
A.C. Lloyd, ‘The Aristotelianism of Eustratios of Nicaea’, in Jürgen Wiesner (ed.), Aristoteles Werk
und Wirkung, Berlin 1987, 341–51.
11
Elias in Cat. 123, 1–3.
12
Photius Bibliotheca (Bekker) 171b33ff.
13
Syrianus in Metaph. 80,4–7; Proclus in Tim. 1. 6,21–7,16.
4 Aristotle Transformed

Saffrey, Chapter 8) and also on the issue of whether God was causally responsible
for the existence of the physical world, which Aristotle denied. But back in
Alexandria, Proclus’ pupil Ammonius (434/45–517/26) was to claim harmony
even on these issues, and, though the debate was not clear-cut,14 his claim was
to prevail. Aristotle, he maintained, accepted Plato’s Ideas,15 at least in the form
of principles (logoi) in the divine Intellect, and these principles were in turn
causally responsible for the beginningless existence of the physical world.
Ammonius wrote a whole book to show that Aristotle’s God was thus an efficient
cause of the world’s existence. The book is lost, but some of its principal
arguments are preserved by Simplicius.16 Ammonius’ interpretation of Aristotle’s
God as Sustainer of the world was passed on through Farabi’s Harmony of
Plato and Aristotle, and eventually influenced the Christian writers of the
thirteenth century, including Thomas Aquinas. It was the Aristotle of the
Alexandrian commentators (and of certain spurious works) that they inherited.
Hence, by an irony, what started in Ammonius as a desire to harmonise Aristotle
with Plato finished in the thirteenth century by helping to make Aristotle safe
for Christianity (Richard Sorabji, Chapter 9). In Simplicius (writing after 532),
who goes furthest of all, it is a formally stated duty of the commentator to display
the harmony of Plato and Aristotle ‘in most things’.17 Philoponus (c. 490–570)
who with his independent mind had thought better of his earlier belief in
harmony, is castigated by Simplicius for neglecting this duty (Koenraad
Verrycken, Chapter 11).18
The idea of harmony was also extended beyond Plato and Aristotle to embrace
the Presocratics. Plato’s pupils Speusippus and Xenocrates saw Plato as being
in the Pythagorean tradition.19 And this view gained support from pseudo-
Pythagorean forgeries of the third to first centuries bc , which drew on the
ideas of Plato and Aristotle, but embedded them in works purporting to come

14
Asclepius sometimes accepts Syrianus’ interpretation (in Metaph. 433,9–436,6), which is, however,
qualified, since Syrianus thinks Aristotle is really committed willy-nilly to much of Plato’s view (in
Metaph. 117,25–118,11; ap. Asclepium in Metaph. 433,16; 450,22). Philoponus repents of his early
claim that Plato is not the target of Aristotle’s attack, and accepts that Plato is rightly attacked for
treating Ideas as independent entities outside the divine Intellect (in DA 37,18–31; in Phys. 225,4–
226,11; contra Procl. 26,24–32,13; in An. Post. 242,14–243,25). Elias also thinks (in Cat. 123,1–3)
there is some disharmony on the theory of ideas.
15
Asclepius in Metaph. from the voice of (i.e., from the lectures of) Ammonius 69,17–21; 71,28; cf.
Zacharias Ammonius PG vol. 85, col.952 (Colonna).
16
Simplicius in Phys. 1361,11–1363,12, translated in Richard Sorabji, Matter, Space and Motion,
London and Ithaca NY 1988, ch. 15.
17
Simplicius in Cat. 7,23–32.
18
Simplicius in Cael. 84,11–14; 159,2–9.
19
See e.g. Walter Burkert, Weisheit und Wissenschaft, Nürnberg 1962, translated as Lore and Science in
Ancient Pythagoreanism, Cambridge Mass. 1972, 83–96.
 The ancient commentators on Aristotle 5

from an earlier period and to be written by Pythagoras and his pupils.20 These
works, though not deceiving everyone,21 were taken by the Neoplatonists as
genuine. Plotinus saw the Presocratics as precursors of his own views,22 but
Iamblichus went far beyond him by writing ten volumes on Pythagorean
philosophy and making maximum use of the pseudo-Pythagorean writings.23
Thereafter Proclus sought to unify the whole of Greek philosophy by presenting
it as a continuous clarification of divine revelation,24 and Simplicius argued for
the same general unity in order to rebut Christian charges of contradictions in
pagan philosophy,25 charges which had been provoked inter alia by Porphyry’s
earlier complaints of conflicts in Christianity.26 Not for the only time in the
history of philosophy (the condemnation of 219 propositions in 1277 provides
another example), a perfectly crazy position (harmony) proved philosophically
fruitful. To establish the harmony of Plato and Aristotle, philosophers had to
think up new ideas, and the result was an amalgam different from either of the
two original philosophies.

Neoplatonist school practices

The Neoplatonist commentaries increasingly come to reflect a teaching

curriculum. Under Iamblichus, twelve dialogues of Plato were selected for study
in a set order, culminating in the two ‘theological’ dialogues, the Timaeus and the
Parmenides.27 Syrianus had Proclus study Aristotle as the ‘Lesser Mysteries’
serving to introduce the ‘Greater Mysteries’ of Plato. As a very quick student, he
took two years over Aristotle.28 Proclus’ pupil Ammonius and subsequent
commentators make it explicit in their introductions to the Categories
commentaries that the eventual purpose of embarking on a study of Aristotle is

20
See Holger Thesleff, An Introduction to the Pythagorean Writings of the Hellenistic Period, Åbo 1961;
Thomas Alexander Szlezàk, Pseudo-Archytas über die Kategorien, Peripatoi vol. 4, Berlin and New
York 1972.
21
Themistius recognised the spuriousness of pseudo-Archytas’ version of the Categories (Boethius in
Cat., PL 64, 162A).
22
Plotinus e.g., 4.8.1; 5.1.8 (10–27); 5.1.9.
23
See Dominic O’Meara, Pythagoras Revived: mathematics and philosophy in late antiquity,
forthcoming.
24
See Christian Guérard, ‘Parménide d’Elée selon les Néoplatoniciens’, forthcoming.
25
Simplicius in Phys. 28,32–29,5; 640,12–18. Such thinkers as Epicurus and the Sceptics, however,
were not subject to harmonisation.
26
Probably in his lost treatise Against the Christians.
27
Anonymous Prolegomena to Platonic Philosophy, L.G. Westerink (ed. & tr.), Amsterdam 1962, ch. 26
(new, revised edition, translated into French, Collection Budé, forthcoming).
28
Marinus, Life of Proclus, ch. 13.
6 Aristotle Transformed

to be carried up to the supreme Neoplatonist God, the One.29 This ascent to God
would presumably be achieved only when the Greater Mysteries of Plato’s
Timaeus and Parmenides were revealed.
Much of the relevant information is found in the ten-point introductions to
Aristotle’s philosophy which are prefixed to Categories commentaries from
Ammonius onwards, and which are described by Praechter in Chapter 2 and
Westerink in Chapter 14 below. The ten points were already laid down, we are
told, by Proclus in his Sunanagnôsis (reading of a text with a master).30 One of
the standard ten points is the aim of studying Aristotle, ascent to God, an idea
still found in Eustratius’ twelfth-century commentary on the Nicomachean Ethics
(Mercken, Chapter 18), and two others are the means to that end, which
determines the canon to be studied, and what part of that canon is to be taken
first. The student is to start with the logical works of Aristotle, and then move on
through ethics, physics, mathematics and theology. Logic is important as the
starting point, but no longer has the monopoly it enjoyed at the time of the very
earliest commentaries.
One of the ten points is devoted to the qualities demanded of a commentator.
It is here that a complete knowledge of Aristotle is demanded, and that Simplicius
requires the commentator to track down the harmony of Plato and Aristotle in
most things, while Elias complains that Iamblichus took the harmonisation too
far. On the other hand, when Alexander denies that for Aristotle man’s rational
soul is immortal, Elias does not view that as a legitimate assertion of disharmony
between Plato and Aristotle, but as opposition to his own school.31
Another introductory heading concerns the reason for Aristotle’s obscurity of
style, which Ammonius explains as designed so that ‘good people may for that
reason stretch their minds even more, whereas empty minds that are lost through
carelessness will be chased away by the obscurity when they encounter sentences
like these’.
The tenth introductory heading breaks up into six or more points about the
particular treatise to follow, the Categories. Ammonius says that each and every
treatise of Aristotle should be introduced by such a set of six points,32 but

29
Ammonius in Cat. 6,9–16.
30
Elias in Cat. 107,24–6.
31
Elias in Cat. 123,1–7. Alexander is also accused of introducing his own doctrines into his
commentaries on Aristotle by Plutarch of Athens, ap. Philoponum in DA 21,20–3. Another aspect
of the discussion of the qualities of commentator and audience in Elias and Olympiodorus concerns
the adage that Socrates or Plato are dear, but the truth is dearer. For this see L. Tarán, ‘Amicus Plato,
sed magis amica veritas’, Antike und Abendland 30, 1984, 93–124.
32
Ammonius in Cat. 7,7–14.
 The ancient commentators on Aristotle 7

Philoponus holds that the points should be made only where called for.33 One of
the six points concerns the authenticity of the ensuing treatise, a question which
Andronicus had had to address, when he constructed the first edition of
Aristotle’s works.34
Even before the ten points on Aristotle’s philosophy, Ammonius and his
successors provide a commentary on Porphyry’s Isagoge (Introduction to the
Categories), and preface it in turn with an account of philosophy in general. The
definition of philosophy as contemplation of death provides the occasion for a
disquisition on suicide. Peter Brown has observed in conversation that the
Greeks would have been very shocked by the Latin-speaking circle of Ambrose
and Augustine, which seized on Greek philosophical texts as they became
available in Latin, and read them in any order. This is the antithesis of Proclus’
‘reading of a text with a master’. However, the formal scheme of the six-point
introduction (along with some material from the ten points and a definition of
philosophy) was eventually passed on to the Latin world in the sixth century by
Boethius in his commentaries on Aristotle’s logical works (Shiel, Chapter 15).
And the influence of Boethius’ introduction on twelfth-century Latin texts has
been detected in fields as far apart as Rhetoric, Grammar, Literature, Medicine,
Law and Theology.35 It is an irony that the Neoplatonists lavished so much
attention on the written text, when Plato preferred the spoken word on the
mistaken ground that you cannot ask a text questions.
I. Hadot has argued, contrary to the suspicions voiced by Praechter, that a
formalised set of six points can be traced back as far as the Christian theologian
Origen in the third century ad in his Commentary on the Song of Songs.36 The
points correspond to various items in the ten-point and six-point scheme of
Alexandrian Neoplatonism. In fact, a number of the points are discussed in the
earliest Categories commentaries, as also are the question of the order of the
categories, and whether they apply to the intelligible world of Ideas (Gottschalk,
Chapter 3). Earlier than that again, we find the tendency to harmonise the
various schools in the work of the Platonist Antiochus (died c. 67 bc ). But, as

33
Philoponus in Cat. 8,7–8.
34
For a history of the study of authenticity in connexion with Aristotle’s works, see Jill Kraye, Pseudo-
Aristotelian Studies, London, forthcoming.
35
A.J. Minnis, Medieval Theories of Authorship, London 1984, citing R.W. Hunt, ‘Introductions to the
Artes in the twelfth century’, Studia Mediaevalia, Festschrift for R.J. Martin, Bruges 1948, 85–112;
E.A. Quain, ‘The medieval accessus ad auctores’, Traditio 3, 1945, 243–56. Praechter (Chapter 2)
finds Boethius’ influence as late as the sixteenth century.
36
I. Hadot, ‘Les introductions aux commentaires exégètiques chez les auteurs néoplatoniciens et les
auteurs chrétiens’, in M. Tardieu (ed.), Les régles de l’interpretation, Paris 1987.
8 Aristotle Transformed

Saffrey points out in Chapter 8, harmony was easier to maintain before

Andronicus’ edition, when Aristotle was known chiefly by his early dialogue
works, written while he was still a member of Plato’s Academy.
School practices are revealed not only in the introductions to commentaries,
but also in the use of the expression ‘from the voice of ’. Nearly all of Ammonius’
commentaries are described in the manuscripts as being ‘from the voice of ’ or
‘from the seminars of ’ Ammonius. In other words, they are Ammonius’ lectures
written up as commentaries by his students.37 Sometimes the name of the
student, Ascelpius or Philoponus, is supplied, and in three of the four cases where
Philoponus is mentioned, we are warned that he has added ‘some reflections
of his own’,38 so that it becomes much harder to decide what belongs to
Ammonius. The tradition of encouraging students to write up one’s lectures and
comments is found as early as Plutarch of Athens and Syrianus. Plutarch read the
Phaedo with Proclus, encouraging him to take notes, so that there might
eventually be a commentary by Proclus on the Phaedo, while Proclus’ Timaeus
commentary represents his record of Syrianus’ teaching, along with his own
critical remarks.39
Relations between teacher and pupil were close in other ways. The
philosophical chairs in Athens and Alexandria were latterly passed from teacher
to pupil. Plutarch of Athens taught Syrianus (his successor) and Proclus. Syrianus
taught his own successor Proclus and Hermeias who took the chair at Alexandria.
Proclus taught Hermeias’ successor, Ammonius, while Ammonius taught
Damascius, who held the chair at Athens, and Olympiodorus, an indirect
successor of his own, as well as the two non-chair-holders, Philoponus and
Simplicius. These ties were in some cases cemented by family connexions.
Hermeias married a relative of Syrianus, originally intended for Proclus, while
Ammonius was Hermeias’ son.
The Physics commentary of Philoponus is not described as being ‘from the
seminars of Ammonius’, but, according to Verrycken in Chapter 11, it incorporates
an early redaction heavily influenced by Ammonius. É. Évrard has shown that
it already exemplifies the custom of double discussion later systematised in
Olympiodorus, Elias, David and Stephanus, and the same can be said of

37
The classic article on this is M. Richard, ‘Apo phônês’, Byzantion 20, 1950, 191–222 (unpublished
English translation by Victor Caston), although the explanation is already given by Praechter in
Chapter 2.
38
The four commentaries are in An. Pr., in An. Post., in DA, in GC. The last three warn of Philoponus’
additions.
39
Marinus, Life of Proclus, chs 12 and 13.
 The ancient commentators on Aristotle 9

Philoponus’ de Anima commentary, which is described as being from the

seminars of Ammonius with the addition of personal reflections by Philoponus.40
Here, indeed, we even have a technical name for the first of the two discussions:
protheôria – a preliminary general discussion of doctrine (424, 4 and 13). The
institution of a general discussion gives the inventive Philoponus a grand
opportunity to introduce his own ideas. The general discussion is followed by a
textual discussion (called by later commentators exêgêsis tês lexeôs) of the same
portion of Aristotle. Unlike the medieval Latin question commentaries and
literal commentaries,41 the two discussions in the Greek tradition are intended
for the same audience. Of the two discussions the general one is preceded by a
single lemma or quotation from Aristotle, while the textual discussion may have
several lemmata, which are sometimes wrongly printed in the CAG edition, as if
they were part of the discussion. Readers of these commentaries should be
alerted to look up both discussions, in order to discover Philoponus’ full view, as
well as consulting such major digressions as the corollaries on place and vacuum,
which interrupt the Physics commentary.
Évrard finds that the commentary is divided into portions which would take
about an hour each to deliver, regardless of whether that splits up or bundles
together the topics which Aristotle is discussing. These hour-long portions must
correspond to a single lecture or lesson (later called a praxis). If Praechter (in
Chapter 2) finds less evidence of hour-long portions in commentaries by some
later authors, it may be that they have been written up in a way that disguises the
lecture presentation more fully. As Praechter noted elsewhere,42 the commentaries
of Simplicius do not exhibit the lecturing structure, and this, along with his
references to readers, rather than to an audience, has suggested that he was
writing outside of a classroom context.
An idea of the topography of a school may be gained from the excavation by
Jean and Janine Balty at Apamea, which literary evidence favours as the city
where Iamblichus taught,43 with Dexippus among his pupils. The Baltys have
suggested that the beautiful mosaic of Socrates marked the site of the school and

40
Philoponus’ general discussion of touch, for example, runs from in DA 407,18 to 422,10 (although
the editor, Hayduck, does not put a proper break here), the exegesis of text from 422,11 to 436,33.
For the Physics commentary, see E. Evrard, L’école d’Olympiodore et la composition du ‘commentaire
à la physique’ de Jean Philopon, Diss., Liège 1957.
41
On these see Anthony Kenny and Jan Pinborg, ‘Medieval philosophical literature’, in Norman
Kretzmann, Anthony Kenny, Jan Pinborg (eds), The Cambridge History of Later Medieval Philosophy,
Cambridge 1982, 11–42.
42
Karl Praechter, ‘Simplikios’, RE 3A, 1 (2nd series) 1927, cols 204–13.
43
See, for example, John Dillon, ‘Iamblichus of Chalcis (c. 240 – 325 A.D.)’, in Wolfgang Haase (ed.)
Aufstieg und Niedergang der römanischen Welt, II 36.2, Berlin 1987, 862–909, at 869–70.
10 Aristotle Transformed

was installed by the Emperor Julian, thirty-five years after Iamblichus’ death and
after the dispersal of his best associates.44 If this is right, I would suggest that the
mosaic found next to it45 represents rhetoric, rather than philosophy, classes (see
illustration). It shows Persuasion (Peithô) and Judgment (Krisis) awarding a
beauty contest to Cassiopeia, who is disrobing to display a comely body that does
not immediately suggest the spiritual philosophy of Iamblichus. In the
conventional story, Cassiopeia was disqualified, and it was a standard part of a
rhetorical student’s training to be given a theme from mythology and told to
argue that the verdict should have gone the other way. The total area of mosaic
seems too vast for the needs of a philosophy school alone, and rhetoric classes
might well go on side by side with philosophy classes. Charlotte Roueché has
drawn my attention to a later analogue in fifth-century Alexandria. There we
learn that, although lecturers often taught in their homes on Fridays, a philosophy
class could be going on in the central school within earshot of a philology class.46
At Alexandria, someone like Philoponus would be likely to teach both philosophy
and philology, though holding a chair only in the latter. Similarly at Apamea,
Iamblichus and his associates might well have helped in the rhetoric teaching.
Iamblichus himself wrote a rhetorical work, as did two of his associates. For the
general student, rhetoric would have been the more vocational and popular
course, but it was also considered a necessary complement to philosophical

44
Jean-Charles Balty, ‘Julian et Apameé, aspects de la restauration de l’Hellénisme et de la politique
antichrétienne de l’empéreur’, Dialogues d’histoire ancienne, 1974, 267–304; Janine et Jean-Charles
Balty, ‘Un programme philosophique sous la cathédral d’Apameé; l’ensemble néoplatonicien de
l’empéreur Julien’, in Texte et Image, Actes du colloque international de Chantilly (13–15 Octobre
1982), Paris 167–76. I am deeply indebted to the Baltys for their hospitality at the site.
45
The mosaics are also displayed consecutively in the museum at Apamea.
46
Zacharias, Life of Severus, p. 23.
 The ancient commentators on Aristotle 11

studies, since even in those days the philosopher’s powers of expression left
room for improvement.47
If the connexion with rhetoric is made out, this will bear on the survival of the
school at Apamea. A larger institution involving rhetoric as well as philosophy
would have been better able to survive the persecution of pagan philosophers
under the Christian Emperor Constantine. Those teachers who stayed, even the
rhetoricians, could, some of them, have been used to teaching philosophy. Hence
at the time of Julian’s munificence, philosophy classes would either have been
continuing, or at least revivable. And Julian will have had more incentive to
install the mosaics, if he was not merely in the position of honouring the memory
of his favourite philosopher, the pagan Iamblichus, but attempting to restore to
an ongoing school its former prestige in philosophy, as part of his short-lived
drive to restore paganism.
I do not think we have to suppose that philosophy teaching at Apamea was
even interrupted, although the standard view is that Aedesius moved Iamblichus’
school to Pergamum at a time of repression under the Christian Emperor
Constantine and his successors.48 In fact, Eunapius says less than this. He
mentions the well-known suppression of pagan temples and disapproval of the
philosopers’ rites,49 and he says that Iamblichus’ former pupils were dispersed all
over the Empire.50 But he does not say that the dispersal was due to repression,
and he may intend it rather as a sign of the school’s influence. As for Aedesius, he
says that he inherited Iamblichus’ school, that out of conviction he turned to the
holier life of a goatherd or cowherd, apparently in Cappadocia, and that he was
persuaded to return to philosophy, after which he set up a chair in Pergamum.51
This is not to say that he moved Iamblichus’ school there. He may have left it
functioning when he took up the life of a herdsman.

The commentators and Christianity

The story of Iamblichus’ school is, however, interesting as a phase in the long
history of interaction between the Neoplatonist commentators and Christianity.
To see the sun rise behind the colonnades at Apamea is to be reminded of

47
See I. Hadot in Chapter 12, 300–1, and further references there. Iamblichus’ rhetorical work is
recorded in Syrianus, in Hermogenem, I, 9, 10ff.
48
e.g. John Dillon, op. cit., 871.
49
Eunapius, Lives of the Sophists, 461; 471; 472 (Wrights’ Loeb translation is bound with Polystratus).
50
Eunapius 461; 462.
12 Aristotle Transformed

Iamblichus’ insistence on performing the traditional animal sacrifices. Eunapius

portrayed him sacrificing at dawn in one of his suburban villas, accompanied by
his associates who had urged him to admit them to his private rites.52 In
sacrificing animals, he was going not only against the Christian Church, which
reaffirmed its opposition to animal sacrifice at the Council of Nicaea in 325 ad,
the probable year of his death, but also against his own pagan teacher Porphyry.
For Porphyry had written a whole book against the sacrifice and eating of
animals.53 On the other hand Porphyry’s attack on animal sacrifice, congenial as
it would have been to Augustine,54 does not show him to be a friend of the
Christians. On the contrary, his onslaught on them in Against the Christians, and
still more in Philosophy from Oracles, was considered the most formidable assault
from the side of pagan Greek thought in the whole of antiquity. His books were
burnt both by Constantine and, a hundred years later, by Theodosius, and he was
answered by many of the main Christian theologians, including Augustine.
Augustine’s own conversion to Christianity was as much a conversion to
Neoplatonism. He describes in Confessions VII 9 the revelation provided by
some ‘books of the Platonists’ which he obtained in Latin translation, and for a
long time he saw no obvious difference between Neoplatonism and Christianity.
But in the City of God, which began to appear in 413, he is aware that there is
much in Porphyry with which he must disagree, even if he concurs with Porphyry
on other points, including, apparently, the right attitude to sacrifice.
Later in the century, the Neoplatonist commentator Proclus had to leave
Athens for a year, to avoid persecution at the hands of Christians for impiety.55
He followed Iamblichus56 in seeing theurgic practice as the route to mystical
union with the gods. He could not foresee that within 25 years of his death

51
Eunapius 461; 464–5; 469.
52
Eunapius 458–9.
53
Porphyry, On Abstinence from Animal Food (English translation by Thomas Taylor). Iamblichus on
the other side claims Pythagoras’ support for the view that human souls do not transmigrate into
the bodies of sacrificial animals, so that these can be killed and eaten (Iamblichus, Life of Pythagoras,
ch. 18, 85). Indeed, he denies that human souls transmigrate into any animals at all (Aeneas of Gaza,
Theophrastus, PG 85, 893A-B; Nemesius Nat. Hom. 51, p. 117 Matthaei). But whether this too
contradicts Porphyry depends on one’s interpretation of Porphyry (see W. Deuse, Untersuchungen
zur mittelplatonischen und neoplatonischen Seelenlehre, Mainz, Wiesbaden 1983, 129–167, and the
briefer overview in Andrew Smith, ‘Porphyrian Studies since 1913’, in Aufstieg und Niedergang, as
above, II , 36.2, 717–73, at 725–7, with further literature cited there).
54
Porphyry’s preference for a spiritual sacrifice (Abst. 2.61) fits perfectly with Augustine’s quotation
from Psalms 51,18 (City 10,4–6). On vegetarianism, however (City 1,20), Augustine seems to reflect
the arguments which Porphyry ascribes to his opponents (Abst. 1.6; 1.18), despite doubts that have
been expressed about his knowledge of this particular treatise of Porphyry (see P. Courcelle, Late
Latin Writers and their Greek Sources, Cambridge Mass. 1969, p. 188, n.177, translated from Les
lettres grecques en Occident, Paris 1948).
55
Marinus, Life of Proclus, 15.
56
e.g. Iamblichus de Mysteriis 2.11.
 The ancient commentators on Aristotle 13

pseudo-Dionysius would appropriate his ideas for Christianity and even apply
his word ‘theurgy’ to the Christian sacraments.57 Proclus’ writings do not appear
to contain many remarks on the Christians, but Alan Cameron has drawn
attention to the use among later Neoplatonists of coded references to Christian
domination,58 and, exploiting these, Saffrey has enlarged the count of barbed
references to the Christians in Proclus.59 One style of reference, ‘the prevailing
circumstances’, will be encountered below.
Coded references are subsequently used by Damascius, the last head of the
Neoplatonist school in Athens. He recounts with pride the story of Proclus’
contemporary Hierocles of Alexandria, who ran up against ‘those in power’ in
Constantinople, spattered with his own blood the (Christian) magistrate who
had him beaten, and called him a Cyclops.60 Damascius further complains
bitterly that the Neoplatonist professor in Alexandria, Ammonius, had come to
some discreditable agreement with the bishop in charge of ‘the prevailing
opinion’ in the 490s for financial reasons.61 It has proved difficult to discover
what the concession was. Westerink (in Chapter 14) rejects the suggestion that
Ammonius agreed to stop teaching Plato, while his own suggestion that
Ammonius agreed to be silent on the pagan belief in the eternity and divinity of
the world is discarded by Verrycken (in Chapter 10). But I think there is no need
to assume that the concession bore on Ammonius’ teaching. It might have
involved an agreement not to parade worship of the pagan gods, and theurgic
belief and practice, in the way that Ammonius’ teacher Proclus had done in
Athens and Iamblichus in Syria, or more simply not to make trouble with
Christians, if he wanted funding.
In a seminal article, Karl Praechter argued that the Alexandrian school made
concessions to Christian monotheism, abandoning the Neoplatonist distinction
between a supreme God, the One, and a second God, the Intellect.62 He further
argued that this concession was already made by the earlier Alexandrian,
Hierocles, in the first half of the fifth century, and that it was continued by

57
pseudo-Dionysius, Letter 9.1, PG 3, 1108A. This and many other references to Christian theurgy are
cited by H.D. Saffrey, ‘New objective links between the pseudo-Dionysius and Proclus’, in D.J.
O’Meara (ed.) Neoplatonism and Christian Thought, Albany NY 1982, 64–74 and 246–8, at 71–2.
58
Alan Cameron, ‘The last days of the Academy at Athens’, Proceedings of the Cambridge Philological
Society n.s. 15, 1969, 7–29. Verrycken, however, points out in Chapter 11 that the same phrases were
sometimes used by Christians.
59
H.D. Saffrey, ‘Allusions antichrétiennes chez Proclus le diadoque Platonicien’, Revue des sciences
philosophiques et théologiques 59, 1975, 553–63.
60
Damascius Life of Isidore 83, 5–11 (Zintzen).
61
Damascius Life of Isidore 250,2; 251, 12–14.
62
Karl Praechter, ‘Richtungen und Schulen im Neuplatonismus’, Genethliakon für Carl Robert,
Berlin 1910.
14 Aristotle Transformed

Simplicius, whose commentary on Epictetus he took to be written in Alexandria

under the supervision of Ammonius. The thesis about Simplicius and Hierocles
has already been overturned by I. Hadot,63 but its application to Ammonius is
now challenged (in Chapter 10) by Verrycken. Ammonius openly recognises the
pagan distinction between the One and the Intellect. He even ascribes such a
distinction to Aristotle, although both divinities, so Verrycken suggests, may be
contained within Aristotle’s idea of the unmoved mover.
Ammonius’ pupil Philoponus was a Christian, and his view that the physical
world, including prime matter itself, had a beginning was a distinctively Christian
one. Residing in Alexandria, he eventually worked out a comprehensive
alternative to Aristotelian physics, a scientific world view tailored to cohere with
his Christian beliefs.64 What Verrycken argues in Chapter 11, however, is that
these ideas involve a complete volte-face on Philoponus’ part. Though a Christian
from the start, he was content at first to reproduce in his commentaries the ideas
of Ammonius. We can see him changing his mind on the eternity of the world,
on Aristotle’s arguments about the need for an infinite force, on the definition of
place, on the impossibility of a vacuum, on the necessity for a fifth element to
constitute the rotating heavens and on the harmony of Plato and Aristotle.
Verrycken further conjectures that these changes came fairly late in Philoponus’
career, starting with the treatise Against Proclus on the Eternity of the World in
529. Thereafter the commentaries on the Posterior Analytics, the Physics and
probably the Meteorology were revised. At any rate, they all contain expressions
of the later view, but, in the first two cases at least, these are found alongside
unrevised expressions of the earlier view.
Verrycken’s case for a volte-face and for a revision of at least two commentaries
is impressively documented. I believe the de Anima commentary may also prove
to contain a later stratum, if there is a reference to the Christian Creator at
330,24, to earlier lectures on the Meteorology at 339,35 and to the Categories
commentary at 391,32.65 Whether or not the change began in 529, that year was
certainly a significant one. Not only did Philoponus publish his onslaught on
Proclus, the greatest of the Athenian Neoplatonists, in defence of the Christian
view, attacked by Proclus, that the physical universe had a beginning, but 529

63
I. Hadot, Le problème du néoplatonisme Alexandrin: Hiéroclès et Simplicius, Paris 1978.
64
Richard Sorabji, ‘John Philoponus’, in id. (ed.), Philoponus and the Rejection of Aristotelian Science,
London and Ithaca NY 1987.
65
Henry J. Blumenthal, however, following Koenraad Verrycken, takes it to be wholly early in
‘Simplicius and others on Aristotle’s discussions of reason’, in Gonimos, Neoplatonic Studies
Presented to Lendeert G. Westerink at 75, Arethusa, Buffalo NY 1988, 103–19.
 The ancient commentators on Aristotle 15

was the year in which the Christian Emperor Justinian finally closed the pagan
Neoplatonist school in Athens. Damascius, the head, along with Simplicius,
Priscian and four others, left shortly afterwards for the court of Khosroes in
Ctesiphon, in present-day Iraq, and archaeological evidence, admittedly
conjectural, has been offered to suggest that they did not return. At any rate, if
the house excavated in Athens is really the one handed down by Plutarch of
Athens to his Neoplatonist successors, the statues carefully placed down a well
were never recovered by them.66
No wonder Simplicius, who was one of the seven philosophers, employs such
bitter invective in his attacks on Philoponus.67 He sees Christianity and
Philoponus its representative as thoroughly irreverent. They deny the divinity of
the heavens, and glorify the things of the body above the things of the mind by
the revolting habit of preserving the relics of martyrs. His own commentaries, in
contrast to the sober and argumentative tone of the Christian Philoponus,
culminate in devout prayers.
In Alexandria, unlike Athens, Neoplatonist teaching continued with little
disturbance for the rest of the sixth century. Westerink describes in Chapter 14
the open way in which the pagan professor Olympiodorus addresses the
Christian pupils in his class. When he speaks of guardian spirits or demons, they
are to think of angels, or of conscience. This he describes, using a code word, as
an interpretation adapted to ‘the prevailing circumstances’. Olympiodorus was
the last of the pagan commentators in Alexandria. The next commentators, Elias,
David, pseudo-Elias and Stephanus were all Christians. But, as Westerink shows,
though using the traditional format and formulae, they did not address
themselves in Olympiodorus’ explicit way to the relation between pagan and
Christian belief.68
The most important remaining example of interaction with Christianity
concerns the Christian Boethius, who wrote in Latin at the beginning of the
sixth century, and passed the commentary tradition on to the Latin-speaking
world. Like the early Augustine, he writes as if there were no difference between

66
Alison Frantz, ‘Pagan philosophers in Christian Athens’, Proceedings of the American Philosophical
Society 119, 1975, 29–38. The identification is, however, conjectural, and it had been doubted that
the scratch marks defacing the floor represent crosses superimposed by Christian occupiers.
67
Philippe Hoffmann, ‘Simplicius’ polemics’, in Richard Sorabji (ed.), Philoponus and the Rejection of
Aristotelian Science, London and Ithaca NY 1987, 57–83.
68
I am not persuaded by the examples of ‘Christianising’ in David offered by Haig Khatchadourian
and A.K. Sanjian, which pertain to ethics and at Prolegomena 115,12–13 (cf. 129,8–10) to the
Creation of man. See the interesting discussion in A.K. Sanjian (ed.), David Anhaghtc, the ‘Invincible’
Philosopher, Atlanta 1986, 55 and 109.
16 Aristotle Transformed

Christianity and Neoplatonism. Some of his treatises are works of Christian

theology, while his commentaries on Aristotle’s logic are works of Neoplatonism.
In his Consolation of Philosophy, paganism and Christianity are combined. The
Consolation has indeed been viewed as a pagan work, on the grounds that it
makes the world eternal. But quite apart from the fact that some Christians
accepted this view, I think that Boethius merely described the eternity of the
world as a hypothesis of Plato and Aristotle.69
In this survey of the relations between the Neoplatonist commentators and
Christianity, we have found a wide spectrum and even a progression of attitudes.
There is Porphyry’s offensive against Christianity, the commitment of Iamblichus
in Syria and Proclus in Athens to pagan ritual under increasing pressure, the
vehemence of Damascius and the bitterness of Simplicius when the Athenian
school was closed. We have seen the Christian counter-attack by Philoponus in
Alexandria and the pacific agreement to differ of the Alexandrians Ammonius
and Olympiodorus. Finally there was the undramatic absorption of the Alexandrian
chair by Christian professors. The Latin-speaking Christian Boethius stands aside
from this progression, since he was one of those who saw Neoplatonism as one
with Christianity and helped it to penetrate deep into Christian sensibility.

Dangers in reading the commentators

Evidently, the theological motive of the Neoplatonist curriculum and the

pressure to harmonise Plato with Aristotle creates dangers, if the commentaries
are read as straightforward guides to Aristotle, without due allowance being
made. Nor is it safe to extract from them the fragments of the Presocratics, or of
other authors, without taking account of the Neoplatonist background against
which the fragments were originally selected for discussion. For different
reasons, analogous caveats apply to fragments preserved by the pre-Neoplatonist
commentator Alexander.70 Henry Blumenthal warns in Chapter 14 of the
amount of Neoplatonist doctrine to be found in the de Anima commentaries,
and shows that Anglo-Saxon scholars have often been unaware of the
Neoplatonist bias. The opposite danger, exemplified by Cousin and Ravaisson, is
accepting the harmony of Plato and Aristotle as literal truth.

69
Richard Sorabji, Time, Creation and the Continuum, London and Ithaca NY 1983, 196.
70
For Alexander’s treatment of the Stoics, see Robert B. Todd, Alexander of Aphrodisias on Stoic
Physics, Leiden 1976, 24–9.
 The ancient commentators on Aristotle 17

The distorting Neoplatonist context, however, does not prevent the commentaries
from being incomparable guides to Aristotle. The commentators’ introductions to
Aristotle’s philosophy insist that the commentator must have minutely detailed
knowledge of the entire Aristotelian corpus, and this they certainly have. Moreover,
commentators are enjoined neither to accept nor to reject what Aristotle says too
readily, but to consider it in depth and without partiality. The commentaries draw
one’s attention to hundreds of phrases, sentences and ideas in Aristotle, which one
could easily have passed over, however often one read him. The scholar who makes
the right allowance for the distorting context will learn far more about Aristotle
than he would on his own.
The de Anima commentaries, which Blumenthal studies, are likely to involve
more distortion than logical commentaries, or than the non-theological parts
of Physics commentaries, just because Plato had less to say on logic and physics.
In fact, the Neoplatonist distortion is so much greater in the de Anima
commentary ascribed to Simplicius than in his other commentaries that this has
been given as one of several reasons for suspecting that it may be by another
author (Priscian). But this hypothesis is contested by I. Hadot in Chapter 12
below.71

The individual commentators

Having described some of the overall trends, I turn now to the individual
commentators. Those whose commentaries we possess fall into three main
groups. First there are the second- and early third-century Aristotelians, Aspasius,
someone using Adrastus, Alexander of Aphrodisias, and, writing in the same
tradition but at a later date, Themistius. Then there is the largest group, the

71
The hypothesis is that of F. Bossier and C. Steel, in ‘Priscianus Lydus in de “In de anima” van
Pseudo(?)-Simplicius’, Tijdschrift voor Philosophie 34, 1972, 761–822, repeated in C. Steel, The
Changing Self, Brussels 1978. The point about the special character of de Anima commentaries is
made by I. Hadot in the appendix to her Le problème du néoplatonisme Alexandrin: Hiéroclès et
Simplicius, Paris 1978, where she resists all but one of the arguments of Bossier and Steel. In
Chapter 12 below, she resists the remaining argument, and the conclusion as well. But one issue, not
previously mentioned, on which the de Anima commentary may differ at least from the Physics
commentary is whether it is qualities (in Phys. 623,11–18) or matter (in DA 134,6) by which bodies
are prevented from occupying the same place. It may also be wondered how the doubt that light is
a body (in DA 134,13–20) is to be squared with the endorsement of Plato’s definition of light as a
kind of fire (in Cael. 12,28; 16,20–1; 130,31–131,1; but cf. in DA 133,32; 134,6). See Richard Sorabji,
Matter, Space and Motion, London and Ithaca NY 1988, 106–7; 108. It may also be wondered why
the de Anima commentary at 141,15–38, does not, like Priscian Metaphrasis 14,10–12, credit the
wave theory of sound to Theophrastus.
18 Aristotle Transformed

commentators writing in the Neoplatonist manner, some of the later ones

Christians, between 250 and about 610, when Stephanus was called from
Alexandria to Constantinople. Finally, there was a revival of commentary-
writing in eleventh-century Byzantium, which reached its peak in the
commentaries of Eustratius and Michael of Ephesus in the twelfth century.
Some of the earliest commentary we have is on Aristotle’s Nicomachean
Ethics. Earliest of all is that of Aspasius of Athens from the first half of the second
century ad, covering Books 1–4, 7 and 8. Then there is the commentary on EN
2–5, printed in CAG 20, which may derive from Adrastus of Aphrodisias.72
Although Gottschalk73 and in (Chapter 18) Mercken reject the hypothesis that it
is actually by Adrastus, Mercken believes that it too may be as early as the second
century ad. Both commentaries are described by Mercken below.
Alexander of Aphrodisias, taught among others by Aspasius’ pupil Herminus,
was appointed to a chair of Aristotelian philosophy some time between 198 and
209. This may have been a chair in Athens, one of the four founded for the four
main philosophical schools by Marcus Aurelius (Emperor 161–180). The satirist
Lucian in The Eunuch represents these chairs as open to competition among any
exponents of the relevant schools who submitted themselves to examination by
a committee of first citizens.74 In Lucian’s version, a eunuch is a leading
competitor. Alexander’s brilliance at filling out Aristotle’s ideas to obtain a
coherent Aristotelian position can be illustrated from many examples. One (not
from a commentary) which I have given elsewhere is his defence of Aristotle’s
claim that space is finite against the objection that an edge of space is conceptually
problematic.75 The need for an edge in a finite universe was not finally
circumvented until the invention of Riemannian geometry in the nineteenth
century.
In fourth-century Constantinople, Themistius (fl. late 340s to 384 or 385) saw
himself as the inventor of paraphrase, wrongly thinking that the job of full-scale
commentary was completed.76 His paraphrases, however, contained a substantial
amount of interpretation, including a whole excursus on the active intellect, and

72
Paul Moraux, D’Aristote à Bessarion, Laval Quebec 1970, 24f; Der Aristotelismus bei den Griechen,
vol. 2, Berlin 1984, 323–30.
73
Hans Gottschalk, in another part of the work from which Chapter 3 is a revised extract: ‘Aristotelian
philosophy in the Roman world from the time of Cicero to the end of the second century ad ’, in W.
Haase (ed.), Aufstieg und Niedergang der römischen Welt II 36,2, 1155, n. 363, commenting on
Kenny, The Aristotelian Ethics, Oxford 1978, 37 n. 3.
74
Lucian’s evidence is used by John P. Lynch, Aristotle’s School, Berkeley, Los Angeles 1972, 169–73.
75
Richard Sorabji, Matter, Space and Motion, London and Ithaca NY 1988, ch. 8.
76
Themistius in An. Post. 1,2–12. See H.J. Blumenthal, ‘Photius on Themistius (Cod. 74): did
Themistius write commentaries on Aristotle?’, Hermes 107, 1979, 168–82.
 The ancient commentators on Aristotle 19

they were widely used in later times. It has been estimated that Philoponus’
Physics commentary draws silently on Themistius 600 times,77 and he had a great
influence on Renaissance Platonism.78
In connexion with the Neoplatonist commentaries, we have already seen how
Porphyry (232–309) put the study of Aristotle on a sure footing within
Neoplatonism, insisting on the compatibility with Plato of Aristotle’s categorial
scheme, while his pupil Iamblichus (c. 240–c. 325) confirmed this harmonisation
with his ‘intellective theory’ of the categories. Iamblichus’ commentaries on
Aristotle are lost, but a collection of surviving fragments has been made by
Larsen.79 The only pupil of Iamblichus from whom a commentary on Aristotle
survives is Dexippus (c. 330: see Chapter 6).80 His Categories commentary
shares with Porphyry’s surviving one the special characteristic of being in
dialogue form.
The house used by the late Athenian Neoplatonists, claimed by Alison Frantz
to be identical with the one excavated by her, was bequeathed to the school by
Plutarch of Athens (died 432). No commentary of Plutarch’s on Aristotle survives,
but the evidence for a de Anima commentary is discussed by Blumenthal in
Chapter 13 below. In fact, most of the Aristotle commentaries of the Athenian
school are lost. This applies to the commentaries of Syrianus (died c. 437), apart
from that on Metaphysics 3, 4, 13, 14, and to those of Proclus (c. 411–485) and of
Damascius, who was head at the time of closure in 529.81
The Alexandrian commentators Ammonius (435/45–517/26) and Philoponus
(c. 490 to 570s) have already been described. Philoponus’ most scintillating
arguments turn the tables on the 800–year-old Aristotelian tradition by arguing
that on the Creation of the world Christianity must be right. If the physical
universe had no beginning, it would already have passed through a more than
finite number of years, and Aristotle declared that passing through an infinity
was impossible. Moreover, the number of years traversed by next year would be

77
H. Vitelli CAG 17, p. 992, s.v. Themistius.
78
This is brought out in the articles by E.P. Mahoney, which Blumenthal discusses in Chapter 5.
79
Bent Dalsgaard-Larsen, Jamblique de Chalcis, Aarhus 1972, gives fragments of the commentaries on
the Categories and Prior Analytics, along with others of uncertain origin. There was also a
commentary on the de Interpretatione, perhaps on the de Caelo, but probably not on the de Anima;
see John Dillon, ‘Iamblichus of Chalcis (c. 240–325 ad )’, in W. Haase (ed.), Aufstieg und Niedergang
der römischen Welt II 36.2, Berlin 1987, 862–909, at 872 and 877.
80
The commentary is translated by John Dillon, London and Ithaca NY 1989.
81
For the other Aristotle commentaries of Syrianus, as known from fragments, see Loredana Cardullo,
‘Syrianus’ lost commentaries on Aristotle’, Bulletin of the Institute of Classical Studies 33, 1986, 112–
24. For the evidence on Proclus’ Aristotle commentaries, see L.G. Westerink in Chapter 13, n. 18, for
that on Damascius’, L.G. Westerink, The Greek Commentaries on Plato’s Phaedo. vol. 2, Damascius,
Amsterdam 1977, 11–12.
20 Aristotle Transformed

greater than infinity. It would be another eight hundred years before medieval
philosophy had even assembled the materials to cope with the latter point.82
I think that Justinian’s closure of the school in Athens was effective.83 But it
has remained a mystery where Simplicius went to write his commentaries. They
seem to have been composed after the brief stay of the seven Athenian
philosophers at Ctesiphon came to an end, as it did by 532. In Chapter 12, I.
Hadot backs the thesis which has been argued by Michel Tardieu in a series of
three writings, that Simplicius went to Harrân (Carrhae), in present-day Turkey
˙
near the Iraq border.84 Some of Tardieu’s best evidence is the occurrence of
place-names from the region of Harrân in Simplicius’ commentaries, and some
˙
part of his further evidence is recorded by I. Hadot below. There is testimony
preserved in Syriac of a Neoplatonist school flourishing in Harrân as late as 943,
˙
and more than seventy years before that, it was from Harrân that Tâbit b. Qurra
˙ ˙
went to found the Platonising school in Baghdad, which became the driving
force behind the diffusion of Greek Philosophy in the Arabic world. If Tardieu is
right, the Athenian Neoplatonist school did not after all disappear from the map
of history. It survived in Harrân through Simplicius, accompanied, so Tardieu
˙
has suggested orally, by Damascius and Priscian. And it was this school’s teaching
which Tâbit b. Qurra transported to Baghdad. At the same time, if all this is
˙
correct, it may still be wondered if Simplicius’ role in the school will have been
that of a teacher, given the absence of signs of a teaching context in his
commentaries.
The continuation of the commentary tradition in Alexandria by Olympiodorus
(495/505–after 565), Elias, David, pseudo-Elias and Stephanus is described by
Westerink in Chapter 14. To the commentaries which he surveys, L. Tarán has
added an edition of an anonymous commentary from this period on the de
Interpretatione.85 Stephanus was called to a chair in Constantinople in 610 or
soon after, and this is usually considered as marking the end of the Alexandrian
school. In a forthcoming article, however,86 Mossman Roueché argues that

82
Richard Sorabji, ‘Infinity and the Creation’, inaugural lecture 1982, repr. in id. (ed.), Philoponus and
the Rejection of Aristotelian Science, London and Ithaca NY 1987, 164–78.
83
The best case against was made by Alan Cameron, ‘The last days of the Academy at Athens’, on
which I have commented in Time, Creation and the Continuum, 199–200.
84
M. Tardieu, ‘Sâbiens coraniques et “Sâbiens” de Harrân’, Journal Asiatique 274, 1986; id., ‘Les
Calendriers en˙ usage à Harrân d’après ˙ les sources˙ arabes et le commentaire de Simplicius à la
Physique d’Aristote’, in I.˙Hadot (ed.), Simplicius – sa vie, son oeuvre, sa survie, Berlin 1987; id.
Coutumes nautiques mésopotamiennes chez Simplicius in preparation.
85
L. Tarán (ed.), Anonymous Commentary on Aristotle’s De Interpretatione (Codex Parisinus Graecus
2064), Meisenheim am Glan, 1978.
86
Mossman Roueché, ‘The definitions of Philosophy and a new fragment of Stephanus of Alexandria’.
 The ancient commentators on Aristotle 21

pseudo-Elias and David may have followed Stephanus rather than preceding
him. Although pseudo-Elias was at one time writing outside Alexandria,
probably in Constantinople,87 he could subsequently have moved to Alexandria,
and the Alexandrian school could have continued into the period of the Arab
capture of the city in 640. Such a time scale would be facilitated by Westerink’s
new refusal below to date Elias’ activity as early as 541.
Much more important than the work of the last Alexandrians was the slightly
earlier activity of Boethius in Rome, who sought to make Aristotle accessible to
the Latin-speaking world. He announced a plan in his second commentary on
the de Interpretatione for translating all the Aristotle he could lay hands on
and Plato into Latin, with commentary.88 He got as far as translating all or nearly
all of Aristotle’s logical works, equipping some logical works and the Physics
with selective scholia or full commentaries and writing monographs on the
syllogism covering the de Interpretatione and Prior Analytics. But then he was
executed, on a charge which he denied, at about the age of 44 or 45 in 525 or 526.
His Consolation of Philosophy is a reflection written in prison, while awaiting
execution, on whether life is governed by fate, chance, or providence. Its
poignancy was to inspire English translations or paraphrases by King Alfred,
Chaucer and Queen Elizabeth I.
As Ebbesen explains in Chapter 16, some commentary had been available in
Latin before Boethius. There had in the fourth century ad been an earlier Latin
adaptation of the Isagoge and commentary on the Categories by Marius
Victorinus, a paraphrase of the Categories (Decem Categoriae) and Latin
translations by Praetextatus of Themistius’ paraphrases of the Prior and Posterior
Analytics.89 Still earlier, in the second century, Apuleius wrote a Latin exposition
of the de Interpretatione.90 But it was predominantly Boethius who handed on
the commentary tradition to the Latin world from the sixth to the twelfth
century. Not only does he preserve six standard points in his introduction to the
first commentary on the Isagoge and to the Categories commentary,91 he also
gives in his Isagoge commentaries a definition and division of philosophy,92 and

87
L.G. Westerink, Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge, Amsterdam 1967.
88
Boethius in Int.2 (Meiser), 79,9–80,1.
89
See P. Hadot, Marius Victorinus, Paris 1971.
90
See J. Isaac, Le Peri Hermeneias en occident de Boèce à saint Thomas, Paris 1953; M.W. Sullivan,
Apuleian Logic, Amsterdam 1967.
91
Boethius in Isag.1 I. 1, 4,17–5,10 (Schepss-Brandt); in Cat. PL 64, 161A–162A; most of the points are
also found in in Int.2 (Meiser), 1–13.
92
Definition, Boethius in Isag.1 I. 3, 7,11–16; division, in Isag.1 I. 3, 7, 11–12, 9; in Isag.2 I. 1–3, 135,5–
143,7.
22 Aristotle Transformed

describes the eventual goal of philosophical study as contemplation of God,93

while in his second commentary on the de Interpretatione, he expresses the
intention of demonstrating the harmony of Plato and Aristotle.94 He views
Aristotle’s Categories as only an introduction for beginners, the heights of
philosophy being reserved by Aristotle for his Metaphysics.95 Although only one
Categories commentary exists, Boethius planned a second one, to show the
higher meaning of the Categories. It was to draw on Pythagorean science, he says,
and so presumably would have followed the model of Iamblichus’ ‘intellective
theory’.96 A reconstruction of the projected commentary is offered by Ebbesen in
Chapter 16. Boethius’ other extant Aristotelian commentaries do in fact come in
pairs. There are two on Porphyry’s Isagoge and two on Aristotle’s de Interpretatione.
In each case, the first commentary is the more elementary. These commentaries
between them cover the syllabus which in the Latin West came to be called the
Logica Vetus, or Old Logic.
The late Alexandrian commentaries and prolegomena of Elias, David and
Stephanus were read in the Byzantine world for some centuries after c.610, when
Stephanus left Alexandria for Constantinople, and the results are evident in
various compendia.97 But the tradition of Greek commentary writing was
revived in that world only in the eleventh century. Of more than a dozen late
Byzantine commentators described in Hunger’s Byzantinisches Handbuch, the
CAG has edited only three.98 Of these by far the most important are Eustratius
(1050/1060–c. 1120) and Michael of Ephesus. Robert Browning has transformed
understanding (Chapter 17) by redating Michael and showing that both
commentators belonged to a circle organised by the Princess Anna Comnena
in twelfth-century Byzantium. Accordingly the completion of Michael’s
commentaries should be put at 1138 or later instead of 1040. Sten Ebbesen’s
study of Michael’s commentary on the Sophistici Elenchi shows him ‘vacuuming’
earlier scholia, amending them and stitching them skilfully together, to make

93
in Isag.1 I. 3, 8,6–19.
94
Boethius in Int.2 (Meiser), 80,1–6.
95
Boethius in Cat. PL 64, 252 B-C.
96
Boethius in Cat. PL 64, 160A-B.
97
See Mossman Roueché, ‘Byzantine philosophical texts of the seventh century’, Jahrbuch der
Österreichischen Byzantinistik 23, 1974, 61–76; ‘A middle Byzantine handbook of logical
terminology’, ibid. 29, 1980, 71–98.
98
Besides Eustratius and Michael, Sophonias is included from the thirteenth century. H. Hunger’s
survey is in Die hochsprachliche profane Literatur der Byzantiner, vol. 1 (=Byzantinisches Handbuch,
part 5, vol.1), Munich 1978, 25–41. He does not include the paraphrase commentary unreliably
attributed to the otherwise unknown Heliodorus of Prusa (CAG 19.2) of which all we can say is
that it must antedate 1366 (see R.A. Gauthier in R.A. Gauthier and J.Y. Jolif, Aristote, L’ethique à
Nicomaque 2, Paris 1970, introduction 106–7).
 The ancient commentators on Aristotle 23

one recension after another, the earliest being perhaps for a class attended by
Anna herself. At the end of the process Ebbesen finds no personal slant imposed
by Michael himself.99 On the other hand, in his commentaries on the animal
works and on the Politics, one would conjecture that he had fewer scholia to rely
on.100 In every case, his concern (Mercken, Chapter 18) is simply to expound
Aristotle. Eustratius, by contrast, is distinctive for his defence of Christianity
and of Platonism along with his penchant for syllogisms, all described by
Mercken.
Michael’s commentaries seem designed partly to cover areas where gaps
had been left, and this may have been at Anna’s direction. Not all of the gap-
fillers are extant, but we have commentaries on the great corpus of biological
works (Parts of Animals, Motion of Animals, Progression of Animals, Generation
of Animals, Parva Naturalia)101 and a small fragment of one on the Politics. The
lost Rhetoric commentary had a few antecedents, but the Rhetoric too had been
comparatively neglected.102 It was probably in this circle that a composite
commentary was constructed on the Nicomachean Ethics (CAG 19.1, pp. 158–
86; CAG 20; CAG 22.3), using Eustratius for Book 1, the second-century
derivative of Adrastus for Books 2 to 5, Eustratius for 6, Aspasius for 8, Michael
for 9, 10 and a further commentary on 5, and a makeshift for the remaining Book
7 (Mercken, Chapter 18, Ebbesen, Chapter 19).103 There is controversy over
Praechter’s view that Michael also supplied the pseudo-Alexandrian commentary
on Metaphysics 6–14, which completes Alexander’s own commentary on 1–5,
although Ebbesen tells me that he still104 agrees with Praechter, as does Mercken
in Chapter 18.

99
Sten Ebbesen, Commentators and Commentaries on Aristotle’s Sophistici Elenchi, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1 (CLCAG ), 268–85.
100
For the biological commentaries in general, see the introduction to A. Preus, Aristotle and Michael
of Ephesus on the Movement and Progression of Animals, Hildesheim 1981.
101
For the Parva Naturalia, there is a commentary on its first treatise, the de Sensu, extant by Alexander.
The paraphrase in Arabic of Aristotle’s History, Generation and Parts of Animals, which A. Badawi
has translated into French and identified as a translation of Themistius, may, according to others,
be merely an Arabic summary of the Arabic translation of those three works of Aristotle. See
A. Badawi, Commentaires sur Aristote perdus en grec et autres épîtres, Beirut 1971, and the review of
it by F.W. Zimmermann and H.V.B. Brown, ‘Neue arabischen Übersetzungen aus dem Bereich der
spätantiken griechischen Philosophie’, Der Islam 50, 1973, 313–24, at 323–4.
102
See C.A. Brandis, ‘Über Aristoteles’ Rhetorik und die griechischen Ausleger derselben’, Philologus,
1849, 34ff., and other references given by Gottschalk in Chapter 3, n.67.
103
Ebbesen sees Michael as a likely compiler, Mercken thinks this quite possible.
104
The similarities to Syrianus (died c. 437) have suggested to some that it predates Syrianus (most
recently Leonardo Tarán, review of Paul Moraux, Der Aristotelismus, vol. 1, in Gnomon 46, 1981,
721–50 at 750), to others that it draws on him (most recently P. Thillet, in the Budé edition of
Alexander de Fato, p. lvii). Praechter ascribed it to Michael of Ephesus in his review of CAG 22.2 in
Göttingische Gelehrte Anzeigen 168, 1906, 861–907. For Ebbesen, see Commentators and
Commentaries on Aristotle’s Sophistici Elenchi (= CLCAG, vol. 7), part 3, p. 87.
24 Aristotle Transformed

About this same time (1130), James of Venice went to Byzantium, perhaps
visiting Michael, as Browning suggests and Ebbesen agrees (Chapters 17 and 19),
and thereafter translated into Latin a commentary derived from Philoponus
on the Posterior Analytics, and scholia on the Topics and Sophistici Elenchi,
as well as writing his own commentaries on the Sophistici Elenchi and Posterior
Analytics. In the same century, Gerard of Cremona (died 1187) translated
Greek commentaries from the Arabic versions: Themistius on the Posterior
Analytics, Alexander on the de Sensu, as well as Alexander’s treatises On the
Intellect (= Mantissa 106–113) and On Time. In the next century, the thirteenth,
translations from both Greek and Arabic became a flood, and the Latin world
was no longer confined to Boethius’ repertoire. Grossteste and William of
Moerbeke both translated from the Greek, while from the Arabic side, translations
were made not only of Greek commentaries but also of works influenced by the
Greek commentaries, including commentaries on Aristotle by Averroes in
the thirteenth century, and non-commentary works by Avicenna and Ghazali in
the twelfth.105

This volume

Six of the chapters in this volume are new (1, 4, 10, 11, 19 and 20). Four are
translated for the first time from French or German (2, 6, 8, 12). Six are
substantially revised (3, 5, 9, 14, 15, 18). Others are revised in more minor ways,
except that in all Greek and Latin quotations have been translated, and where
these quotations formed a substantial part of the original (15, 17), the change is
significant. After my introductory Chapter 1, I have included a translation of
Praechter’s monumental review of CAG (Chapter 2), which anticipated in
outline much of the scholarly research that has been conducted since. In Chapter
3, Gottschalk has added some new reflections on Stoic categories to his work on

105
For Grosseteste, see Mercken, Chapter 18. For Moerbeke, see P. Thillet, Alexandre d’Aphrodise de
Fato ad Imperatores, version de Guillaume de Moerbeke, Paris 1963, with bibliography. For James of
Venice, see Ebbesen, Chapter 19, and L. Minio-Paluello, ‘Jacobus Veneticus Grecus’, Traditio 8, 1952,
265–304; id., ‘Giacomo Veneto e l’Aristotelismo Latino’, in Pertusi (ed.), Venezia e l’Oriente fra tardo
Medioevo e Rinascimento, Florence 1966, 53–74, both reprinted in his Opuscula, 1972. For Gerard
of Cremona, see M. Steinschneider, Die europäischen Übersetzungen aus dem arabischen bis Mitte
des 17 Jahrhunderts (repr. Graz 1956); E. Gilson, History of Christian Philosophy in the Middle Ages,
London 1955, 235–6 and more generally 181–246. For the translators in general, see Bernard G.
Dod, ‘Aristoteles Latinus’, in N. Kretzmann, A. Kenny, J. Pinborg (eds), The Cambridge History of
Later Medieval Philosophy, Cambridge 1982. Note that in translating Islamic dates based on Hegira,
it is necessary to take account of the Muslim year being shorter by 10 or 11 days.
 The ancient commentators on Aristotle 25

the beginning of of the commentary tradition. Sharples in Chapter 4 explores

the little-known persona of Alexander, and considers to what extent his
commentaries and Quaestiones may reflect his teaching practices. Blumenthal
has revised his presentation of Themistius as the last of the Aristotelians, to
take account of Mahoney’s evidence of a rival Neoplatonist influence on him.
P. Hadot’s Chapter 6 on Porphyry and Dexippus represents one of the most
important treatments of the harmony of Plato and Aristotle. In Chapter 7,
Ebbesen attempts a reconstruction of Porphyry’s semantics and his theory of
concept formation. In Chapters 8 and 9, Saffrey and Sorabji take the question of
the harmony of Plato and Aristotle from Syrianus and Proclus, who were still
partly sceptical, on to Ammonius who extended harmonisation to Aristotle’s
theology, and handed down the idea of Aristotle’s God as a Platonic Creator
through the Arabs to the thirteenth-century West. Verrycken presents an entirely
new view of Ammonius and Philoponus in Alexandria (Chapters 10 and 11),
conceding to Praechter only that Ammonius simplified Neoplatonist theology,
by cutting out the henads and triads of his teacher Proclus. I. Hadot provides a
comprehensive account of Simplicius in Chapter 12, and establishes his Aristotle
commentaries as being in Cael., in Phys., in Cat., in that order, with a lost in
Metaph. and, after the Physics and Metaphysics commentaries, in DA . Westerink’s
masterly survey of the Alexandrians and their introductions is brought up to
date in Chapter 14, and Blumenthal’s warning of Neoplatonist influence in the de
Anima commentaries is updated by his note in Chapter 13 on the authenticity of
Simplicius in DA .
All discussion of Boethius’ commentaries is indebted to Shiel’s exhaustive
comparison of the relevant texts, which are here presented by him for the first
time in translation. There has been less accord over the startling conclusion he
drew from that comparison, that Boethius was quite lacking in originality.
According to Shiel, Boethius simply borrowed his comments from Greek
annotations in the margin of his copy of Aristotle, translating into Latin as he
went, and in Chapter 15 Shiel explains why he is unmoved by subsequent
objections to his interpretation. In Chapter 16, I have included a presentation by
Ebbesen of the opposing case for Boethius’ originality.
Browning’s work on the circle of Anna Comnena altered the picture of the
late Byzantine commentators, and in Chapter 17 he has translated the funeral
oration from which he drew his evidence. In Chapter 18, Mercken gives an
account of the main Ethics commentaries, a corpus which, on his dating, includes
the earliest extant commentaries, written by Aristotelians of the second century
ad, and then leaps to some of the latest Greek commentaries from the hands of
26 Aristotle Transformed

Eustratius and Michael in the twelfth century.106 Mercken further brings out the
influence these commentaries exerted on Albert the Great and Bonaventure,
once they had been translated into Latin by Grosseteste.
Ebbesen takes a dim view of the logic in the Neoplatonist commentaries,
but none the less stresses its influence (Chapter 19). As in Averroes’ Arabic
commentaries (Sorabji, Chapter 9), so here, the work of Philoponus got
misascribed to Alexander. Philoponus’ comments on the Posterior Analytics,
thus disguised, were picked up by James of Venice, perhaps from the library of
Michael of Ephesus, and so promoted thirteenth-century realism about the
objects of scientific knowledge. By contrast the nominalism characteristic of the
eleventh to twelfth centuries was partly encouraged by the tradition of Porphyry
and Boethius, that Aristotle’s Categories is about words, insofar as they signify
things. This is in spite of the fact that Porphyry and Boethius themselves are
better described as conceptualists107 than as nominalists.
This completes the study of the individual commentators. Chapter 20
represents a harvest of another kind. Study of the commentators makes it
possible to trace for the first time the development of concepts and theories very
different from Aristotle’s, and in Chapter 20, Mueller is able to chart the evolution
of the view that mathematical objects have their being in the mind.

The importance of the commentaries

The 15,000 pages of the CAG are the longest corpus of Ancient Greek philosophy
that has not previously been translated into English or other modern European
languages. But a programme is now under way which aims to translate a
substantial proportion, including all the most important Greek commentaries,
with fragments preserved in Arabic and other languages, and with related
texts.108
I have already explained the value of the works as commentaries, but it will
now be clear that they are much more than commentaries. Commentary writing
was one of the ways of doing philosophy, and the works therefore represent the

106
The Ethics commentary assigned to Heliodorus of Prusa is probably later. It belongs to any time up
to the fourteenth century.
107
For conceptualism in other senses, see Mueller in Chapter 20 and Mercken in Chapter 18, reporting
A.C. Lloyd on Eustratius’ conceptualism.
108
This is the series ‘The Ancient Commentators on Aristotle’, edited by Richard Sorabji, published by
Duckworth and Cornell University Press, from 1987 onwards.
 The ancient commentators on Aristotle 27

thought of the Neoplatonist and Aristotelian schools, as well as expounding

Aristotle. Furthermore, they embed fragments from all periods of Ancient Greek
philosophical thought. That is how many of the Presocratic fragments were
assembled, for example, and the interest of these is widely known. But the other
fragments that remain provide a panorama of every period of Ancient Greek
philosophy.
The commentaries represent a missing link in the history of philosophy. The
Latin-speaking Middle Ages obtained their knowledge of Aristotle at least partly
through the medium of the commentaries. We have already seen how that
medium could pass on a transformed Aristotle, one whose God has become a
Creator of the world, and whose active intellect, it may be added, provides the
hope of immortality for the human soul.Without knowledge of the commentaries,
we cannot understand the Aristotle of the later Middle Ages. Again, the ancient
commentaries are the unsuspected source of ideas which have been thought,
wrongly, to originate in the later medieval period. It has been supposed, for
example, that Bonaventure in the thirteenth century invented the ingenious
arguments based on the concept of infinity which attempt to prove the Christian
view that the universe had a beginning. In fact, Bonaventure is merely repeating
arguments devised by the commentator Philoponus 700 years earlier and
preserved in the meantime by the Arabs.109 Bonaventure even used Philoponus’
original examples. Again, the introduction of impetus theory into dynamics,
which has been called a scientific revolution, has been held to be an independent
invention of the Latin West, even if it was earlier discovered by the Arabs or their
predecessors. But recent work has traced a plausible route by which it could have
passed from Philoponus, via the Arabs, to the West.110
The new availability of the commentaries in the sixteenth century, thanks to
printing and to fresh Latin translations, helped to fuel the Renaissance break
from Aristotelian science. For the commentators record not only Aristotle’s
theories, but also rival ones, while Philoponus as a Christian devises rival theories
of his own and accordingly is mentioned in Galileo’s early works more frequently
than Plato.111

109
Richard Sorabji, Time, Creation and the Continuum, London and Ithaca NY 1983, ch. 14.
110
Fritz Zimmermann, ‘Philoponus’ impetus theory in the Arabic tradition’, in Richard Sorabji (ed.),
Philoponus and the Rejection of Aristotelian Science, London and Ithaca NY 1987, 121–9.
111
William A. Wallace, Prelude to Galileo, Essays on Medieval and Sixteenth-Century Sources of Galileo’s
Thought, Dordrecht 1981, 136 (contrast 196–7); Charles Schmitt, ‘Philoponus’ commentary on
Aristotle’s Physics in the sixteenth century’, in Richard Sorabji (ed.), Philoponus and the Rejection of
Aristotelian Science, 1987.
28 Aristotle Transformed

The commentaries are also a source, we have seen, for the history of school
practices in the Neoplatonist period. Much of their significance was already
foreseen in Karl Praechter’s seminal review of CAG, here translated from the
German of 1909 as Chapter 2. He adds their utility as evidence for changes in
grammar and vocabulary. As regards the latter, the indices of the current
translation series are designed to help keep track of changes in word-meaning.
Readers will find in Simplicius’ commentaries not only a record of earlier
Greek thought, but also, in the case of the Categories commentary, an admirable
gymnastic for students. For the Categories is a central, and apparently simple,
metaphysical text of Aristotle’s, but the commentary tradition represented by
Simplicius takes it apart with a thoroughness unmatched in modern discussions,
and forces the reader to think through the very basis of the metaphysics. It is of
immediate philosophical interest and requires no background knowledge, to
join in the ancient discussion of whether Aristotle’s other categories can all be
reduced to relatives, or whether the definition of relatives can itself be made
water-tight. Philosophical work on these ancient debates is just beginning to
come out in quantity, and it is the mirror image of modern work on whether
relational properties can be eliminated as merely supervenient, or whether
quantum physics proves the contrary.112
As the commentators become more accessible, I expect more information to
come to light on the history of concepts and theories. In forthcoming work, John
Ellis tells the story of a perceived clash between Aristotle’s Categories and his de
Anima.113 Individual qualities cannot exist separately from what they are in,
according to the Categories. But cannot an apple’s individual fragrance float off
into the surrounding air? Only if accompanied by some of the apple’s substance,
Ammonius suggests. But this conflicts, as Philoponus comes to see, with the de
Anima’s insistence that our sense of smell does not operate by direct contact
with the apple’s substance.
This provides only one instance of a wholesale re-evaluation of Aristotle’s
treatment of the five senses, in which I see two major tendencies.114 First, the

112
Paul Teller, ‘Relational holism and quantum mechanics’, British Journal for the Philosophy of Science
38, 1987, 71–81, and for work on the ancient debates see the number of publications since 1985 in
the bibliography under Logic – Categories and Isagoge.
113
John Ellis, ‘Can an apple’s fragrance be separated from it?’, in preparation.
114
Richard Sorabji, ‘Aristotle’s theory of sense perception’, in M. Nussbaum and A. Rorty (eds),
Aristotle’s de Anima, forthcoming Oxford 1990; ‘From Aristotle to Brentano: the development of the
concept of intentionality’, in H.J. Blumenthal and H.M. Robinson (eds), volume on Aristotle and the
Aristotelian traditions, forthcoming 1991.
 The ancient commentators on Aristotle 29

commentators dematerialised Aristotle’s account of the sensory processes: the eye

jelly does not become coloured; sound involves only a wave, not a moving block
of air. Secondly, they came to notice that the sensory processes are much more
diverse (witness smell) than Aristotle had allowed in their degree of materiality.
To the move towards dematerialisation the Arabic writers, Avicenna in particular,
contributed the idea of an intention, understood by him, and still by Thomas
Aquinas, as some kind of meaning or message. But by the time of Brentano in the
nineteenth century, the idea of awareness of that message had become prominent.
Thus it was only through a series of distortions by the commentators that
Brentano was able to read into Aristotle’s treatment of sense perception his own
seminal idea of an intentional object. He thought of an intentional object, like a
hoped-for fortune, as an object which does not have to exist in reality in order to
serve as the object of which our mind is aware. By reading this influential idea
into Aristotle, he lent authority to it, and to his proposal that the mental could be
distinguished from the physical by the fact of its being directed to intentional
objects of this kind. This little story illustrates that the role of the best commentators
is to reinterpret rather than to reflect Aristotle, and that, as in the case of Brentano,
reinterpretation can prove more fruitful than fidelity.
Philosophical commentary is important also in the Islamic and Indian
traditions; indeed the whole of traditional Indian philosophy takes the form of
commentary. A recent conference assessed whether commentary was compatible
with originality and found that in all three traditions it was.115 There is originality
not only of interpretation, but also of ideas. From the Greek commentaries
we may recall Philoponus’ introduction into dynamics of impetus theory, which
has been called a scientific revolution, his defence, endorsed by Galileo, of
motion in a vacuum as not implying infinite speed, and his stunning objections
to an infinite past. We may recall Simplicius’ treatment of prime matter as an
extension which, when endowed with properties, constitutes body, and
Alexander’s explanation of the edge of space. There is the discussion in Syrianus
and Proclus of whether bodies can coincide in exactly the same place and
the debate in many commentators on the reduction of other properties to
relatives.116

115
Conference held by Professor Daya Krishna in Jaipur, India, 10–14 March 1989, under the auspices
of the Indian Council of Philosophical Research.
116
I have written about all except the last in the following places: Time, Creation and the Continuum,
ch. 14; Philoponus and the Rejection of Aristotelian Science, chs 1 and 9; Matter, Space and Motion,
chs 1, 7, 8, 9 and 14.
30 Aristotle Transformed

The CAG texts

I shall conclude with a list of the commentaries on Aristotle in the CAG edition
and a list of the principal commentators with dates.

I Alexander in Metaphysica (in Metaph.) 1–5 and ps.-Alexander in

Metaph. 6–14, M. Hayduck, 1891.
II 1 Alexander in Analytica Priora (in An. Pr.), M. Wallies, 1883.
2 Alexander in Topica (in Top.), M. Wallies, 1891.
3 ps.-Alexander (Michael of Ephesus) in Sophisticos Elenchos (in SE ),
M. Wallies, 1898. (But see S. Ebbesen’s edition in Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.2, 153–99, of an earlier
version also by Michael.)
III 1 Alexander in de Sensu (in Sens.), P. Wendland, 1901.
Alexander in Meteorologica (in Meteor.), M. Hayduck, 1899.
Fragments of Alexander’s commentaries have now been collected as
follows:
2 Alexander on the intellect from in de Anima (in DA ), in Paul Moraux,
Alexandre d’Aphrodise, Exégète de la Noétique d’Aristote, Liège and Paris
1942, 203–21.
Alexander in Analytica Posteriora (in An. Post.), in Paul Moraux, Le
Commentaire d’Alexandre d’Aphrodise aux Secondes Analytiques,
Peripatoi 13, Berlin 1979.
IV 1 Porphyry Isagoge (Isag.) and in Categorias (in Cat.), A. Busse, 1887.
2 Dexippus in Cat., A. Busse, 1888.
3 Ammonius in Porphyrii Isagogen (in Isag.), A. Busse, 1895.
4 Ammonius in Cat., A. Busse, 1895.
5 Ammonius in de Interpretatione (in Int.), A. Busse, 1897.
6 Ammonius in An. Pr., M. Wallies, 1899.
V 1 Themistius in An. Post., M. Wallies, 1900.
2 Themistius in Physica (in Phys.), H. Schenkl, 1900.
3 Themistius in DA , R. Heinze, 1899.
4 Themistius in de Caelo (in Cael.), Latin and Hebrew, S. Landauer,
1902.
5 Themistius in Metaph. 12, Latin and Hebrew, S. Landauer, 1903.
6 ps.-Themistius (Sophonias) in Parva Naturalia (in PN ), P. Wendland,
1903.
VI 1 Syrianus in Metaph. 3, 4, 13,14, W. Kroll, 1902.
 The ancient commentators on Aristotle 31

2 Asclepius in Metaph. 1–7, M. Hayduck, 1888.

VII Simplicius in Cael., J.L. Heiberg, 1894. (New edition in preparation,
P. Hoffmann).
VIII Simplicius in Cat., C. Kalbfleisch, 1907.
IX Simplicius in Phys. 1–4, H. Diels, 1882. (New edition in preparation,
L. Tarán).
X Simplicius in Phys. 5–8, H. Diels, 1895. (New edition in preparation,
L. Tarán).
XI Simplicius (?) in DA, M. Hayduck, 1882.
XII 1 Olympiodorus Prolegomena (Proleg.) and in Cat., A. Busse, 1902.
2 Olympiodorus in Meteor., W. Stuve, 1900.
XIII 1 Philoponus in Cat., A. Busse, 1898.
2 Philoponus in An. Pr. 1, Philoponus (?) in An. Pr. 2, M. Wallies, 1905.
3 Philoponus in An. Post. 1, Philoponus (?) in An. Post. 2, Anon. in An.
Post. 2, M. Wallies, 1909.
XIV 1 Philoponus in Meteor., M. Hayduck, 1901.
2 Philoponus in de Generatione et Corruptione (in GC ), H. Vitelli,
1897.
3 ps.-Philoponus (Michael of Ephesus) in de Generatione Animalium (in
GA ), M. Hayduck, 1903.
XV Philoponus in DA 1–2, Philoponus (?) in DA 3, M. Hayduck, 1897.
XVI Philoponus in Phys. 1–3, H. Vitelli, 1887.
XVII Philoponus in Phys. 4 and fragments, H. Vitelli, 1888.
XVIII 1 Elias in Isag. and in Cat., A. Busse, 1900.
2 David Proleg. and in Isag., A. Busse, 1904.
3 Stephanus in Int., M. Hayduck, 1885.
XIX 1 Aspasius in Ethica Nicomachea (in EN ), G. Heylbut, 1889.
2 Heliodorus (?) in EN, G. Heylbut, 1889.
XX Composite commentary on EN, G. Heylbut, 1892:
Eustratius in EN 1, 6.
Derivative of Adrastus in EN 2–5.
Anon. in EN 7.
Aspasius in EN 8 is printed in XIX , 1.
Michael of Ephesus in EN 9–10.
XXI 1 Eustratius in An. Post. 2, M. Hayduck, 1907.
2 Anonymous Neobarii and Stephanus in Rhetorica (in Rhet.), H. Rabe,
1896.
XXII 1 Michael of Ephesus in PN, P. Wendland, 1903.
32 Aristotle Transformed

2 Michael of Ephesus in de Partibus Animalium (in PA ), in de Motu

Animalium (in MA ), in de Incessu Animalium (in IA ), M. Hayduck,
1904.
3 Michael of Ephesus, in EN 5, M. Hayduck, 1901.
XXIII 1 Sophonias paraphrase in DA, M. Hayduck, 1883.
2 Anon. paraphrase in Cat., M. Hayduck, 1883.
3 Incert. paraphrase in An. Pr., M. Wallies, 1884.
4 Anon. paraphrase in SE, M. Hayduck, 1884.

Supplement

I 2 Priscian of Lydia Metaphrasis in Theophrastum (in Theophr.) and

Solutiones ad Chosroem (Chosr.), I. Bywater, 1886.
II 1 Alexander de Anima (DA ) and Mantissa (Mant.), I. Bruns, 1887.
2 Alexander Quaestiones (Quaest.), de Fato (Fat.), and de Mixtione
(Mixt.), I. Bruns, 1892.

Principal Greek commentators

Andronicus of Rhodes edition perhaps 60 bc

Boethus Andronicus’ pupil, first century bc
Lucius and Nicostratos Platonist critics, early second century ad
Adrastus first half of second century ad
Aspasius first half of second century ad
Alexander of Aphrodisias appointed to chair between 198 and 209 ad
Porphyry 232–309 ad
Iamblichus c. 240–c. 325 ad
Dexippus fl. c. 330 ad
Themistius fl. late 340s to 384 or 385 ad
Plutarch of Athens died 432 ad
Syrianus died c. 437 ad
Proclus c. 411–485 ad
Ammonius 435/45–517/26 ad
Damascius head of Athenian school at time of closure,
529 ad
Philoponus c. 490–570s ad
Simplicius wrote after 529 ad
 The ancient commentators on Aristotle 33

Olympiodorus 495/505 – after 565 ad

Elias probably a pupil of Olympiodorus
Stephanus called to Constantinople in, or soon
after, 610 ad
pseudo-Elias either before or after Stephanus
David akin to pseudo-Elias, and close in date, but
not his source
Eustratius 1050/1060 – c. 1120 ad
Michael of Ephesus commentaries completed 1138 ad or later
34
 2

Review of the Commentaria in

Aristotelem Graeca*
Karl Praechter

The Berlin Academy is about to finish a tremendous undertaking which developed

organically from their edition of Aristotle, itself appearing between 1831 and
1870 and now the basis of all Aristotelian studies. Even in organising that earlier
edition, the significance of the commentators for the recensio as well as for the
interpretation of the Aristotelian texts could not remain unnoticed. Hence
extracts from some of the commentary literature were compiled under Brandis’
supervision in a fourth volume. In the framework of the Aristotle edition, they
occur only as a side concern. The commentators – they could not have been
otherwise in this context – appear only as a tool for research on the Philosopher,
without regard to their own value for the most diverse areas of study. Yet the
preface to the fifth volume already betrays a change in attitude towards the
commentary literature. We learn there that the remaining excerpts Brandis
collected did not go to press because a sampling of individual chapters simply
could not be a replacement for the commentaries in full, and a critical edition of
the latter was still, as before, lacking. Thus, instead of continuing these extracts,
Syrianus’ commentary on the Metaphysics appeared complete – as much as is
extant, at any rate – in the fifth volume, edited by Usener. It was with this in mind
that the Academy decided in 1874 to produce a new and independent edition of
the Greek commentators on Aristotle. A committee consisting of Bonitz, Zeller
(who put forward the proposal), and Vahlen was entrusted with management of
the enterprise and Adolf Torstrik with the redaction (after Torstrik’s death in
1877, Hermann Diels took his place). For the preparation, selection, and survey of
the extant tradition, and the collation of manuscripts, as well as for editing the

* Translated by Victor Caston.

35
36 Aristotle Transformed

individual parts of the corpus, a circle of colleagues who had been schooled
proficiently in philology was assembled immediately and extended over the years.
The execution of the plan went forward at a swift and sure pace. In 1882, the first
volume, comprising xxxiii plus 800 pages, had already resulted from their edition
of Simplicius on the Physics. By 1892, H. Usener was able to review eighteen
prepared and edited parts of the collection,1 and today the last part is in press.
The scholarly work whose precipitate is contained in this considerable series of
volumes represents an imposing amount of self-discipline. For all the works which
are included here, everything had to be built from the ground up. The examination
of the tradition’s often truly entangled relationships, which in many cases posed
very difficult riddles concerning the question of authorship, had to be handled
everywhere ex integro. Even where an editio princeps did not have to be prepared,
earlier editions offered only minimal help, as they rested on unsatisfactory
foundations. Usener’s own very meritorious edition did not remove the need for a
new foundation, as the collation of relevant manuscripts at his disposal had been
inaccurate. Considering how great the danger is of endlessly postponing the
conclusion with work as vast as that which had to be conducted here, by a series of
men and with the complicated organisation of such an undertaking, one must
express the highest admiration for Diels. His talent for organisation above all takes
the credit for this great project not only being finished, but even having reached its
conclusion within a very short allotment of time (relatively speaking); and yet he
found both the energy and the time to supervise the entire undertaking step by
step, indicating the way, checking and advising at every point, with the most
faithful care. The pleasant uniformity of the entire corpus is a closely related
benefit of this working procedure. In spite of the considerable number of co-
workers and the freedom which obviously had to be granted to those involved to
fulfil their tasks in detail, the collection has the appearance of the work of a single
hand; this not only affords aesthetic pleasure, but also uncommonly facilitates the
use of the commentaries. On a single point only it would have been preferable, and
also possible, to have kept the army of compilers within certain bounds through
more definite marching orders: I have in mind the arrangement of the indices
verborum. For external reasons, it was of course unavoidable that these indices
should be limited to a selection of what is lexically and grammatically important.2

1
Gött. gel. Anz., 1892, 1000f.
2
Grammatical matters are in general taken up only where they can be assigned to a particular Greek
entry (where an is missing with the optative, ean with the indicative, and so on).
 Review of the Commentaria in Aristotelem Graeca 37

Viewpoints, however, often differ on the ranking of importance, in accordance

with the subjective perception of the editor. To exclude this subjective element
completely and to regulate uniformly the production of these indices in detail, for
authors of very different periods and diverse idiosyncracies, would be inconceivable.
Nevertheless, it is an avoidable inconvenience when the indices to the different
works of a single author (as, for example, with Michael of Ephesus) or different
books of a single work (as, for example, with Michael’s commentary on the Ethics)
exhibit considerable variations between editors and lack the idiosyncratic
expressions3 and philosophical terms (such as dikaiosunê in the Ethics) recorded
there. This is a point, however, of minor significance. Whoever utilises these
indices for grammatical and terminological studies and for questions of literary
criticism should simply keep this state of affairs in mind, so as not to be led to false
conclusions ex silentio.
As for the rest, these editions deserve the highest praise in every respect for
their external arrangement as much as for their internal realisation. The
extremely practical organisation of reports on the manuscript tradition and
other aids which come into consideration for the recensio is especially to be
commended. This has been a persistently weak point in our editing methods.
Many otherwise excellent editions suffer from an aggravating clumsiness in the
organisation of the praefatio, with the result that the user who has been led only
occasionally in the course of his work to the writer in question and who wants to
form his own judgment on the authentic text of some passage or other must
unfortunately work through an introduction of fifty or more closely printed
pages, just to obtain a glimpse of the elementary facts of the tradition and to be
able to make use of the apparatus for a questionable passage. This precipice has
been circumvented with great skill in the commentary collection. The praefatio
of each part reports in the most concise manner, essentially indicating only the
results of the investigation, the worth and interrelationship of the textual sources
employed for the constitution of the text and represented in the apparatus by its
sigla. A supplementum praefationis adds a description of the collected manuscripts
that were not considered or mentioned in the praefatio and were not used for the
recensio (which are characterised, so far as is possible, by lists of variations for
particular sections), and gives more detailed reports on translations and editions.
This method, which is adhered to throughout the entire corpus (with trivial
discrepancies in the distribution of material between the praefatio and

3
See Gött. gel. Anz., 1906, 879, n. 1.
38 Aristotle Transformed

supplementum), unites splendidly the possibility of a quick and convenient

overview with a thorough introduction to the tradition, and should be considered
an exemplar for editions of ancient authors.4
These prefaces deserve as much recognition for their content as for their
form. Concerning the appraisal of some of the tradition’s sources, of course, one
may differ. There is evidence everywhere, however, of the industrious and
conscientious investigation and survey of the material, and of sensible judgment.
The same holds true for the editing of the author’s text which follows the praefatio
and supplementum. What deserves special notice above all is the distancing
from wilful conjecture, a superiority all the more worthy of recognition, since
these texts – which as regards grammar are based in essence on the ancient
tradition, yet in detail make the numerous concessions of its late period –
are often a truly great temptation for critical emendation among philologists
used to working on [earlier] antiquity. In this regard, these editions ought
to provide a methodological precedent, and editors of Byzantine texts
especially should be directed to the cautious and circumspect procedure followed
here.
Separate indices of words, names, and passages cited from Plato and Aristotle
conclude the volumes (or their sections), all of which we can be very thankful
for, as they provide access for the first time to much which the text offers that is
valuable regarding language and content.
The final verdict to be pronounced on the whole Greek commentator corpus
must be that a model edition has been produced, which should be greeted most
enthusiastically and gratefully as such and for the influence which it is bound to
exercise on the art of editing and on philological methods.5
Yet we must welcome it not only for this reason. This collection is also of the
greatest importance because of the authors represented here. It is not yet the
moment to measure the profit which various fields of study will derive from it,
since its exploitation has hardly begun and a long time will pass before it reaches
a final conclusion. There is, however, much of a general nature that can be
recognised even now. From the ore of precious Presocratic fragments, which is
currently minted in Diels’ edition of the Presocratics, one can foresee what the
history of Ancient Greek philosophy has yet to gain from fragments of

4
Gratifyingly, Ernst Diehl has followed a similiar procedure in his edition of Proclus’ commentary
on Plato’s Timaeus, which was suggested by Diels and supported by the Berlin Academy.
5
[New editions are, however, being prepared of Simplicius in Phys. by L. Tarán, and of Simplicius in
Cael. by P. Hoffmann. – Ed.]
 Review of the Commentaria in Aristotelem Graeca 39

Theophrastus, Eudemus, Epicurus, the Stoics, etc. The abundant advantages

which Aristotelian textual criticism also gains from the commentators does not
concern us here. For this and several other points, Usener’s thorough review
(mentioned above) should be consulted.6 It is our concern here only to emphasise
what the new corpus offers to Byzantine studies, which carry off the lion’s share,
so to speak. If one places the beginning of the Byzantine period of Greek
literature in the time of Constantine I (as has been argued convincingly by K.
Krumbacher),7 all of the authors included here – excluding Aspasius, Alexander
of Aphrodisias and Porphyry – belong to the Byzantine period, and, of the three
just named, Porphyry influenced the Byzantine school for centuries with his
introductory logical work, thus forging, through his being frequently commented
upon and excerpted, the first link in a long chain of Byzantine publications. Our
corpus also immediately confirms Krumbacher’s division of epochs rather than
the one previously customary, which recognises the year 529 as the border
between antiquity and the Byzantine Middle Ages. Porphyry does not in fact
belong to the Constantinian period, but his life extends quite close to it, and with
him a new direction begins. A deep chasm separates him from Alexander of
Aphrodisias, something without analogy in the sixth century. Alexander’s
interest still lies in puzzles, and, although he wishes only to be an interpreter of
Aristotle, he not infrequently arrives at perceptions of his own which deviate
from Aristotle. For Porphyry, in contrast,8 it is merely a matter of making
accessible to the understanding of a wider circle what was recognised as logic
par excellence, that is, Aristotelian logic, and above all else of making it useful for
school – hence his short, introductory handbook on the Five Terms (pente
phônai); hence the new pedagogical experiment of his Categories commentary
in the form of a catechetical dialogue. This pedagogical interest in the
development of new methods for making Aristotle accessible persisted. Dexippus
uses a different form of dialogue-exegesis, while Themistius in his [use of]
paraphrase brings an old form of explanation to a new level of respectability.9
Something of the spirit of Scholasticism already breathes in these works:
Aristotelianism has grown stiff, but new life stirs in the manner in which it is

6
See n. 1.
7
Gesch. d. byz. Litt.2, 2.
8
[For warnings about Porphyry’s supposed neutrality in the logical commentaries, see now P. Hadot
in Chapter 6. De Anima commentaries are, of course, less neutral still, see Henry J. Blumenthal,
Chapter 13 and I. Hadot, Le problème du néoplatonisme Alexandrin: Hiéroclès et Simplicius, Paris
1978, appendix 193–202. – Ed.]
9
This methodological effort issues in a late second blossoming in Sophonias. Cf. the beginning of his
paraphrase of the de Anima.
40 Aristotle Transformed

handed on. The combination of Aristotelian logic with the Neoplatonic system,
however, is entirely scholastic, and though first championed by Porphyry, it
endures from then on, finding its perpetuation in the linkage of Aristotelianism
with Christian dogma. Of course, this combination was not suddenly engendered
by Porphyry without preparation from those who came before. Middle Platonism
of the second century ad, as Albinus demonstrates, had already incorporated
Aristotelian logic into its system, but in such a way that it was attributed to Plato.
From the fact that Plato on occasion made use of this or that form of argument,
it was taken as proof that he had also established this form theoretically; and in
this manner they succeeded, beginning with these few points, in claiming
essentially all of Aristotelian logic as Plato’s. It is with this that Porphyry broke.
Now Aristotelian logic as such, and not covered in Platonic colours, formed the
basis for the study of philosophy in the schools of the Neoplatonists as well, and
thus the exegesis of Aristotle also came to hold an extremely important place in
the Platonic school, which it never lost.
Thus we find traces in our commentaries of an important change that was
complete by the third century ad and causes us to mark a break in the
development of commentary activity before Porphyry. In contrast, the closing of
the Athenian school in the sixth century left no trace at all in the commentaries.
On the contrary, the works of this century are inseparably connected to one
another, and nowhere does there split even the most insignificant seam which
could be used for the division of two epochs. In particular, the opposition
between pagan and Christian is not at all prominent and forms no boundary.
Within the school of Ammonius Hermeiou, Ammonius, Asclepius and Simplicius
are pagan; no definite conclusion can be reached concerning Olympiodorus; and
Philoponus is a Christian. Yet their commentaries give no indication of it.10 If the
Categories commentaries of Simplicius, Olympiodorus and Philoponus are
compared with each other, the stamp of the same school is recognisable even
down to details. The same holds true for the next generation of Olympiodorus’
students, Elias and David. Their names suggest a Christian confession of faith.
Their work remains wholly within the tradition of Ammonius’ school.

10
[For Olympiodorus’ paganism, see now L.G. Westerink, pp. 331ff, and for pagan and Christian
controversy in the commentaries of Simplicius and Philoponus see K. Verrycken, Chapter 10 and P.
Hoffmann, ‘Simplicius’ polemics’, in Richard Sorabji (ed.), Philoponus and the Rejection of
Aristotelian Science, London and Ithaca NY 1987. For David’s Christianising tendency, see Haig
Khatchadourian, ‘Universals in David, Boethius and al-Farabi’s summary of Porphry’s Isagoge’, in
Avedis K. Sanjian (ed.), David Anhaghtc, the ‘Invincible Philosopher’, Atlanta, Georgia 1986. It is not
clear whose pupil David was. – Ed.]
 Review of the Commentaria in Aristotelem Graeca 41

On yet another point, the commentator corpus reflects a fundamental

characteristic of Byzantine literature. If we disregard those pieces which cannot
be attributed to a specific author or precisely dated, the entire collection falls
chronologically into two groups which are very unequal in extent. The first
group consists of the following: from the Peripatos, Aspasius (in the second
quarter of the second century ad ;11 CAG XIX 1), Alexander of Aphrodisias
(who was Professor of Peripatetic Philosophy in Athens between 198 and 211;
CAG I, II , III ), and Themistius (from the second decade until the end of the
fourth century ad ; CAG V); from the Neoplatonic school, Porphyry (232 or 233
until 301; CAG IV 1), Dexippus (about the middle of the fourth century; CAG
IV 2), Syrianus (who became the Diadochus of the Platonic school in 431/2;
CAG VI 1), Ammonius (a student of Proclus, who died in 485; CAG IV 3–6)
and his school: Asclepius (CAG VI 2), Simplicius (CAG VII -XI ), Olympiodorus
(CAG XII ), John Philoponus (CAG XIII -XVII ), and Elias and David (CAG
XVIII ). Somewhat later still than David is Stephanus, whom the Emperor
Heraclius called to Constantinople soon after 610.12 We thus have an
uninterrupted sequence of literary publications from the second century until
into the seventh, and between the lifetimes of the men named above there is
hardly the space of a generation which, here or there, is not covered. Yet with
Stephanus the thread breaks off suddenly, to be spun again only centuries later.
The great hole, which gapes throughout Byzantine intellectual life from the
middle of the seventh century onwards, also makes itself noticeable in the
commentary literature. Only here it extends even further than it does in other
areas. In the ninth century, when new life stirred elsewhere, here there reigned a
perfect silence – and for long afterwards. It revives only in Psellos’ period. The
date of Michael of Ephesus (CAG II 3, XIV 3, XX , XXII ; he is also the pseudo-
Alexander of in Metaph. CAG I; cf. Gött. gel. Anz. 1906, pp. 882ff ) cannot be
determined with certainty, to be sure, but in all probability he belongs to the
eleventh century,13 and only with difficulty to an earlier period. He is also
probably the contemporary of Eustratius (CAG XX , XXI 1).14 Aside from these
two men, the Byzantine renaissance is also represented in our corpus by

11
The dating is taken from Zeller here and in what follows.
12
cf. H. Usener, de Stephano Alexandrino, Bonner Lektionskatal. Summer 1879, 5. [Whether Stephanus
is later than David is questioned by Mossman Roueché: see Chapter 1 – Ed.]
13
cf. Gött. gel. Anz., 1906, 902ff; Arist., Pol., O. Immisch (ed.), Leipzig 1909, xviif. [But see now Robert
Browning, p. 399, for the twelfth-century date – Ed.]
14
cf. K. Krumbacher, Gesch. d. byz. Litt.2, 430.
42 Aristotle Transformed

Sophonias (at the turn of the thirteenth and fourteenth centuries;15 CAG V 6,
XXIII 1).16
How will Byzantine studies profit from the collection of the commentators on
Aristotle?
First of all, material will become available which is invaluable for the history
of the Greek language, for the lexicon as well as for grammar. Without exception,
every author included in the collection has in general become accessible to
linguistic research for the first time in these editions, which have been neatly and
conscientiously prepared on the basis of a careful survey of the tradition. An
abundance of grammatically significant writings can now be traced through the
centuries in almost uninterrupted continuity. Peculiarities of the later language
have long been known from more or less secure examples, but no judgment was
possible concerning their currency and significance, as, for example, with the
absence of an with the optative and in the apodosis of counterfactual hypothetical
sentences, hotan and ean with the indicative and the optative, hote, hênika, heôs,
hou, prin with the subjunctive, the confusion in the use of negatives (oute for
oude, etc.), and peculiar uses and placements of de. All of these can now be
observed within a larger context. It is clear that textual criticism of Byzantine
writers must cull diverse advantages from the insights to be gained here. Literary
criticism does not leave empty-handed either, as is demonstrated by the question
concerning the pseudo-Alexander of the in Metaph. and his relation to Syrianus;
with the help of the linguistic indices, it allows of a definite solution. It can only
be hoped that these commentaries will soon be exploited for their grammatical
significance in earnest. Obviously, it is not merely a matter of establishing the
specific deviations from classical grammar; it should also be considered how
individual authors place the forms accepted by classical grammar next to one
another and opposite the changes. Another interesting problem would be the
relationship between contemporary writers in other areas and these men, who,
on the one hand, are led by their subject and tradition to a linguistic conservatism,
and yet, on the other hand, are led to a certain vital sense of the language through
the oral lecture style.
The commentator corpus will greatly advance work on the philosophical
lexicon. Diels’ work on elementum and stoikheion (‘element’), the paradigm of a
history of philosophical terms, still stands entirely alone. Further investigations

15
ibid.
16
The name Heliodorus, which appears as the author of a commentary on the Ethics (XIX 2), rests on
a falsification. Cf. L. Cohn, Berl. philol. Woch., 1889, col. 1419.
 Review of the Commentaria in Aristotelem Graeca 43

are without question needed for the preparation of the Greek Thesaurus. For
a delimited, yet still quite broad area, the Aristotle commentaries provide
abundant material. A lexicon of philosophical terms built on this foundation
could form the basis for further investigations into the philosophical terminology
of late antiquity and the Middle Ages, and would prove especially useful for
theology.
If we proceed from language to literary form, then the Berlin Academy’s
undertaking valuably enriches our knowledge here as well. The Aristotle
commentaries are in large part transcripts from oral lectures and thus were not
originally conceived as written works, but became such only subsequently. Owing
to their origin, such works exhibit certain peculiarities that have occupied
philology repeatedly in recent years,17 since the interest shown in Cynic-Stoic
diatribe necessarily leads to it. In particular, H. von Arnim’s thorough investigation
of the works of Dio Chrysostom was of as great consequence to his diatribes as to
his sophistic speeches.18 New light was shed especially on the duplicates in Dio’s
work, which von Arnim attributed to the repetition of the same lecture, thus
connecting them with the peculiar origin of this literature. In the same way, the
peculiarities of the composition, language and style of the oral lectures which
became written works through transcription have also begun to be traced in the
area of Christian sermon literature.19 The Academy’s commentary collection
draws our attention to a new species of this genre, namely, the lecture transcribed
from the lectern.20 Generally, in a version of these notes produced with the aim of
publication, the tangible traces of the oral lecture have been effaced, such as any
reference to the audience as companions (hetairoi) or the use of address in the
second person plural, while changes which indicate written composition, such as
‘we write (graphomen)’ and the like, are introduced. Nevertheless, the original
state of affairs is frequently betrayed in the occasional overlooked passage:21
numerous items in the collection already confirm in their titles that the viva vox
of a lecturer is recorded here with the phrase ‘from the mouth of so-and-so (apo
phônês tou deinos)’.22 Next to the name of the professor, the name of the audience
member himself may appear in the heading as well, as in the transcription which
Asclepius made from Ammonius’ series of lectures on the Metaphysics. Indeed, it

17
References to the literature can be found in P. Wendland, Gött. gel. Anz., 1901, 780, n. 1.
18
Leben und Werke des Dio von Prusa, Berlin 1898, 171ff, 282ff.
19
P. Wendland, Gött. gel. Anz., 1901, 780ff.
20
cf. also Freudenthal, Hellenistische Studien III , 303 on lecture notes in antiquity.
21
cf. Gött. gel. Anz., 1905, 518f; ibid. 1906, 903.
22
[See M. Richard, ‘Apo phônês’, Byzantion 20, 1950, 191–222. – Ed.]
44 Aristotle Transformed

could also happen that the student disseminated his transcription solely in his
own name, without thereby committing a transgression in the eyes of his teacher.
When the not yet twenty-year-old Proclus read Plato’s Phaedo with the elderly
Plutarch (of Athens), he was encouraged by Plutarch to write down the exegesis,
and egged on by the remark that then there would be a Phaedo commentary in
circulation under Proclus’ name too.23 By this indifference to intellectual property,
we can explain a discrepancy between David’s and Elias’ introductions to
philosophy. In response to Callimachus’ epigram (Wilamowitz) 23, David (CAG
XVIII 2, p. 31,34) cites two verses under Olympiodorus’ name (‘to which the
philosopher Olympiodorus said [pros touto de eipen Olumpiodôros ho
philosophos]’), which in Elias (CAG XVIII 1, p. 14,8) are ascribed to the speaker
himself (‘I would say to him the contrary, in the following words [egô de ta’nantia
ekeinôi phêmi epê houtôs]’). Although according to the title24 the commentary
appeared as his own, Elias has made a literal transcription here and adopted the ‘I
(egô)’ of his teacher, while David has substituted ‘Olympiodorus the philosopher’.
It is clear that in this branch of transcribed literature, namely, in the exegetical
lectures of the philosophical schools, duplicates must have appeared very
frequently. Few professors caused as much confusion as the Alexandrian Hierocles,
whose lecture series on the Gorgias was transcribed twice by Theosebios, with the
result that when the versions were compared with each other, it was felt that, apart
from the fundamentally Platonic character of the exegesis, nothing was the
same.25 In any event, what usually occurred was that the same notes formed the

23
Marinus, Vita Procli sec. 12. The inclusion of both names in the title, as in Asclepius’ case, is of
course not entirely impossible here either. According to Porphyry’s Vita Plotini sec. 3 (at the end),
Amelius dedicated his transcript of Plotinus’ lectures to Ustillianus Hesychius, thus laying claim to
it as his own literary property. Whether Plotinus was mentioned as the originator, we do not know.
In the Phaedrus commentary that is extant under the name of Hermeias (ed. P. Couvreur, Paris
1901), the discussion is interrupted at 92,6 by the report of an objection raised by ‘our companion
Proclus (ho hetairos Proklos)’. Then it is recounted what ‘the philosopher (ho philosophos)’ responded
to this objection. This ‘philosopher’ can only be Syrianus, with whom Proclus and Hermeias studied
together. The entire section interrupted by the objection belongs to Syrianus, and since this does not
in any way stand out from the rest of the commentary, it is likely the whole commentary belongs to
Syrianus as well. Objections and refutations are brought forward in identical circumstances,
154,21ff, 28ff.
24
On the title, cf. A. Busse, praef., vff. If the form of the title accepted by Busse on p. vii (‘Introduction,
by the grace of God, to Porphyry’s Isagoge from the mouth of the philosopher Helios (Prolegomena
sun theôi tês Porphuriou eisagôgês apo phônês Heliou philosophou)’) were correct, then it would
indeed be possible to suppose an analogous title (with ‘from the mouth of (apo phônês)’) for the
preceding ‘Introduction to Philosophy (Prolegomena tês philosophias)’. Then one would have to
assume that Elias read his literal transcription from Olympiodorus to the listeners of his own
lecture series, which is very unlikely. The ‘from the mouth of (apo phônês)’ of the Laurentianus MS
could easily have arisen, however, from the effect of the preceding ‘from the mouth of David (apo
phônês Dabid)’. The tradition that Busse rests his text on does not recognise it.
25
Phot. Bibl. 338b35ff. B.
 Review of the Commentaria in Aristotelem Graeca 45

basis for the lecture’s repetition, at which point changes of detail could be
transcribed into the text or on inserted leaves, or even given merely oral expression.
At this point, one can imagine that a series of lectures might be transcribed not
only from different repetitions, but also copied down on each occasion by several
listeners, of whom one may have perhaps followed the lecture word for word,
while others paraphrased it in different ways, combining their transcriptions later
before publication. One can also further imagine that the teacher’s expanded and
revised notes may have been at the disposal of some of the students, and finally
that the activity of the listener himself also made itself noticeable in additions,
elaborations, transpositions, and the like. In this way, much of what strikes the
reader of the commentaries can be explained: contradictions (apart from those
which have their origin in the impulsiveness of an oral lecture, or, as with Proclus’
commentary on Plato’s Timaeus, in the stirring, colourful manner of the
commentator26); appending the same discussion in one transcript to one point,
and in another transcript to another point of the presentation;27 the greater
completeness of the commentary of a later generation in contrast with that of an
earlier one;28 and above all the duplicates, which abound in many of these Aristotle
commentaries. It goes without saying that athetising in order to eliminate such
unevenness is completely inadmissable in these circumstances. Our task rather
can only be to establish all the publications for each work in detail, to examine
their special peculiarities case by case, and to trace the ramifications of this for
each commentary, for each commentary series (such as the succession of
Categories commentaries in Ammonius’ school), and ultimately for the entire
literary genre (which raises questions of language and the form of presentation).
If Olympiodorus can, after discussing the views of the Peripatetics and Stoics on
the place of logic in Plato’s philosophy, say ‘In my eyes you are both right’ (Proleg.
17,20),29 then here the freedom and informality of oral speech are clearly revealed.
Likewise, Michael of Ephesus in Nic. Eth. CAG XX , p. 598,19ff.30

26
cf. Gött. gel. Anz., 1905, 533. As regards contradictions, see also 1908, 234.
27
cf. ibid. 1908, 228f.
28
cf. ibid. 1904, 377ff.
29
‘The arguments of the Peripatetics and the Stoics have this character, while the divine Plato thinks
it (sc. logic) a part and a tool (sc. of philosophy), from which [it follows that] there is no need of
preparation, “for as far as I am concerned,” he says, “you both win”. For you both get a rich supply of
proofs, just because etc. (Hoi men oun tôn Peripatêtikôn kai tôn Stôkôn logoi touton ekhousi ton
tropon. Ho de theios Platôn kai meros autên (sc. tên logikên) oietai einai kai organon (sc. tês
philosophias), hothen ou deitai kataskeuês· amphoteroi gar, phêsin, emoi nikate· dia touto gar
epikheirêmata euporeite hekateroi, dioti k.t.l.)’.
30
Here the audacity is admittedly diminished by an addition, which I have italicised: ‘He (sc. Aristotle)
says, once the omissions and that which must be supplied from elsewhere are brought together, that in
accordance with the assumptions of the Epicurean and later Stoic philosophers concerning
46 Aristotle Transformed

The close relation between the commentaries and the oral lecture brings us to
another point: the commentator corpus furthers our knowledge of philosophical,
and especially exegetical, teaching. However pressing its need, a history of
philosophical instruction still eludes us. Such a history alone could teach us to
understand how ancient philosophy was able to fulfil the vast cultural mission
which befell it in antiquity as sovereign in the realm of Weltanschauung, and in
the Middle Ages as the ‘handmaiden of theology’. Ivo Bruns’ monograph on the
activity of Epictetus’ school31 demonstrates how a great deal of valuable material
still lies fallow even with much-read authors. As philosophical instruction after
the second century ad was essentially exegetical, the commentaries immediately
come to the centre of our attention. To elucidate their method, and thus the
teaching method of the philosophical schools, is a pressing task. Of course, if its
method is to be properly understood, the philosophical commentary should not
be isolated from the context of commentary literature as a whole. Nevertheless,
the investigation might well begin with the philosophical commentary, since
massive amounts lie before us, thus offering a firm foundation for research. At
the forefront stands the Aristotle commentary. For study began with Aristotle,
with the aim of progressing towards Plato.32 Among Aristotle’s works, the logical
ones were put first, and among these again, the Categories.33 Still before these,
Porphyry’s Isagoge served as the classic preparation for the Categories. Its exegesis

happiness, one can attribute a share of happiness even to the non-rational animals, while according
to myself and Plato and others who along with us would place happiness in the intellective life, it is
impossible for the non-rational animals to be happy in that way, etc. (legei de (sc. ho Aristotelês) hôs
sullexamenous ta paraleleimmena kai hôn prosüpakouein exôthen khrê, hoti kata men tas tôn allôn
philosophôn Epikoureiôn te kai tôn husteron Stoikôn peri eudaimonias hupolêpseis dunatai tis
eudaimonian metadidonai kai tois alogois zôiois kat’ eme de kai Platôna kai tous allous hosoi tên
eudaimonian en noerai zôêi histômen, adunaton kata tautên eudaimonein ta aloga tôn k.t.l.)’.
31
de Schola Epicteti, Kiel 1897.
32
Marinus, Vita Procli sec. 3: ‘at least, in less than two full years, Syrianus read through all of Aristotle’s
philosophy together with Proclus . . . and having been sufficiently led through these as through a
kind of preliminary initiation and lesser mysteries, he began to lead him into the initiation of Plato
(en etesi goun oute duo holois pasas autôi (sc. tôi Proklôi) tas Aristotelous sunanegnô (sc. ho Surianos)
pragmateias . . . akhthenta de dia toutôn hikanôs hôsper dia tinôn proteleiôn kai mikrôn mustêriôn eis
tên Platônos êge mustagôgian)’. Cf. also the sequence in sec. 12: ‘So, he read Aristotle’s de Anima and
Plato’s Phaedo with Plutarch (anaginôskei oun para toutôi (sc. tôi Ploutarkhôi) Aristotelous men ta
peri psukhês, Platônos de ton Phaidôna)’. Elias, in Cat. 123,9: ‘It is necessary for him (sc. the
interpreter of Aristotle) to know everything of Plato’s so that he may demonstrate Plato’s self-
consistency, making Aristotle’s philosophy an introduction to Plato’s (dei auton (sc. ton Aristotelous
exêgêtên) panta eidenai ta Platônos hina sumphônon heautôi ton Platôna apodeixêi ta Aristotelous
tôn Platônos eisagôgên poioumenos)’. For the later period, cf. perhaps the Byzantine school dialogue,
Byz. Zeitschr. II , 1893, 99, l. 79; Theodor. Prodr. Speech on Isaac Komn., Byz. Zeitschr. XVI , 1907,
115, ll. 118, 123 (previously, ‘the divine Plato (Platôn ho theios)’ represented the highest level of
philosophy). Treatise on the philosopher Joseph, Byz. Zeitschr. VIII , 1899, 11, ll. 15, 19.
33
Ammonius, in Cat. 6,2ff; Simplicius, in Cat. 6,2f; Philoponus, in Cat. 5,15ff; Olympiodorus, in Cat.
8,29ff; Elias, in Cat. 117,17ff (in the last three, there are also several views which deviate somewhat
from those of the former).
 Review of the Commentaria in Aristotelem Graeca 47

initiated the course.34 From here it was thus a short step to preface this series of
lectures with a general introduction, containing discussions on the concept and
division of philosophy and similar questions, as was done by Ammonius (CAG
IV 3), Elias (CAG XVIII 1) and David (CAG XVIII 2). It is surely no accident
that, among the extant commentaries, only those to Porphyry’s Isagoge provide
such general prolegomena, while those on Aristotle and Plato contain
introductions specifically relating only to Aristotle and Plato and their works.
Ammonius provides the simplest of these general introductions. It has become
so elaborate in Elias and David, his pupils at one remove, that it has developed
into a special segment of the lecture series with its own title, though its affiliation
to the exegesis of Porphyry is assured by references. The complete course
of lectures might, in keeping with modern custom, perhaps be entitled
Explanation of Porphyry’s Isagoge with an Introduction to Philosophy. This
general introduction is matched by a special introduction to the study
of Aristotle at the opening of the Categories lecture series,35 which is divided
into two parts:36 the first concerns the philosophy and writings of Aristotle in
general, the second the Categories in particular. The first comprises ten points,
which are enumerated by Ammonius, Simplicius, Olympiodorus, Philoponus
and Elias, in varying sequence and somewhat different formulations. The

34
Porphyry himself designated his Isagoge [Introduction] as a prerequisite for the study of the
Categories (Isag. 1,3). Its position at the beginning of the entire course of study is clear in Ammonius’
explication of The Five Terms (pente phônai, as Porphyry’s Isagoge was called) 1,2 (cf. also 13,10
introductory lectures [eisagôgikais akoais]); 22,23: ‘This is first in order, for if it leads into Aristotle’s
Categories, which is about the simple terms (phônai) which themselves are the beginning of logic,
it is clear that this is first in order in logic (prôton de esti têi taxei· ei gar eis tas Aristotelous Katêgorias
eisagêi, hai peri tôn haplôn eisi phônôn, hautai de tês logikês eisin arkhai, dêlon hoti prôton esti têi
taxei tês logikês)’. Here too the ancient tradition was preserved in the Middle Ages: Nicephorus
Blemmydes’ Autobiography (Heis.) 2,28: ‘completely educated as regards terms (phônai), categories
(katêgoriai), and on interpretation (hermêneia)’. Cf. further Byz. Zeitschr. II , 1893, 98, l.16
35
cf. the beginning of Philoponus’ commentary to the Categories: ‘Just as when we began the Isagoge
[only Porphyry’s Isagoge could be meant; the plural occurs also at 12,19], we spoke of what pertains
to all philosophy, and then marked off the object of the book then before us, so too as we now begin
before the first (prôtou) Aristotelian treatise, let us speak of what pertains to all Aristotelian
philosophy, and then we will define the object of the present book, the Categories (Kathaper
arkhomenoi tôn Eisagôgôn elegomen ta pros pasan philosophian sunteinonta, epeita aphorizometha
ton skopon tou prokeimenou bibliou, houtô kai nun arkhomenoi pro tou (prôtou?) Aristotelikou
sungrammatos eipômen ta pros pasan philosophian sunteinonta tên Aristotelikên philosophian, eith’
houtôs ton skopon tou prokeimenou tôn Katêgoriôn bibliou diorisômetha)’.
36
Olympiodorus has three parts at 1,7 and accordingly in the exposition as well (3ff ): Aristotelian
philosophy, logic, categories. The second section contains the discussion of the question whether
logic is a part (meros) or a tool (organon) of philosophy. That this is a subsequent elaboration of the
original bi-partite schema becomes quite apparent from the fact that section III immediately picks
up where section I leaves off. The tenth and last point in section I raises the question (1,23 = 12,18):
‘How many and what sorts of things must be said about each treatise beforehand? (posa kai tina dei
hekastou sungrammatos prolegesthai)’. The usual points are under discussion: theme (skopos), utility
(khrêsimon), order (taxis), title (the reason for the title) (epigraphê (hê tês epigraphês aitia)), author
(whether the book is authentic) (sungrapheus (ei gnêsion to biblion)), structure (the division into
48 Aristotle Transformed

differences need not be mentioned here; it is sufficient to elaborate the points

according to Simplicius (CAG VIII 3,20ff; the sequence in Ammonius is palpably
false): (1) ‘In how many and what sorts of ways did the philosophical schools get
their names? (kata posous kai poious tropous tas onomasias eskhon hai kata
philosophian haireseis)’; (2) ‘What is the division of the Aristotelian treatises (so
that it becomes clear where in the sequence we should place the present book)?
(tis hê diairesis tôn Aristotelikôn sungrammatôn (hina kai to prokeimenon hopou
khôrou taxômen genêtai dêlon))’; (3) ‘Where should one begin Aristotle’s
treatises? (pothen arkteon tôn Aristotelous sungrammatôn)’; (4) ‘What is the end
of Aristotle’s philosophy? (ti to telos estin tês Aristotelous philosophias)’; (5) What
leads us to this end? (tina ta agonta hêmas pros to telos)’; (6) ‘What is the style of
the Aristotelian treatises? (ti to eidos tôn Aristotelikôn sungrammatôn)’; (7) ‘Why
did the Philosopher choose unclear diction? (dia ti tên asapheian epetêdeusen ho
philosophos)’; (8) What sort of interpreter is required for such arguments? (poion
dei ton exêgêtên einai tôn toioutôn logôn)’; (9) ‘What sort of audience should be
admitted? (poion dei ton akroatên paralambanesthai)’; (10) ‘How many chapters
should one expect for each Aristotelian treatise, of what sort, and on what
ground? (posa dei prolambanein hekastês Aristotelikês pragmateias kephalaia kai
poia kai dia poian aitian)’.37 In this list, two things ought to surprise us. First, why
should there be a chapter here on the names of the philosophical schools, which
contrasts so oddly with all the other sections? Naturally, it chiefly concerns only
the name (of Aristotle’s school) ‘Peripatetic’. Its explanation has given rise to the
entire chapter, which, incidentally, is very old, judging from its content.38 While
this state of affairs has become obscured in Ammonius (1,4.13ff ) and
Olympiodorus (1,16; 3,8ff ), it is quite evident in Simplicius (3,19) and Elias
(107,8). Philoponus (1,7) also emphasises the Aristotelian point of departure,
although admittedly in a very clumsy manner.39 Elias even considers it necessary

chapters) (diaskeuê (hê eis ta kephalaia diairesis)). Section III begins with the application of these
points to the Categories, and section II intrusively intervenes. In Simplicius (8,9ff; 9,4ff ), the
original sequence is clearly recognisable.
37
According to Elias [in Cat. 107,24–6 – Ed.], this schema dates back to Proclus. Some traces can be
found in Proclus’ commentary on the Timaeus: I 348,15ff: ‘secondly, he differentiated the arguments,
thirdly he prepares the listener in such a way that the subsequent arguments will be accepted
(deuteron ton tropon aphôrisato tôn logôn, triton parakeuazei ton akroatên, hopôs prosêkei tous
mellontas hupodekhesthai logous)’; 355,3f: ‘and such is the form of the arguments and such is the
listener (kai hopoion tôn logôn eidos kai hopoios ho akroatês)’.
38
cf. Diels Dox. Gr., 246.
39
‘Firstly, in how many ways the philosophical schools are named, so that we will know to which
school the Philosopher belonged (Prôton kata posous tropous onomazontai hai tôn philosophôn
haireseis, hina gnômen ek poias haireseôs ên ho philosophos)’.
 Review of the Commentaria in Aristotelem Graeca 49

to justify the extension of the inquiry to the names of all schools.40 In any event,
the first point can be formulated on Elias’ model: ‘Why is Aristotle’s philosophy
named thus, viz. “Peripatetic”? (dia ti houtôs ônomasthê he Aristotelous
philosophia, Peripatetikê)’. With that, however, the matter is not yet closed.
Whoever reads through the list of ten problems will ask himself, ‘Why, alongside
questions which concern Aristotle and Aristotelian studies in general, does there
arise this detail of language, which contrasts so strikingly with the remaining
nine points?’ There must be a hundred more such questions of detail, all with the
same right to be considered. Why were not all of them handled or none at all? I
mention the entire question only because it is an example of how our
commentaries further scholarly research that has been concerned with other
problems in recent years, and in turn how new light is thrown on our
commentaries by such publications. The discovery of Didymus’ commentary on
Demosthenes’ speeches41 has provided an opportunity for discussions that
frequently touch on general questions concerning the method of ancient
commentaries. In this connection, Fr. Leo42 emphasises that the actual
commentary (hupomnêma) was intended to be read together with the text of the
author, that it is only a portion of the book, of the commented edition. Here lies
the solution, I believe, to the puzzle mentioned above. What holds true for the
written commentary (hupomnêma) also holds true for the transcribed oral
commentary. It presupposes that there is in the lecturer’s hands a text of the
writer to be explicated, which, as a part of the school library, is also available to
the audience. The exegesis refers to this text step by step. Obviously, the text also
traditionally contains the author’s biography and a list of his writings as an
introduction. Here, the lecturer only needs to add what seems appropriate to
him. The section on the names of the Peripatetics and of other philosophical
schools is probably such an addition, in fact, to a biographical introduction.
Perhaps the origin of the name ‘Peripatetic’ was not mentioned at all in the
introduction, or perhaps it was mentioned in isolation but the speaker thought
it appropriate to set it in the context of the names of the schools and so to add
this customary chapter here.

40
‘And the first of these ten is, why is Aristotle’s philosophy named thus, viz. “Peripatetic”. But since
this is particular and unphilosophical (for the philosopher is fond of the universal), the general
features must be looked into as for what and how many reasons the philosophical schools are
named (Kai prôton esti tôn deka, dia ti houtôs ônomasthê hê Aristotelous philosophia, Peripatêtikê. all’
epeidê touto merikon kai aphilosophon (philokatholou gar ho philosophos), ta katholou zêteisthô dia
tinas kai posas aitias ônomasthêsan hai kata philosophian haireseis)’.
41
Berliner Klassikertexte, herausg. v. d. Generalverwaltung der Kgl. Museen zu Berlin, Heft I, Berlin
1904.
42
Nachrichten von der Kgl. Gesellsch. d. Wiss. z. Gött., philol.-hist. Klasse 1904, 258.
50 Aristotle Transformed

In the same way, the second shock the ten membered schema delivers is also
softened. The second point concerns the question of how to divide Aristotle’s
writings. Nowhere is there offered what one would immediately expect, namely,
a list of these writings. The individual works, as a rule preceded by ‘as (hôs)’
or ‘such as are (hoia estin)’, rather appear only as examples for the rubrics of
the division (diairesis), in which the individual commentators vary in
comprehensiveness when citing them. Surely the school library’s text of Aristotle
contained a list of writings, but one without an easily recognisable or memorable
principle of organisation (just like the lists which have come down to us from
Diogenes Laertius, the Anonymous Menagii and Ptolemy), and the division
(diairesis) given by the commentators was intended to bring organisation and
perspicuity to the collection.43 The first and second points are thus only
supplements to the Aristotle edition. To combine these two points (especially the
first) with the remaining ones, which are of a completely different nature, into a
unbroken unity through a consecutive numbering was surely foolish. What can
be offered in its defence, however, is the attempt, on the one hand, not to
complicate the arrangement of the introduction further by distinguishing
between matters supplementary to the edition and questions for independent
discussion; and, on the other hand, to establish by the introduction of the same
continuous numbering an easily surveyable correspondence between the
overview of the problem and its detailed treatment (e.g. in Ammonius 1,4ff;
1,13ff; 3,21ff; 5,31ff; Olympiodorus 1,15; 3,8ff; 8,29ff; etc.).
The schema which provides the basis for the special introduction to the
Categories which follows the general introduction to Aristotle also occurs in the
introduction to other works and obviously at some point in time possessed a
general validity.44 It comprises the following points: the aim of the writing in

43
Fr. Littig, Andronikos von Rhodes, II Teil, Erlangen 1894 (Progr.), 16ff, traces this division back to
Andronicus, but his demonstration of this does not seem successful to me. I would not, of course,
offer as a successful counterargument that the Aristotelian treatise de Interpretatione, which was
rejected by Andronicus (see the entry in Littig III 1895, 28ff ), is considered here as genuine, since
this work was taken from the list of writings in use and could have been inserted into Andronicus’
division. On the Arabic sources in which the same principle of division is carried out, see Fr. Littig,
Andronikos von Rhodos, I Teil, Munich 1890 (Progr.), 20ff. On this question, cf. also A. Baumstark,
Aristoteles bei den Syrern vom V.-VIII. Jahr., I, Leipzig 1900, 70ff.
44
cf. Ammonius, in Porph. Isag. 21,6ff: ‘We should also mention the things which the philosophers call
prolegomena or technical preliminaries in every book (dei de hêmas eipein kai ta pros tôn
philosophôn houtô prosagoreuomena prolegomena êtoi protekhnologoumena epi pantos bibliou)’.
These prolegomena can be cited with great frequency in the commentary literature after Ammonius.
Outside the CAG, for example, Olympiodorus, in Gorg. (Jahn) 108ff, in Alcib. (Creuzer) 3ff (3 aim
(skopos), 9 utility (khrêsimon), 10 order (taxis), 11 division into chapters or parts (eis ta kephalaia
êtoi merê diaireseôs)) can be cited. These points also extended to Latin commentary literature, as, for
 Review of the Commentaria in Aristotelem Graeca 51

question (ton skopon),45 its authenticity (to gnêsion), its order in the list of
readings (tên taxin tês anagnôseôs), the reason for its title (tên aitian tês
epigraphês), its division into chapters (tên eis kephalaia diairesin), and the
question as to which part of philosophy (or of the particular field of study in
question) the writing concerned belongs (hupo poion meros anagetai to paron
sungramma). Occasionally, the method of instruction (ho tropos tês didaskalias)
occurs as a further subject of discussion. The order and formulation of this
chapter vary;46 the complete list, even apart from the the method of instruction
(tropos tês didaskalias), is not always considered, and at times it is explicitly
stated that one of the points can be left aside in consideration of special
circumstances, while at other times this restriction intervenes silently.
The material which the Aristotle commentaries provide offers an opportunity
for the investigation of this schema’s history, something which is highly desirable.
Fr. Littig traces its first instantiation to the new founder of Aristotelian studies in
the first century bc , Andronicus of Rhodes.47 What militates against this,
however – at least, if my observations are correct – is the fact that then the
schema would have disappeared from the literature for centuries, only to surface
again in Ammonius’ time.48 Of course, it must be admitted that some of the
points unified in the schema are mentioned several times after Andronicus. The

example, in Boethius, in Isag. (Brandt) 14,17. Thomas Aquinas can also be cited as an example from
a later period, cf. in Aristot. de Anima 2a: ‘One renders one’s hearers well disposed by showing the
usefulness of the knowledge, by first stating the order and division of the treatise (benevolum . . .
reddit ostendendo utilitatem scientiae, docilem praemittendo ordinem et distinctionem tractatus)’.
Further: In Boethii Severini de philosophiae consolatione opus Iohannis Murmellii Ruraemundensis
praelibatio (1514, printed in the Boethius edition, Basel 1570), 898: ‘There are some things to be
studied first with which the minds of students are fittingly prepared for what is to be treated; and
these are Boethius’ life, the title of the work, the quality of the style [by which surely the manner of
instruction (tropos tês didaskalias) is meant], the intention of the writer, the number of books, what
the purpose of the work is, and to what part of philosophy it is to be assigned (. . . quaedam praelibanda
sunt quibus studiosorum animi ad ea quae tractanda sunt non incongrue praeparentur. sunt autem
haec: Boethii vita, titulus operis, styli qualitas, scribentis intentio, numerus librorum, quae sit operis
utilitas et ad quam id philosophiae partem referatur)’.
45
[Skopos, literally ‘aim’, and accordingly Praechter translates it Zweck. But in fact it is used, like
prothesis, for the theme, and this would have been a more helpful rendering. As noted above,
the introductions do ask about the aim or end of Aristotle’s philosophy, but the word there used is
telos. – Ed.]
46
The compilation given in the text above is that of Ammonius, in Porph. Isag. 21,8ff. The manner of
teaching (tropos tês didaskalias) (the phrase is from Plato, Cratylus 435E) is in Ammonius, in Porph.
Isag. 23,18; 26,5; in Cat. 66,6 (cf. 80,15); Philoponus, in Cat. 102,16; Elias, in Porph. Isag. 35,7; 40,8;
in Cat. 129,1; David, in Porph. Isag. 80,14; 93,25; 95,10 (cf. also p. x, 2 lines from the bottom, in the
preface to David’s commentary on the Categories); Anon., in de Interp., schol. in Aristot. (Brandis),
93a11; Proclus, in Resp. II . 8,8f; Anon., Proleg. Phil. Platon. (Herm) 27. 220,27f.
47
Andronikos von Rhodos II . 19f.
48
[But see now Hans B. Gottschalk, in Chapter 3, and I. Hadot, ‘Les introductions aux commentaires
exégétiques chez les auteurs néoplatoniciens et les auteurs chrétiens’, in Michel Tardieu (ed.), Les
règles de l’interprétation, Paris 1978. – Ed.]
52 Aristotle Transformed

question concerning the aim of the writings is very widespread. Andronicus

spoke of the purpose (prothesis) of the Categories, according to Schol. in Aristot.
(Brandis) 81a27; cf. also Alexander, in An. Pr. 8,2ff, 9,5. The aim (skopos) of
Platonic dialogues plays a major role in Iamblichus’ exegesis,49 and his successors,
such as Proclus (and later Olympiodorus) emphasised this point heavily.
Alexander also mentions the reason for the title (aitia tês epigraphês), in An. Pr.
42,23 (cf. also in de Sens. 2,1: ‘Even the title of the book reveals the purpose of the
treatise itself (dêloi de kai hê epigraphê tou bibliou tên prothesin tês kat’ auto
pragmateias)’). Alexander puts purpose (prothesis) and utility (khrêsimon)
together, at in Top. 1,3 (here the use (khrêsimon) was given by Aristotle himself);
cf. also 26,24; 125,4–8. The sequence in which the Platonic writings should be
read is an old problem (Albinus, Isag. 4; Diogenes Laertius 2,62; and so on).
Calcidius combines the writer’s purpose (scriptoris propositum) and the order
(ordinatio) (in Tim. 4; the division into chapters is meant in the last expression,
cf. ch. 7), while Dexippus combines aim (skopos) and title (epigraphê) (in Cat.
5,26–30); both of these men belong to the fourth century. Hierocles combines
aim (skopos), order (taxis) and name (epônumia), in Carm. Aur. (Mullach) 417a.50
I have not, however, encountered the set schema mentioned above anywhere
before Ammonius.51 If further investigations do confirm it to be of later origin,
this schema will be useful for literary criticism.
The inauthentic Hippolytus fragment 7 from the introduction to the Psalms is
ascribed to Origen in the Syrian Ambros., while the other three Greek MSS used
by H. Achelis provide no name. It begins: ‘And this is the aim and utility of the
book; but the reason for its title is this. (Kai houtos men tês bibliou ho skopos kai
to khrêsimon. hê de aitia tês epigraphês hautê)’. If fr. 9 – which according to Barber.
III 59 likewise belongs to Origen and is again left anonymous by the three Greek
MSS – is of the same origin, then the order (taxis) may be added to the three
points (140,17 Ach; what precedes is a discussion of the number of the Psalms,
which could correspond to the division into chapters (eis kephalaia diairesis), but
the technical expression is not used). The combination of the three points (still
more, if four) arouses the most lively suspicion against its being attributed to

49
cf. Gött. gel. Anz., 1905, 526f.
50
‘And this is the object of the epic verses and the order: to present the philosophical character to the
listeners before the other lessons [i.e. the order (taxis) of the poems is that they should be read
before any other reading, and their aim is to represent the philosophical character (kharaktêr
philosophos) to the listeners]; they are called “golden” because of their being, as far as epic poems go,
the best and divine (Kai houtos men ho skopos tôn epôn, kai hê taxis, kharaktêra philosophon pro tôn
allôn anagnôsmatôn ektheinai tois akroatais khrusa de tên epônumian eskhe dia to hôs en epesin
arista einai kai theia)’.
51
In whose time it appears to have been rather generally widespread already; see nn. [37 and] 44.
 Review of the Commentaria in Aristotelem Graeca 53

Origen, since the prolegomena schematism is unlikely, even thus abridged, for
the beginning of the third century. A Latin commentary by Oribasius on the
Hippocratic Aphorisms52 would also be revealed as a forgery because the preface
immediately mentions the aim53 of the writing (‘. . . to this he has directed his
study (. . . eo studium suum dirigit)’, p. 6), then its utility (‘it is useful (utilem vero)’,
p. 7), and its authenticity (p. 7), after which the rest of the usual questions follow:
‘to what part of the medical art is this book to be assigned? (in quam partem artis
medicae liber hic redigatur)’ (p. 8); ‘if you ask in what order the rest is to be read
(caeterum si quo ordine legendus sit quaeras)’ (p. 9); ‘then if one asks of the
method of instruction (ho tropos tês didaskalias) (modus doctrinae si quaeritur)’
(p. 9); ‘But in how many parts is the present book divided? What is its title (aitia
tês epigraphês)? (at in quot partes praesens liber dividitur? quid est aphorismus?)’.
If we proceed from the introduction to the actual commentary, we find – in
Olympiodorus, Elias and David, in the third book of Philoponus’ commentary
on the de Anima53a and in Stephanus – an organisation of the content which is
instructive with respect to commentary methods and teaching. A lesson (praxis)54
is divided, insofar as the material allows, into two parts, a general discussion
(theôria) and an explanation of the text (lexis), of which the first provides an
exegesis of a portion of text as regards its main content and its larger context,
while the second provides a detailed exegesis focusing on individual sentences
and words.55 With the paucity of immediate testimonies concerning the external
conditions of teaching, the question arises as to whether knowledge can perhaps
be gained from the written transcripts of the lessons (praxeis) as to the extent
and duration of the lecture. Was it fixed to a determinate period, corresponding
to our class hour, or in delimiting it was it primarily finishing the subject matter

52
Oribasii medici clarissimi commentaria in Aphorismos Hippocratis hactenus non visa, Ioannis
Guinterij Andernaci doctoris medici industria velut e profundissimis tenebris eruta et nunc primum
in medicinae studiosorum utilitatem aedita. Basileae. Ex officine And. Cratandri. 1535 I must thank
my friend in Bern, S. Singer, for verifying the title of the Berner Stadtbibliothek’s text (no. 133).
53
[See n. 45 above. – Ed.]
53a
[Now commonly, but not with certainty, ascribed to Stephanus. – Ed.]
54
Prattein as a scholastic term means ‘to treat exegetically’: Proleg. Plat. Phil. (Herm.) 26, 219,29f: ‘Of
these [certain Platonic dialogues], it is worthwhile to inquire about the order because everybody
expects these to be treated exegetically also. First then one should explain the Alcibiades (toutôn de
axion esti tên taxin zêtêsai, dioti kai toutous êxiôsan pantes prattesthai. prôton toinun dei ton
Alkibiadên prattein)’. 220,13f: ‘But since some think it worthwhile to give an exegesis of the Laws
and the Republic . . . (epeidê de kai tous Nomous kai tas Politeias prattein axiousi tines)’. When used
absolutely, it means ‘to perform an exegesis’. Praxis as ‘exegesis’ is broader in meaning than the
individual exegetical lecture or its lecture plan; cf. Marinus, Vita Procli 22: ‘For with measureless
love of toil, he used to give five exegeses, and sometimes even more, in the same day, usually writing
about 700 lines (philoponiai gar ametrôi khêsamenos exêgeito tês autês hêmeras pentê, hote de kai
pleious praxeis kai egraphe stikhous ta polla amphi tous heptakosious)’.
55
For further details, cf. Gött. gel. Anz., 1904, 382ff; 1905, 532f; 1906, 898f; 1908, 222ff.
54 Aristotle Transformed

that was decisive? (Although even in this case the duration of time could not go
completely unnoticed, since a seminar (sunousia) could not possibly last an
entire day or longer). The significant difference in the size of individual lessons
(praxeis) within one and the same commentary seems to suggest the latter
alternative.56 Yet the transcript’s asymmetries could be in play here too. It is, for
example, quite natural that there is less to be taken down in a class hour in which
a small number of facts are proved in detail through examples, than in a lecture
which is saturated from start to finish with important and well integrated
material. We must therefore look for more reliable criteria. As such, the following
considerations come to mind. (1) The treatment of a lemma in a general
discussion (theôria) and an exegesis of text always ends together with the full
lesson (praxis), not even occasionally in the middle of one so as to make the
general discussion (theôria) of a new portion of text begin in the middle of
the lesson (praxis) in question. The latter might be thought to be the case if the
lecturer had to fill out a determinate period of time with his lecture. Judging
from the overall character of these lectures, it is difficult to believe that an
agreement between the lecture’s constraints on content and time was worked out
through previous experiment or calculation. (2) There are also never two groups
comprising general discussion (theôria) and lexis interpretation joined together
in a single lesson (praxis), as would be expected occasionally if the lesson (praxis)
had to fill a certain amount of time. (3) The close of a lesson (praxis) is announced
by a reference to the subject matter. Thus Stephanus, in Int. 34,1ff: ‘But this
suffices for the present general discussion. We will learn how he does this [the
solution of a puzzle by Aristotle], with the grace of God, in another general
discussion. With the text clear and all that is in it well discussed here, we shall
bring the present lesson to a close (all’ arkei tauta têi parousêi theôriâi. pôs gar
touto poiei en heterai theôriâi sun theôi mathêsometha. saphous de tês lexeôs ousês
kai pantôn tôn en autêi kalôs theôrêthentôn en toutois tênde tên parousan praxin
katapausômen)’. If all these facts militate against an analogy to our class hour,
then, on the other hand, natural limitations on a lecture suggest certain temporal
boundaries arising from the circumstances themselves, in that very extensive
general discussions (theôriai) can be divided into several lessons (praxeis) and
after a long general discussion (theôria) the interpretation of the text (lexis)
assigned to the following lesson (praxis).57
56
[É. Évrard contrasts Philoponus’ Physics commentary, in which lessons last about an hour, regardless
of the divisions of the subject: L’école d’Olympiodore et la composition du ‘commentaire à la Physique’
de Jean Philopon, Diss. Liège 1957. This commentary provides an earlier use of the format than
those listed by Praechter above. – Ed.]
57
Supporting evidence can be found in Gött. gel. Anz., 1908, 222–4.
 Review of the Commentaria in Aristotelem Graeca 55

These are only a few of the numerous questions concerning commentary

method and the organisation of teaching whose investigation the new
commentators’ texts will encourage. Of others, I will mention only the coexistence
of dialogue-58 and lecture-commentaries, which are likewise a reflection of two
forms of teaching, having their parallel in the seminars and lectures of university
instruction. The dialogue-commentary of the sort found in Porphyry’s in Cat. is
at once related to the catechising also represented in other areas of ancient and
Byzantine literature59 and to the oral examination.60 The different types of
commentaries (as regards content and form) should be recalled as they are
mentioned in the loci classici, Simplicius, in Cat. 1,8ff and Sophonias, in DA l,4ff.
Among them, the paraphrase, as uniquely depicted by Sophonias (1,11ff ), is a
remarkable example, even in this purely technical literature, of the Greeks’
surviving feeling for style, which objects to the disunity and disharmony of the
commentary’s usual form, owing to the alternation from the author’s words in
the lemma to the commentator’s subsequent explanation, and to his tearing the
writer’s text to pieces.61
If we turn to the commentaries’ significance for our understanding of
philosophy in late antiquity and the Byzantine period, then it is to be highly
valued here as well. The development of philosophical ideas in this period was
realised essentially in the form of exegesis. This is a sufficient reason for writers of
the history of philosophy to regard it primarily from this point of view, and to
direct their attention to the different methods and changing principles of exegesis
and the correspondingly varying conception the authors have of Plato and
Aristotle. Proclus’ commentary on Plato’s Timaeus, for example, contains an
abundance of material for the history of exegesis that is still little used; the
Aristotle commentaries are likewise very fertile. When this task is approached
one day in earnest, there will develop a sharper eye for fine distinctions, and the
individual features of commentators who for us still disappear in the great mass
will become recognisable, affording a glance into the development of the exegesis

58
Porphyry’s and Dexippus’ commentaries on the Categories, and Boethius’ first commentary on
Porphyry’s Isagoge.
59
cf. e.g. Cicero, de Partitione Oratoria; Philo, Quaestiones in Genesim and in Exodum; C. Chirius
Fortunatianus, Ars Rhetorica; Bakcheius, Eisagôgê technês mousikês; Manuel Moschopulos,
erôtêmata grammatika and peri skhedôn.
60
Byz. Zeitschr. II , 1893, 97ff.
61
Sophonias, in de An. 1,11ff: ‘But pretending to be Aristotle himself and using his own voice as a
mask in order that the whole may be seen as one and not cut apart, they [namely, the paraphrasts]
have left his own text aside, neither dividing it nor unifying it in their treatises (auton gar hupoduntes
Aristotelên kai tôi tês autaggelias proskhrêsamenon prosôpeiôi, hôs eusunopton kai to pan hen eiê kai
mê diakoptoito, tên men lexin parêkan autên, oute diêirêmenên outh’ hênômenên tois hupomnêmasi
suntaxamenoi)’.
56 Aristotle Transformed

of Aristotle, and thereby into the development of philosophical views as well.

Elsewhere,62 I have attempted to show that we obtain a new image of Iamblichus
as soon as we investigate his principles of exegesis, a depiction which helps us to
understand the powerful inspiration with which this man inflamed his students.
With respect to exegesis, Syrianus, Ammonius, Asclepius and Simplicius are also
fascinating.63 Above all, however, the impressive corpus of Aristotle commentators
will teach us to depart from the one-sidedness which is still predominant in the
writing of the history of Greek philosophy, and which has completely assimilated
the philosophical development from the third century ad onwards to the on-
going formation of the Neoplatonic system.64 Even E. Zeller has been influenced
by this tendency. For example, he discusses Themistius under the Neoplatonists,
even though admittedly none of Neoplatonism’s distinguishing characteristics
are evident in him.65 It is simply that under the prevailing paradigm no other
lodging is to be found. This is the punishment for having held the significance of
exegesis in too little esteem. For the philosophers of this time, it is not at all a
matter of system building, of the quest for truth on the path of independent
speculation. Plato and Aristotle are to be explained – the truth lies resolved in
their works. Thus, the emphasis of Aristotle exegesis lay in itself, even where
influenced by Neoplatonic views. Yet frequently this influence also retreats
entirely, especially in the area of logic, which is neutral by its own nature. This
non-partisan commentary activity, which is independent of the religio-
philosophical tendencies of Neoplatonism, is of immense importance to the
greater historical context. It paves the way for ancient philosophical instruction
in the Christian world. With exegesis conducted in this way, there was little or
nothing to prune away for it to be acceptable to the adherents of the new religion
too. With all the fanatical hatred with which the adherents of the old and new
religions battled, here was a field on which they could in fact meet in peace. Thus,
the philosophical schools were able to become Christian. How peaceful the
completion of this transition may have been is demonstrated by the relation of
Philoponus, David and Elias to Ammonius, Simplicius and Olympiodorus.66
Ammonius’ course on the Categories could be read before Christians as much as
Elias’ course before pagans. In this way the Aristotle commentaries lead us

62
Gött. gel. Anz., 1905, 525ff.
63
ibid. 1903, 526.
64
On this matter, Paul Tannery has made some excellent remarks in his essay, ‘Sur la période finale de
la philosophie grecque’, Revue philosophique XLII , 1896, 266ff.
65
[But cf. Henry J. Blumenthal in Chapter 13. – Ed.]
66
[But for Simplicius, see P. Hoffmann, n. 10 above and for Philoponus, K. Verrycken, Chapter 11. –
Ed.]
 Review of the Commentaria in Aristotelem Graeca 57

unnoticeably to Christian scholasticism. Boethius, the founder of Aristotelian

studies in the Christian West, belongs by nature to our commentators series,
though his language separates him.67 The continuing effect of the Greek
commentators on Byzantine intellectual life – which must have been mediated by
teaching during centuries of literary unproductivity as well – can now be traced
with the help of these new complete texts. Interest in Aristotle led to the creation
of new commentaries during the Byzantine renaissance, from which the Academy
edition has brought to life a truly respectable interpreter of Aristotle in Michael
of Ephesus.68 The multifaceted character of his Aristotelian studies is noteworthy.
Only now can that be judged. The new edition offers for the first time a more
reliable foundation for the several commentaries which do not appear in the
tradition as his work (or not incontestably) and could only be attributed to him
on the basis of research into their language and content. Michael commented on
the most varied writings in the Aristotelian corpus, as few before him had done.
From the Organon, he worked on the Prior and Posterior Analytics, the Topics, the
treatise de Interpretatione (cf. CAG IV 5, p. xlv), and the Sophistici Elenchi
(contained in CAG II 3); from the Metaphysics, he commented upon books 7–14
(according to his in Parva Naturalia, CAG XXII 1, p. 149,15), while books 6–14
are extant under Alexander of Aphrodisias’ name in CAG I;69 of the works on
natural science, he commented on the Physics and the de Caelo. He did not halt,
as the rest of the Aristotle interpreters did (with few exceptions),70 before the
anthropological and zoological writings, with their abundance of details from
natural history. It appears that it was a medical and scientific interest, due to
either predisposition or career, which led him to be active exegetically here as
well. His commentary on the Parva Naturalia is extant. According to a list of
Aristotle exegetes contained in a manuscript,71 he commented on the Historia

67
The Wiener Akademie has included him in the plan of the Corpus scriptorum ecclesiasticorum
Latinorum, and in 1883 the edition was commissioned to G. Schepss. Cf. praef., p. v, of the first
volume which has appeared in the meantime: Ancii Manlii Serverini Boethii in Isagogen Porphyrii
commenta. Copiis a Georgio Schepss comparatis suisque usus rec. Samuel Brandt, Vindob. Lips. 1906
(Corp. script. eccl. Lat., vol. 48). This supplements Usener’s note, Gött. gel. Anz., 1892, 1008.
68
For details on Michael and his works, cf. Gött. gel. Anz., 1906, 879ff; O. Immisch (ed.), Pol., xviff.
69
[The ascription of 6–14 to Michael is still controversial. See e.g. Leonardo Tarán, review of Paul
Moraux, Der Aristotelismus vol. 1 in Gnomon. L.6, 1981, 721–50, at 750; P. Thillet in the Budé edition
of Alexander de Fato, p. lvii; Sten Ebbesen, Commentators and Commentaries on Aristotle’s Sophistici
Elenchi, Corpus Latinum Commentariorum in Aristotlem Graecorum 7.3, Leiden 1981, 87. – Ed.]
70
Boethius at least planned a translation and exegesis of all Aristotelian works (according to the
second edition of the commentary on the de Interpretatione (Meiser) 79,16ff ). Georgios Pachymeres
treated the works on natural science: in de Gen. Anim. and in de Part. Anim. are known from the
library catalogue of cod. Paris 2328, according to Cramers Anec Paris. I 393; and Theodorus
Metochites did as well, cf. K. Krumbacher, Gesch. d. byz. Lit.2, 552f.
71
In cod. Marc. 203 (H. Usener, Rhein Mus. XX , 1865, 135f) and cod. Vat. 241 (M. Hayduck, CAG 18.3,
p. v).
58 Aristotle Transformed

Animalium. While nothing is otherwise known of these works, we do nevertheless

have his commentaries on the de Generatione Animalium (CAG XIV 3), the de
Partibus Animalium, the de Motu Animalium and the de Incessu Animalium. A
further work (proposed by Michael at in Parv. Nat. 149,16, but cited as complete
in the manuscript list) dealt with the de Coloribus. Michael’s de Anima was
probably also a commentary. The commentary to books 5, 9 and 10 of the
Nicomachean Ethics are also extant. Finally, Michael gave his attention to two
works, of which the first has rarely and the second (as far as we know) has never
been commented on in the Greek language: the Rhetoric and the Politics. That
Michael was involved with the latter has only recently been ascertained. O.
Immisch, in the appendix to his edition of Aristotle’s Politics, published the
scholia of cod. Hamiltonianus,72 whose author refers (ad loc. 1282a3ff ) to his
exegesis of the treatise de Partibus Animalium. The conclusion which Immisch
drew from this (p. xvi), namely, that we are dealing with Michael of Ephesus here,
can be confirmed. Aside from him, Georgios Pachymeres is the only other
commentator on the de Partibus Animalium we know of.73 The few indeterminate
allusions to contemporary affairs74 are insufficient to draw a conclusion, but the
language is sufficient. A whole series of Michael’s idiosyncracies recur in this
small scholia selection.75 Immisch (p. xviii) has already indicated the cultural-
historical significance of this commentary on the Politics. Aristotle’s Politics was
made accessible to the West in the thirteenth century and formed the foundation
for the development of political theory, which was based on Aristotle for centuries
thereafter. The Byzantine renaissance, as regards interest in this treatise, is ahead
by about two centuries.75a
Michael’s on the whole sober and intelligent exegesis may have furthered
Aristotelian studies considerably in his time and can also provide essential
services to the inexperienced reader of Aristotle today. Above all, he seeks to

72
Aristotelis Politica, post F. Susemihlium recog. Otto Immisch, Lipsiae 1909 (Bibl. Teubn.), 295ff.
73
See n. 70 above.
74
Immisch, xviif.
75
A comparison of the scholia with the compilations in Gött. gel. Anz. 1906, 885ff yields this result.
Substantial agreements of characteristic expressions can be demonstrated. ‘The crows among us
(hoi kath’ hêmôn korakes)’ of 303,34 (Immisch) can be compared to ‘the crows among us (tous kath’
hêmas korakes)’ of in Eth. 587,20. Immisch has already emphasised (p. xviii) that corresponding to
the exegesis of the Politics referred to as being in the future in the commentary on the Ethics, there
is an exegesis of the Ethics cited in our scholia. Besides, an agreement in diction demonstrates that
it is the extant Ethics commentary which is in question: both in Eth. X 578,19 and in Pol. (Immisch)
305,21, in deviating from Aristotle, speak of the happy man who is politically active (politikos
eudaimôn).
75a
[One century, given Browning’s redating of Michael to the twelfth century in Chapter 17. This also
affects the next paragraph. – Ed.]
 Review of the Commentaria in Aristotelem Graeca 59

make Aristotle’s concise expression more intelligible through paraphrase and to

elucidate the context by demonstrating the connection of thought; he also
frequently eases one’s concentration through examples. Yet he also has value for
the experienced reader owing to his preservation of many sorts of valuable
materials, however trivial they were to the author’s overall historical-philological
interests, and owing to the perspective he permits on the philosophical tendencies
of his time. It cannot be doubted that such a literarily productive man, who in
addition – unless everything is deceptive – held a teaching position, was linked
through numerous connections to his contemporaries and their views. It is a
sign of how insufficient our knowledge is of the personalities and intellectual life
in the eleventh century that we are virtually unable to discover a single one of
these connections, and that for us such a man is an almost completely isolated
phenomenon, to whom we cannot form ties anywhere and whose date we can
supposedly determine according to only very general indices. The century of
Michael Psellos, Eustratius, and Ioannes Italos serves as a point of contact for a
more detailed characterisation of the philosophical writer, even when it is a
commentator working largely with an inherited legacy. Yet the material lies
partly unedited in the dust of libraries, partly in editions which are unusable for
scholarly purposes, and the rest offers few returns for the moment in its isolation.
Acquisition kindles new desires: now that Michael of Ephesus has been given
back to us, the more anxious is our longing for the resuscitation of Psellos and
his follower, Ioannes Italos. If those men were joined from the following centuries
by Nicephorus Blemmydes, Georgios Pachymeres, Theodorus Metochites, and
perhaps Leon Magentinos (who was originally accepted in the plan of the
Academy’s enterprise, but then struck from it) together with their philosophical
work, then our knowledge of the philosophy of the later Byzantine period would
be placed on a substantially broader and more secure foundation. If we may be
permitted to hope that the Academy and the directors of the commentary
venture will also expeditiously join in here – as they have often stimulated and
supported the editions of related writers in connection with the corpus of
Aristotle commentators – then they can be certain of the renewed thanks of all
Byzantine specialists.76

76
As a demonstration that the editions of philosophical comentaries can be of benefit to the exegesis
of even quite distant literature, one may be casually directed to Byz. Zeitschr. XIV, 1905, 219, ll. 7ff,
490f (the passage of Elias can be found in the Academy edition, 6,32ff; cf. also 24,21ff ). The
definition of philosophy mentioned in the first passage is one of the six treated in the Isagoge
commentaries of Ammonius, Elias and David.
60 Aristotle Transformed

As an addendum, a few words may be said here as regards the Supplementum

Aristotelicum. It is intended to complete the Aristotle edition as well as the
commentator edition, and it contains, besides the rediscovered Aristotelian Ath.
Pol. (III 2), a number of writings which, without being in the form of a
commentary, belong to the testimonies of Aristotelian studies in antiquity and
the Middle Ages. Both parts of the first volume will interest the Byzantine
specialist. The writing published in the first part belongs to the encyclopaedic
literature which Constantinus Porphyrogenitus engendered. It is a work of
zoological excerpts, whose title reports the content and sources as follows:
‘Summary of the investigations of land, air, and sea animals, elaborated for
Constantine the Great King and Emperor. Aristophanes’ [the Alexandrian
Philologist’s] abridgement of Aristotle’s On Animals, arranged by animal and
what is said about them by Aelianus, Timothy [the Philologist from Gaza about
500 ad ] and others (Sullogê tês peri zôiôn historias khersaiôn ptênôn te kai
thalattiôn Kônstantinôi tôi megalôi basilei kai autokratori philoponêtheisa.
Aristophanous tôn Aristotelous peri zôiôn epitomê hupotethentôn hekastôi zôiôi
kai tôn Ailianôi kai Timotheôi kai heterois tisi peri autôn eirêmenôn)’. Among the
excerpts, those from Timothy deserve special attention for their stylistic
character. On this matter, cf. H. Usener, Gött. gel. Anz. 1892, p. 1019, where
he confirms the observance of the law of sentence conclusions discovered by
W. Meyer. The second part of the volume brings us to the period of the Athenian
philosophers’ emigration to the court of the Persian king Chosroes, after the
closing of the school in Athens. Priscian took part in this exodus, and his
Solutiones is a work that is directly linked with this outcome, but at the same
time also deals with the question of the origin of Neoplatonism and makes use
of important material. The writing appears here complete for the first time. Also
Priscian’s Metaphrasis of Theophrastus’ de Sensu et de Imaginatione, which is of
importance primarily for Theophrastus, yet also for Neoplatonism, will become
available for the first time with this edition.
Despite its brevity, this survey of the Academy’s commentary enterprise may
provide some idea of the riches which are here offered to scholarship, while at
the same time suggesting the tasks which it incurs. May their resolution be
vigorously engaged on all fronts. For their great gift, the highest gratitude is
owed to the Academy and all those who worked with the project, and above all
to their most deserving and untiring director.
 3

The earliest Aristotelian commentators*

Hans B. Gottschalk

The tradition of writing commentaries on Aristotle’s works began about the

middle of the first century bc , in a (successful) attempt to revive the Aristotelian
school after a period of eclipse lasting a century and a half. The reasons for
this eclipse are obscure and probably complex, but one factor which seemed
important at the time was the inaccessibility of Aristotle’s most important
philosophical writings, the so-called pragmateiai or ‘school-treatises’. Andronicus
of Rhodes set out to remedy this by preparing a new, critical edition of these
works; he received some help from Tyrannio of Amisus, a well-known grammarian
settled at Rome, who supplied him with copies of Aristotelian manuscripts which
had found their way into the library of a certain Apellicon and had been taken
from Athens to Rome by Sulla after his sack of Athens in 86 bc , but Andronicus
seems to have been fully responsible for the editorial work.1 However, this edition

* This is a reprint, with some additions and revisions, of pp. 1089–112 and 1150–1 of an article
entitled ‘Aristotelian philosophy in the Roman world from the time of Cicero to the end of the
second century ad ’ in Aufstieg und Niedergang der römischen Welt (ANRW ) II 36 pt. 2, 1079–174.
Some of the additional material was first presented in a paper read at the Institute of Classical
Studies, London, in February 1988; I am very grateful to Richard Sorabji for his helpful comments
on this. The most important general works dealing with this period are E. Zeller, Die Philosophie der
Griechen III 14, 1909, 641–53; F. Überweg and K. Praechter, Grundriss der Geschichte der Philosophie
des Altertums, 11th edition, Berlin 1920, 568ff, 12th edition, Berlin 1926, 556ff (the 11th edition
contains very full bibliographies up to the time of the First World War, while that of the 12th is more
selective); P. Moraux, Der Aristotelismus bei den Griechen I-II , Berlin 1973–84 (where no volume
number is given, the references are to vol. I). For logic, see K. Prantl, Geschichte der Logik im
Abendlande I, Leipzig 1855, 582ff and W. and M. Kneale, The Development of Logic, Oxford 1962,
181ff. Further references are given by Moraux passim and in my ANRW article, 1081ff.
1
cf. ANRW 36.2.1083ff. On Andronicus’ edition see F. Littig, Andronicus von Rhodos I, Munich
1890, II -III , Erlangen 1894–5; M. Plezia, De Andronici Rhodii studiis Aristotelicis, Cracow 1946;
P. Moraux, Les listes anciennes des ouvrages d’Aristote, Louvain 1951, 283ff, Aristotelismus I 58ff;
I. Düring, Aristotle in the Ancient Biographical Tradition, Göteborg 1957, 472ff (this work, a
collection of testimonia with commentary, is henceforth abbreviated AB ).

61
62 Aristotle Transformed

was only one part of a larger scheme aimed at making Aristotle’s writings available
in a form which contemporary philosophers would find suitable to their needs
and interests. In one sense it was selective: it included the main philosophical
school-treatises of Aristotle2 and some by Theophrastus, but not, it seems, the
purely factual writings and collections of material or the published ‘exoteric’
works by which he was chiefly known to Cicero and his contemporaries.3 But it
was accompanied by commentaries on some treatises and a catalogue which, in
the tradition of Callimachus’ Pinakes, included not only a detailed enumeration
of Aristotle’s (and Theophrastus’) writings, but a certain amount of biographical
and other material helpful for understanding Aristotle and his work. Our most
important information about the character of the edition comes from a passage
of Porphyry’s Life of Plotinus:

Since Plotinus had entrusted to me the task of arranging and emending his
books . . . I decided first of all not to allow them to remain in a random
chronological order as they had been issued; but following the example of
Apollodorus of Athens and Andronicus the Peripatetic, of whom the first
collected (the works of) the comic writer Epicharmus into ten volumes and the
other grouped the writings of Aristotle and Theophrastus into treatises, bringing
together those on related subjects, in the same way I grouped the fifty-four books
of Plotinus I had into six Enneads.4

The result can be seen by comparing the catalogue of Aristotle’s writings

preserved in some Arabic sources,5 which is derived from that of Andronicus,

2
Excluding the de Interpretatione, which Andronicus thought spurious (Ammon. in Int. 5,28ff, cf.
Philop. in DA. 27,21f) and Metaph. 2. On the collection of Aristotle’s letters he is said to have made
see below, p. 58.
3
cf. Moraux, Arist., 63.
4
Porph. Vit. Plot. 24 = AB T 75g; see also ch. 26 fin, about the summaries he added to his edition.
Jacoby has doubted the existence of Andronicus’ edition, arguing that Porphyry here refers only to
his catalogue (Comm. on FGH 244 F 213–18, p. 795), but Porphyry was using Andronicus as a
precedent for his own edition of Plotinus. Cf. Moraux, Arist., 59, R. Pfeiffer, History of Classical
Scholarship, Oxford 1968, I 264. On Porphyry’s editorial work see now M.-O. Goulet-Cazé, in id.,
L. Brisson, D. O’Brien et al. (eds), Porphyre, La Vie de Plotin, I, Travaux préliminaires (Histoire des
doctrines de l’ant. class. VI ), Paris 1982, 280ff, and D. O’Brien, ibid. 350ff; on the character of his life
in general, H.J. Blumenthal, ‘Marinus’ Life of Proclus: neoplatonist biography’, Byzantion 54, 1984,
471.
5
This catalogue formed part of a Life of Aristotle by a certain Ptolemy-al-Garîb (P. the Stranger or the
Unknown) which until recently was only known from excerpts in later Arabic ˙ compilations (edited
by Düring, AB 185ff ); a complete Arabic version has recently turned up in a manuscript at
Constantinople, cf. I. Düring, ‘Ptolemy’s Vita Aristotelis rediscovered’, in R.B. Palmer, R. Hamerton-
Kelly (eds), Philomathes: Studies . . . in Memory of P. Merlan, The Hague 1971, 264–9, and M. Plezia,
‘De Ptolemaeo Pinacographo’, Eos 63, 1975, 37–42 (Prof. Plezia has kindly informed me that he and
Prof. J. Bielawski hope to publish a complete edition in the next three years or so). In the preface to
this life, published in translation by Düring loc. cit., the writer refers to the work of Andronicus but
states that he did not have it to hand at the time of writing (and presumably had to rely on memory
 The earliest Aristotelian commentators 63

with the Hellenistic catalogue preserved by Diogenes Laertius (5.22ff ). Ptolemy-

al-Garîb includes, as items 29–56, nearly all the school-treatises extant today, with
˙
the same titles and, except for the Poetics and Metaphysics, the same number of
books. Diogenes lists many of the same writings, but usually as short works of
from one to three books, with different titles from those known to us. For example,
instead of our eight-book Physics, Diogenes has a three-book Peri phuseôs
(no. 90), one book entitled Phusikon (no. 91) and one Peri kinêseôs (no. 115);
probably all three were books later incorporated in the Physics. Diogenes’ Êthika
in five books (no. 38) may be our Eudemian Ethics without the three books in
common to it and the Nicomachean Ethics;6 Ptolemy has the Eudemian Ethics,
under that name, in eight books (no. 36). Our Rhetoric in three books is catalogued
by Ptolemy (no. 39), but Diogenes has a Tekhnê rhêtorikê in two books (no. 78)
and the third book separately under the title Peri lexeôs, divided into two books
(no. 87). A particularly striking case is that of the Prior Analytics, which Diogenes
describes as having no fewer than nine books; it must have consisted of quite
short sections which Andronicus collected into the two long books we still have
today.7 Some at least of Theophrastus’ writings were treated in the same way.
A subscription at the end of book 7 of the Historia Plantarum tells us that
Andronicus gave it the place and title it has today, while Hermippus knew it
(presumably together with book 6) by the title, On under-shrubs and grasses.8
This process of consolidation did not originate with Andronicus, but can be
traced to the beginning of the Peripatetic school. Andronicus himself, when
considering the place of the Postpraedicamenta in the Corpus, referred to earlier
attempts to arrange the logical treatises.9 In Diogenes’ catalogue we find the

and notes made earlier). The writer of the Greek original was until recently taken to have been a
Platonist of the fourth century ad ; see Plezia loc cit., A. Dihle, ‘Der Platoniker Ptolemaios’, Hermes 85,
1957, 314ff (=id. Antike und Orient. Gesammelte Aufsätze, V. Pöschl, H. Petersmann (eds), Heidelberg
1984, 9ff ), Moraux, Arist., 60 n. 6. But see now M. Plezia, ‘De Ptolemaei Vita Aristotelis’, in J. Wiesner
(ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet I, Berlin 1985, 1–11; here he argues
persuasively that the author of the Vita was a Peripatetic, not a Platonist, and repeats his earlier proof
that he was not the main source of the extant Neoplatonic ‘Lives’ of Aristotle. The identification with
P. Chennos, a grammarian and mythographer active in the second century ad, has now been
exploded. Cf. below p. 60 and now D. Gutas, ‘The spurious and the authentic in the Arabic Lives of
Aristotle’, in J. Kraye et al. (eds), Pseudo-Aristotle in the Middle Ages, London 1986, 15–36
6
This is the traditional view (Moraux, Listes, 80f), but recently Kenny (The Aristotelian Ethics, Oxford
1978, 39ff ) has suggested that it may be Books 1–5 of the extant EE, while the titles Peri philias
(no. 24) and Peri pathôn orgês (no. 37) represent our books 7(-8) and 6 respectively. On the EN,
which is found in neither catalogue, see below, p. 68.
7
cf. Moraux, Listes, 87f, Arist., 60ff, with more examples.
8
Peri phruganikôn kai poiôdôn; the title is taken from the opening words of Book 6. The subscription
is found in the oldest MS , U, and two later ones; in the Aldine edition it has been transferred to the
beginning of Book 8 (see further Regenbogen, RE Suppl. VII , 1940, 1373f, 1451f, with references).
9
Simpl. in Cat. 379,8ff, cf. Moraux, Arist., 99 n. 12.
64 Aristotle Transformed

Politics already collected into eight books (no. 74), although there can be no
doubt that this work consists of parts written at widely different times, and since
the author of the catalogue seems unsure whether to attribute it to Aristotle or
Theophrastus, it may have been Theophrastus who first brought them together.
Eudemus seems to have done the same for the Physics; in his own Physics he
treated the same topics as Aristotle in much the same order.10 No doubt they
were following lines laid down by Aristotle himself; the cross-references still
found in his works, most if not all of which must go back to him, show that he
meant them to be seen as parts of a comprehensive scheme. But for whatever
reason, the following generations of Peripatetics did not continue the work, and
it was left to Andronicus to complete it.
The justification of what he did was contained in his book about Aristotle’s
writings. Its exact title is not known, but it was a substantial work in at least five
books. Its main feature was a catalogue of Aristotle’s works, genuine and spurious,
which is said to have enumerated no fewer than a thousand books.11 This part of
the work occupied at least three books; Book 3 is quoted as the source of a
discussion of Aristotle’s Physics12 and Book 5 seems to have contained the list of
‘hypomnematic’ writings.13 The logical works will have been dealt with before
the physical, either in Book 2 or the earlier part of Book 3. In addition Andronicus’
book contained a transcript of Aristotle’s will and is quoted as the source of a
pair of letters, almost certainly spurious, said to have passed between Aristotle
and Alexander the Great;14 they could have been part of the collection of ‘Letters
of Aristotle found by Andronicus’ listed as item 96 of Ptolemy’s catalogue.15
The presence of these items suggests that the book may also have included
a biography of Aristotle, or at least a summary of the main facts of his life,

10
See especially Eudemus fr. 98 and Wehrli’s commentary, p. 88f; C. Brandis, ‘Über die Schicksale der
aristotelischen Bücher’, Rh. Mus. 1, 1827, 281ff.
11
Elias in Cat. 113,17ff, cf. 107,11ff = AB T 75p; also the introduction of the Arabic Life, cited above,
n. 5. For discussion of the content of Andronicus’ book see Littig II 18ff, Plezia, Andronicus, 26–35,
Moraux, Arist., 63–94.
12
Simpl. in Phys. 923,7ff.
13
It is referred to under item 97 of Ptolemy’s catalogue. ‘Hypomnematic’ is a portmanteau term used
to denote collections of material and other writings which are neither systematic school-treatises
nor finished literary works; the extant Problems and Mir. Ausc. are examples of this class of writing.
Its earliest extant occurrence in this sense appears to be in the Neoplatonic Introductions to
Aristotle’s philosophy, but it may be earlier. See further P. Moraux, Les Listes Anciennes d’Ouvrages
d’Aristote, Louvain 1951, 114ff, 145ff, and Düring AB 230, 446ff.
14
Will: Vita Aristotelis Marciana 43 (AB p. 105). Letters: see below n. 24 and text.
15
See Plezia, Arist. Epist. Fr. (op. cit. n. 24 below) 18, 84ff. According to the Arabic version of Ptolemy,
the collection was in twenty ‘parts’ or ‘sections’, and Plezia argues that this means that there were
twenty letters, not twenty books of letters.
 The earliest Aristotelian commentators 65

presumably in Book 1.16 While there is no direct evidence for this in the form of
quotations, the parallels favour the inclusion of some biographical data; even if
Porphyry’s Life of Plotinus is not admitted as evidence, some material of this
kind seems to have been included in Callimachus’ Pinakes,17 and the Neoplatonic
Lives of Aristotle contain some very valuable information not preserved
anywhere else, e.g. the chronological discussion taken from Philochorus
(Vit. Marc. 9, 12). Since it is unlikely that the author of their immediate source
studied the Hellenistic historians himself, this must have come from a good
intermediate source, and Andronicus is the only respectable source he is known
to have used.
But the main emphasis was undoubtedly on Aristotle’s writings, their
authenticity, internal structure, the connection between them and their place in
the Aristotelian corpus. Andronicus’ purpose must have been to explain his
grouping of shorter treatises in pragmateiai and the sequence of pragmateiai in
his edition, and also the relationship of these works to the literary dialogues and
the large mass of hypomnematic writings, collections of material and so on,
which were listed in his catalogue but apparently not included in his edition. The
details, unfortunately, are obscure. We are told that in his view the study of
philosophy should begin with logic,18 and it is reasonable to infer that the logical
works, which he first organised as a corpus, stood at the beginning of his edition
and headed the list of pragmateiai in his catalogue, as they do in Ptolemy’s.
Attempts have been made to recover more details from the introductions to some
of the extant commentaries on Aristotle, particularly from the classification of
Aristotle’s writings found at the beginning of certain sixth-century commentaries
on the Categories. But while the questions raised by these commentators and
many of the answers they suggest clearly belong to an older tradition, all we can
really say about Andronicus is that he probably discussed some of the same
topics; we know hardly anything of his treatment or the internal arrangement of
his work.19 We can still see, however, that he employed the techniques of
Hellenistic scholarship. His catalogue included incipits and information about
the length, in lines, of the works he listed, and letters and personal writings were
listed at the end.20 No doubt he compared different manuscripts, although

16
Littig II 18; Plezia, Andronikos 18–26 attempts a detailed reconstruction of this biography. But
Düring, AB p. 422 has denied that Andronicus wrote a Life of Aristotle in any form.
17
Call. fr. 429, 438 Pfeiffer; Pfeiffer, Hist. Cl. Schol. I 129.
18
Philop. in Cat. 5,18ff, Elias in Cat. 117,24 = AB T 75n, 77a (p. 455).
19
See further Düring, AB 444ff, Moraux, Listes, 146ff, id., Arist. 58ff, with references.
20
See item no. 97 in Ptolemy’s catalogue (AB p. 230).
66 Aristotle Transformed

positive evidence is rare.21 But he also used the available indirect evidence, both
internal and external. He condemned the de Interpretatione as spurious because
it contains what appears to be a false reference to the de Anima,22 and in
connection with Books 5 and 6 of the Physics he cited Eudemus’ Physics and
correspondence with Theophrastus as well as the cross-references in Aristotle’s
other writings.23 He also quoted a pair of spurious letters said to have passed
between Aristotle and Alexander the Great, with the intention, seemingly, of
emphasising the importance of the ‘esoteric’ writings comprised by his edition in
comparison with the well-known ‘exoteric’ literary works.24 The emphasis
throughout was on higher rather than lower criticism, and we can see why
Tyrannio left the work to him; only an Aristotelian with a full philosophical
training could have completed it successfully.
Andronicus is said to have been the eleventh Peripatetic scholarch. The
authority for this is Ammonius, writing in the fifth century ad, and unfortunately
the same Ammonius also tells us that Boethus, Andronicus’ pupil, was eleventh
scholarch.25 The discrepancy has led Düring to reject Ammonius’ testimony
altogether, but it is unreasonable to forego the only evidence we have about
Andronicus’ position in life for what may be a mere slip,26 and Ammonius’
information may have come from a good source: he used the Life of Aristotle and
catalogue of his writings compiled by Ptolemy, who for his part used Andronicus’

21
The most definite evidence would be Dexippus in Cat. 21,18ff, on Cat. 1a1, but comparison with
Simplicius’ comments on the same passage (29,29ff ) suggests that Dexippus may have been
mistaken and it was Porphyry who compared manuscripts at this point; cf. Plezia, Andronikos, 7;
Moraux, Arist., 102. A variant at Phys. 202a14 reported by Simpl. in Phys. 440,14 as from Andronicus
may have been due to conjecture or carelessness, cf. Moraux, Arist., 113ff. The other passages
mentioned by Brandis, Über die Reihenfolge der Bücher des Aristotelischen Organons und ihre griech.
Ausleger, Abh. Akad., Berlin 1833, 273 are not concerned with textual criticism.
22
Ammonius in Int. 5,28ff; cf. Anon. in C. Brandis, Scholia gr. in Aristotelem, Berlin 1836, 94a21ff;
Philop. in DA. 27,21ff. The passage in question is de Int. 16a8, but as the later commentators point
out, Andronicus seems to have misunderstood it. Cf. Moraux, ‘La critique d’authenticité chez les
commentateurs d’Aristote’, Mélanges Mansel, Ankara 1974, 274ff; id., Arist., 117ff.
23
See Simpl. in Phys. 923,7ff; Eudemus fr. 6 and 98 Wehrli; AB T 75m; cf. Plezia, Andronikos, 33ff;
Moraux, Arist., 115f.
24
Aulus Gellius 20.5 = AB T 76f; cf. M. Plezia, Aristotelis Epistularum fragmenta cum Testamento,
Warsaw 1961, 42f, 127ff.
25
Ammonius in Int. 5,24 ‘(Andronicus), who was eleventh <head of the School> after Aristotle’; id. in
An. Pr. 31,11 ‘But Boethus, the eleventh <head of the School> after Aristotle’. Ammonius’ statement
about Andronicus is repeated by Elias and others (AB T 75p), but their testimony has no
independent value; cf. Düring, AB , 416ff, 420f. Lynch, 204 claims that the second Ammonius
passage is ‘an obviously apocryphal and harmonising list of diadochoi’ and that Ammonius ‘went on
to list as Boethus’ followers such men as the Neo-Platonists Porphyry and Iamblichus’. It is clear that
he has misunderstood Ammonius’ êkolouthêsen and knows no more of his text than the brief
extract printed by Düring.
26
One way of reconciling both statements would be to suppose that Ammonius included Aristotle in
his count in the first passage and excluded him in the second; cf. Zeller, III 1, 642 n. 5.
 The earliest Aristotelian commentators 67

book.27 So Andronicus, who must have said something about the history of the
Peripatetic school (if only to explain why the school-treatises had been allowed
to fall into oblivion), may have mentioned his scholarchate himself. If Ammonius
and Elias are right to this extent, it follows that he did his work in Athens, and
this would explain why he had to rely on Tyrannio for copies of the Apellicon
manuscripts. Düring’s assertion that he prepared his edition at Rome is baseless.28
A fiction also is the ‘Roman edition’ of Aristotle so beloved by compilers of
handbooks,29 unless we want to dignify the bad commercial copies of Apellicon’s
manuscripts, about which Strabo tells us, with this title.30
However, the very existence of the Peripatos at this time has been called into
question in two recent books, in which it is argued that all the philosophical
schools, with the possible exception of the Epicurean, ceased to exist after Sulla’s
siege of Athens in 86 bc caused the destruction of their buildings and the
dispersal of their members.31 Thereafter the schools continued in an ideological
sense only, with no permanent institutional centre, and their most important
representatives spent their working lives in places other than Athens, especially
in Rome and the great cities of Syria and Asia Minor. Lynch has succeeded in
demonstrating that we have no positive evidence for the existence of the Peripatos
in the imperial age and that the men described as scholarchs in modern accounts
of this period have no identifiable claim to the honour.32 But while it is right to
suspend judgement where no evidence exists, there is some evidence for the first
century bc : not only the report that Andronicus and Boethus were scholarchs,
but also a remark in the Index Academicorum Herculanensis that Ariston and
Cratippus ‘abandoned the Academy and became Peripatetics’.33 This cannot mean

27
cf. Elias in Cat. 107,11ff = AB T 75 p. 3; Ptolemy’s catalogue item 96–7, etc.
28
AB p. 420 and elsewhere. It was swallowed by J.P. Lynch, Aristotle’s School, Berkeley 1972, 203ff,
whose vaunted critical faculty works in a curiously intermittent fashion.
29
It was invented by H. Usener, ‘Ein altes Lehrgebäude der Philologie’, in Kl. Schriften II , Leipzig and
Berlin 1913, 306f (first edition in Sitzungsber. d. philos.-philolog. und hist. Kl. d. Bayer. Ak. d. Wiss.,
1892, 582–648); Kl. Schriften III 150ff, who claims that the main part of the editorial work was done
by Tyrannio, acting as literary adviser to Cicero’s publisher friend Atticus, for publication by Atticus.
For appropriate comment see A.E. Housman in the introduction to his edition of Lucan, pp. xiii–
xviii; also Littig, Andronikos II , 8ff.
30
As is done by Littig, Andronikos I, 12f. Cf. Strabo 13.1.54 p. 609 C = AB T 66b.
31
Lynch, Aristotle’s School, 163ff; J. Glucker, Antiochus and the Late Academy, Hypomnemata 56,
Göttingen 1978, 98ff, 330ff.
32
See ANRW 36.2.1080 n. 5.
33
Index Acad. Herc., col. 35. 10ff, p. 112 Mekler ‘Of whom Ariston and Cratippus left the Academy and
became Peripatetics’. Glucker 99ff discusses this passage at length and tries to alter the accepted
reading in such a way as to support his contention that Antiochus’ ‘Academy’ was something
different from the school founded by Plato. This does not concern us here, but he also says (96, 113)
that Dio of Alexandria, who is named a few lines previously (col. 35 line 8) went over to the
Peripatos with the other two, and this is clearly wrong.
68 Aristotle Transformed

merely that they showed an interest in Peripatetic philosophy or that Ariston

made a name as a commentator on Aristotle, for Eudorus the Academic and
Athenodorus the Stoic did the same without losing their membership of their
respective schools. Moreover, since the school of Antiochus recognised no
fundamental distinction between Academic and Peripatetic beliefs, it would
have been pointless to describe any of its members as apostates for teaching
Peripatetic doctrines. The report only makes sense if there was some organisation
recognised by contemporaries as the legitimate successor of Aristotle’s school
and official representative of his teaching, for Ariston and Cratippus to join.
Undoubtedly the school’s activity must have been interrupted for a time by the
siege of Athens and its aftermath,34 and Andronicus may not have been able to
work in the old buildings near the Lyceum, outside the city-wall. But the life of a
learned institution does not depend on buildings or even administrative
continuity; what matters is that there should be an identifiable body of men
whose members feel that they are the heirs and representatives of an intellectual
tradition and that their claim should be admitted by those contemporaries who
are sufficiently interested to hold any decided view on the matter.35
A more difficult question is the date of Andronicus’ edition. Düring puts it
between 40 and 20 bc ;36 his chief reason for putting it so late is Cicero’s silence,
which would indeed be inexplicable if the work was done in Rome during his
lifetime. But as we have seen, Andronicus almost certainly worked at Athens, and
most of the evidence points to an earlier date. Apellicon’s library came to Rome
about 84 bc , Tyrannio about 71; Sulla’s library seems to have been accessible to
Sulla’s friends, including Cicero, and Tyrannio was highly regarded in their
circle,37 so the rediscovery of Aristotle’s manuscripts can hardly have happened
later than 60 bc . Andronicus must have been considerably older than Strabo, who
was born about 63 bc . Strabo says that he himself was an associate, possibly a
pupil, of Boethus, a pupil of Andronicus,38 and in his enumeration of eminent

34
Perhaps this was what induced Andronicus to devote himself to literary work, much as the forced
closure of the Neoplatonic school at Athens some six centuries later induced Simplicius to write his
invaluable commentaries.
35
cf. H. Weissert, ‘Über Universitätsjubiläen’, Ruperto-Carola 64, 1980, 5–8, who points out that many
universities which claim a history of several centuries have not in fact been continuously active
throughout these periods.
36
Düring, AB p. 421; for earlier proponents of this view see Brink, RE Suppl. VII , 1940, 938.
37
See the passages quoted in AB T 74c.
38
Strabo 16.757 = AB T 75b ‘In my time, famous philosophers born at Sidon are Boethus, with whom
I studied Aristotelian philosophy, and . . .’. Moraux, Arist., 54 argues that sumphilosophein with the
dative always means ‘to be a pupil of ‘, but the available evidence is not sufficient to allow a firm
conclusion. But since Strabo does not refer to anyone else as his teacher, most scholars have
understood him to mean that he was a pupil of Boethus; see Zeller III 1, 646 n. 2; Brink 938.62f;
F. Susemihl, Geschichte der gr. Lit. in der Alexandrinerzeit, Leipzig 1892, 307 n. 354.
 The earliest Aristotelian commentators 69

Rhodians places his name between those of Stratocles, a pupil of Panaetius,

and Leonidas, possibly a pupil of Posidonius;39 he does not name him among
his own contemporaries. All this would tend to make Andronicus roughly
contemporaneous with Cicero, and one piece of evidence points to an upsurge of
activity in the Peripatetic school about 60 bc . Not long after the death of Antiochus
of Ascalon, when his brother Aristus was head of the Academy, two of its members,
Ariston of Alexandria and Cratippus of Pergamum, went over to the Peripatos;
the exact date is unknown, but Antiochus died in 68 bc and Cratippus was an
established Peripatetic teacher when Cicero met him in Ephesus in 51.40 Each
later became eminent in his own way, Ariston as one of the first generation of
commentators on Aristotle, Cratippus as a popular teacher whose pupils included
Cicero’s son Marcus.41 They must have had a good reason for transferring their
allegiance, and it is a plausible guess that what attracted them was the renaissance
of Aristotelian studies which seemed likely to result from Andronicus’ researches;
such a move would have been a natural outcome of Antiochus’ advocacy of a
return to the ‘ancient’ (we might almost say ‘classical’) thinkers of the fourth
century. Eudorus and Athenodorus, who must have written their commentaries
on the Categories after Andronicus’ edition had come out, flourished in the middle
of the first century bc ; the latter was an old man when he returned to his native
city not long after the battle of Actium.42 All this would indicate that Andronicus
began his work in the sixties and published his edition and catalogue, not
necessarily all at once, during the following decades.43

39
Strabo 14.655; cf. Zeller 590n, 606n.
40
Index Acad. Herc. 35.8ff, p. 110ff Mekler. Most of the other evidence has been collected in Mekler’s
notes ad loc.
41
Cic. de Off. 1.1.1, etc; see below.
42
Strabo 14.674. For the chronology of Ariston and Eudoros see P.M. Fraser, Ptolemaic Alexandria,
Oxford 1972, I 488ff (text), II 707ff (notes), who points out that both men must have been
considerably older than Strabo. Unfortunately he is inclined to accept Düring’s late date for
Andronicus (n. 97) and this forces him to suggest that Ariston and Eudoros may have written their
commentaries on the Categories before Andronicus. But this is very unlikely, as we shall see.
43
This would be confirmed if it were certain that Cic. Orator 192 refers to Andronicus’ edition of Ar.
Rhet. 3, 1408b32ff, as Usener claimed (Kl. Schr. II , 306); but since he does not give a title, it is not
certain that the reference is to this book (cf. Düring, ‘Notes on the history of the transmission of
Aristotle’s writings’, Göteborgs Årsskrift 56, 1950, 3, 38f), and even if it is, Cicero might have used an
earlier version, perhaps even one of the bad commercial copies of which Strabo speaks. The earliest
extant writer to quote from a certainly Andronicean text of Aristotle seems to be Dionysius of
Halicarnassus, who cites passages from Rhetoric 3 as standing en têi tritêi biblôi tôn rhêtorikôn
tekhnôn (Comp. c. 25, cf. Epist ad Amm. 1.8). Cf. F.H. Sandbach, ‘Plutarch and Aristotle’, Illinois Class.
Stud. 7, 1982, 229. Recently Moraux (Arist., 57f) has suggested that Andronicus may have begun
even earlier, ‘soon after the death of Apellicon and the confiscation of his library’. This is not
impossible, although Moraux is inclined to discount Plutarch’s story that Andronicus received help
from Tyrannio; he could have started his work before learning of the papers in Rome, and then
asked Tyrannio for copies.
70 Aristotle Transformed

One difficulty remains. It is often said that Cratippus was head of the Peripatos
in the mid-forties, when Cicero’s son Marcus was his pupil at Athens.44 The
evidence is supposed to be a remark of Cicero’s, who had a high regard for
Cratippus and wrote of him as ‘of all the Peripatetics whom I have heard, in my
judgment easily the foremost’.45 But if Cicero meant to convey that he was
scholarch, this expression would be strangely inappropriate, like saying that a
man is ‘easily Vice-Chancellor of his university’, and all the other evidence is
against this supposition.46 When Cicero first met him in 51, he resided at
Mytilene, and he was still there in 48, when he met Pompey fleeing after his
defeat at Pharsalus. By 44 we find him at Athens. Through Cicero’s good offices
he obtained Roman citizenship, but at the same time Cicero persuaded the
Council of the Areopagus to pass a special motion requesting him to remain at
Athens. These facts do not suggest that he had an official position there, but that
he seriously considered moving on to Rome and was with difficulty persuaded to
stay.47 Even at this time, when Marcus Cicero was under his care, his associates
were men he had brought with him from Mytilene; and when Cicero’s friend
C. Trebonius invited Marcus to accompany him to Asia, he proposed to take
Cratippus as well so that Marcus’ studies would not be interrupted.48 These
passages are enough to indicate what his position was: he was head of a private
school of philosophy existing side by side with the official Peripatos, perhaps
with a special following among young Romans finishing their education in
Athens.49 His relations with Andronicus and his followers may have been
strained. Unlike his fellow-convert Ariston, he took no part in the work on
Aristotle’s school-treatises, but seems to have been a Peripatetic of the Hellenistic
type, teaching rhetoric as well as the more popular kind of philosophy.50 On the
other hand Andronicus must have thought his own work much more important
than retailing moral uplift to the gilded youth of Rome. As for Cicero, his

44
e.g. Fraser I, 489; cf. Littig, Andronikos I, 7ff.
45
Cic. Tim. 1; cf. Off. 3.2.5: ‘My <friend> Cratippus, the foremost philosopher in living memory’; Off.
1.1.2: ‘The foremost philosopher of this age’. For Cratippus’ career, see A. O’Brien-Moore, ‘M. Tullius
Cratippus, priest of Rome’, Yale Class. Stud. 8, 1942, 25–8 and Moraux, Arist., 223–56.
46
For the non-specific use of princeps qualified by an adverb cf. Epist. ad fam. 13.78.1.
47
Lynch, 204ff; Moraux, Arist., 223ff; Glucker, Antiochus, 114ff, has a different and to my mind far-
fetched explanation of this episode.
48
Marcus Cicero in Epist. ad Fam. 16.21.5; Trebonius, ibid. 12.16.2.
49
It is tempting to speculate that one motive for his migration to Athens was the hope of becoming
scholarch, but that he was defeated. In that case the man who vacated the scholarchate in 47 or 46
must have been Andronicus, and his successor Boethus. But this is very uncertain.
50
cf. Cicero Brutus 250, Moraux, Arist., 227ff. This need not invalidate the suggestion made above,
that he and Ariston were attracted to the Peripatos by Andronicus’ work on the school-treatises; it
may only mean that Cratippus found he had bitten off more than he could chew.
 The earliest Aristotelian commentators 71

predilection for Cratippus and silence about Andronicus can be explained on

both intellectual and personal grounds. In spite of his strictures on the Hellenistic
Peripatos, the Hellenistic tradition was the one in which he had been brought
up.51 Philosophy for him was a gentlemanly accomplishment, and he knew
Aristotle almost entirely from his elegant dialogues. He would have found the
school-treatises barely comprehensible and Andronicus’ detailed editorial
labours abhorrent. At the same time he was a very grand personage. When he
visited Greece he did not expect to call on philosophers – they were expected to
wait on him. Cratippus made it his business to wait on eminent visitors from
Rome and to tell them what they wanted to hear;52 Andronicus, it seems, did not.

II

Andronicus performed his task well. He not only established the form and canon
of Aristotle’s writings which, with comparatively slight modifications, we still use
today, but initiated a way of doing philosophy which was to predominate among
Aristotelians to the end of antiquity and to spread to the adherents of other
schools. Their work was firmly centred on the Aristotelian writings. Much of it
consisted of straight exegesis and even where they disagreed with Aristotle’s
doctrine or were dealing with different problems from his, they often chose to
present their views as an interpretation or development of his ideas. Philosophical
debate could take the form of an argument over the authenticity or meaning of
a text.53
Andronicus presented Aristotle’s philosophy as a system like those of the
Stoics and Epicureans, which his adherents were expected to understand and
propagate, with such additions as might be needed from time to time. It has been
suggested that this entailed a distortion of Aristotle’s thought.54 But while it is

51
His approval of Cratippus may be evidence that he took his opinion of the Hellenistic Peripatos at
second hand, perhaps from Antiochus.
52
Cic. Tim. 1; Plut. Pomp. 75.
53
cf. R.B. Todd, Alexander of Aphrodisias on Stoic Physics, Philosophia Antiqua 28, Leiden 1976,
2ff, 8ff. The same approach can be found among Platonists; see Dörrie, Platonica Minora, 308,
H.J. Blumenthal, ‘Plotinus in later Neoplatonism’, in H.J. Blumenthal and R.A. Markus (eds),
Neoplatonic and Christian Thought (Essays in honour of P.H. Armstrong), London 1981, 213f, with
references, and D.A. Russell, Plutarch, London 1973, 64. It is criticised by Seneca Epist. 108.23.
54
Düring, AB p. 422f; id., Aristoteles, Heidelberg 1966, 42; id., RE Suppl. XI , 1986, 199, 319ff; his
insistence that Aristotle was a ‘Problemdenker’ seems to be derived from N. Hartmann (whose
historical essays have been collected in vol. II of his Kleinere Schriften, Berlin 1957), but I am not
sure that Düring has understood his meaning fully. Cf. J.L. Ackrill, Aristotle the Philosopher, Oxford
1981, 1ff.
72 Aristotle Transformed

true that many of Aristotle’s books were not originally written for the context in
which Andronicus placed them, and Andronicus ignored and probably had no
means of knowing the dates at which particular books were composed, a
systematic tendency was inherent in Aristotle’s philosophy from the start. This
is not only shown by the elaborate cross-references and the summaries and
synopses by which his writings are articulated, but by his use of a limited number
of well-established principles to solve the problems presented by his investigation
of different areas of reality. In making these tendencies explicit, Andronicus
continued a development begun by Aristotle’s earliest pupils, as we have seen,
and adopted the only way to make Aristotle’s teaching fruitful in the circumstances
of his time.
One result was to throw up a new set of problems concerning the interrelation
of Aristotle’s works. They could take several forms. Real or apparent differences
of doctrine or terminology between various parts of the Corpus had to be
explained and harmonised; this was done in the commentaries written under
Andronicus’ influence, especially those of Boethus.55 But there was also the more
general question of the way in which the parts of philosophy were connected
and the order in which they should be studied. Such questions had long exercised
the Stoics, and their tripartite division of philosophy into logic (to logikon),
natural philosophy (to phusikon) and ethics (to êthikon) was the position from
which the Aristotelians began; they could do so the more readily because
Aristotle occasionally used the same classification.56 In general, however,
Aristotle approached the problem in a different way. His distinction between
theoretical, practical and productive ‘sciences’ was not a division of philosophy
only but of all types of intellectual endeavour.57 Leaving the last aside, the first
two coincide reasonably closely with the Stoics’ physical and ethical branches of
philosophy; but since Aristotle’s subdivision of theoretical science (into First
Philosophy, Natural Philosophy and Mathematics) left no obvious place for
logic, his followers felt free to place it in a separate class, as the Stoics had done.

55
There is some evidence that such harmonisation may sometimes have led to textual interpolation:
see J. Brunschwig, ‘Observations sur les manuscrits parisiens des “Topiques” ’, in G.E.L. Owen (ed.),
Aristotle on Dialectic, Oxford 1968, 16ff (on Top. 120a6–11).
56
Ar. Top. 105b19: ‘Propositions and subjects of discussion are divided into three: some are ethical,
some concerned with natural philosophy, some logical.’ Cf. Alex. Aphr., ad loc., p. 93,20ff. The same
division is attributed to Plato by writers from Cicero (Acad. Post. 1.5.19) to Aristocles of Messene
(fr. 1 Heiland = Euseb. PE 11.3.6; cf. F. Trabucco, ‘Il problema del “De philosophia” di Aristocle di
Messene e la sua dottrina’, Acme 11, 1958, 136f) and to Xenocrates by Sext. Emp. Adv. Math. 7.16 (=
fr.1 Heinze = 82 Isnardi Parente).
57
For Aristotle’s classification of the sciences and the problems it raises see H.H. Joachim’s commentary
on the EN, Oxford 1951, 1ff.
 The earliest Aristotelian commentators 73

The omission was explained by saying that logic was not a part of philosophy
coordinate with the other two, but a tool used in both theoretical and practical
philosophy.58 There is no direct evidence for attributing this argument to
Andronicus, but an indirect link can be established. On the related question of
the order in which Aristotle’s works should be read, Andronicus held that the
logical works should come first, on the ground that neither of the other branches
of philosophy could be learned properly without a grasp of the basic rules of
argument.59 This implies that he held, and probably originated, the view of logic
as a tool. Boethus may have disagreed; he argued that the study of philosophy
should begin with physics because it is closest to everyday experience.60 When he
wrote, it seems, opinion in the school was still fluid, but later Andronicus’ view
became canonical.
As this example shows, Andronicus’ decisions were not always taken as the
last word, and several additions were made to the Aristotelian canon by his
immediate successors. The last six chapters of the Categories, the so-called
Postpraedicamenta, which Andronicus did not regard as belonging to that
treatise, were treated as part of it by most subsequent commentators, including
Boethus;61 the only exception of which we know is the anonymous writer of two
pseudo-Archytean tracts which together repeat the content of the extant
Categories, with minor modifications.62 The de Interpretatione, which Andronicus
declared spurious, was accepted as genuine by all who came after him.63 Nicolaus

58
Alex. Aphr. in Top. 74,26ff; 94,8ff, quoting Arist. 104b2f; Ammon. in Cat. 4,28ff; Simpl. in Cat. 4,22ff;
Olympiod. in Cat. 7,24ff, Philop. in Cat. 4,23ff; Elias in Cat. 115,14ff; Anon. de Arte Logica (in the
preface of Busse’s edition of Olympiod. in Cat.) p. xif. All these writers describe logic as an organon
of philosophy and logical writings as organika, but the use of ‘Organon’ as a collective title for
Aristotle’s logical works seems to have no ancient authority. The summary of Aristotle’s philosophy
given by Diogenes Laertius 5.28ff (see ANRW II , 35.2. 1129) has a different scheme, based on bi-
partition: philosophy is divided into practical and theoretical, practical into ethics and politics, and
theoretical into physics and logic; but the writer adds that logic is a tool, not a part of philosophy in
its own right. Cf. Moraux, ‘L’exposé de la philosophie d’Aristote chez Diogène Laërce V, 28–34’, Rev.
Philos. de Louvain 47, 1949, 7ff.
59
Philop. in Cat. 5,18ff; cf. Elias in Cat. 117,22ff.
60
Philop. in Cat. 5,16f; Moraux, Arist., 143ff.
61
The evidence is indirect. Boethus included these chapters in his commentary on the Categories
(see Simpl. in Cat. 433,28ff, on 15b1), while Andronicus is not quoted on this part of Aristotle’s
work (Simplicius’ reference on p. 385,3f is to his views about the categories ‘Action’ and ‘Passion’,
cf. ibid. 322,15). Cf. Plezia, Andronicus, 42; Moraux, Arist., 99, 163; and id., Mélanges Mansel (above,
n. 22) 276.
62
See ANRW 36.2.1131ff. The fact that ps-Archytas wrote up both sections of the Categories is some
evidence (but by no means conclusive) that Andronicus accepted both as genuinely Aristotelian,
even if he did not believe that they formed a single work. Cf. Moraux, Arist., 99 against Plezia,
Andron., 35; but in Mél. Mansel, 272 Moraux still accepts the older view that Andronicus probably
regarded the Postpraedicamenta as spurious. See also Boethius in Cat. 1, Migne PL 64,162A.
63
Ammon. in Int. 5,28ff, and the parallels cited by Moraux, Arist., 117 n. 3; cf. Prantl, Gesch. d. Logik I,
547; Zeller II 2, 69 n. 1.
74 Aristotle Transformed

of Damascus, active in the reign of Augustus, referred to two books which

Andronicus apparently did not know, Theophrastus’ Metaphysica and the short
essay which today forms the second book (alpha elatton) of Aristotle’s Metaphysics,
but which must have gained its present place after the formation by Andronicus
of the thirteen-book Metaphysics listed in Ptolemy’s catalogue.64 It is not known
how Nicolaus obtained his extra material. He may have found some manuscripts
in the libraries of eastern Mediterranean cities, as Drossaart Lulofs has suggested
(p. 28); his diplomatic activity on behalf of Herod the Great involved much
travelling and may have presented him with good opportunities. But it is also
possible that Andronicus’ work was continued by members of his school after his
death and Nicolaus simply reflects a later stage in its progress. At the level of
textual criticism we hear of Andronicean readings which were not accepted by
the majority of later commentators.65 Nevertheless Andronicus’ edition was the
foundation of all subsequent work, much as I. Bekker’s edition of 1831 has been
the basis of the scholarly activity of more recent times. No less important was his
example. In the ensuing centuries, philosophical discussions among Aristotelians
took place against a background of philological and historical criticism.
Aristotle’s writings were not all studied with equal intensity. The biological
works were almost entirely neglected; the only writer known to have given them
more than a passing mention was Nicolaus of Damascus, who included
compendia of the zoological works and the de Plantis in his summary of
Aristotle’s philosophy,66 and no full commentaries were written on any of them
in antiquity. The Politics received little attention, perhaps because it was too
remote from the political realities of the time. The Rhetoric and Poetics were
ignored by the philosophers; their influence, and that of Theophrastus’ rhetorical
works, has to be traced in the writings of professional rhetoricians.67 The fortuna

64
See the subscription to Theophr. Metaph. = Nic. Damasc. T 7.1 (edited by H.J. Drossaart Lulofs,
Nicolaus Damascenus on the Philosophy of Ar., Philosophia antiqua 13, Leiden 1969) and for Arist.
Metaph. 2, Nicolaus F 21, with Drossaart Lulofs’ comments (p. 30, 137, 139). It is not likely that
Andronicus knew bk. 2 as part of his version of the Metaphysics, as W.W. Jaeger suggests (Studien
zur Entstehungsgesch. d. Metaphysik d. Aristoteles, Berlin 1912, 178); even if the later commentators,
faced with the fait accompli that the books of the Metaphysics were labelled A-N, occasionally
treated bk. 2 as a mere appendage of bk. 1, there would have been no reason for Andronicus to do
so when he originated this arrangement. Cf. Plezia, Andron., 53f. The scholia of MS Par. Gr. 1853 on
the opening of Metaph. 2 have been discussed by S. Bernardinello, ‘Gli scolî alla Metafisica di
Aristotele nel F 234r del Parisinus Graecus 1853’, Elenchos 3, 1982, 39–54.
65
See above, n. 21 and ANRW 36.2. 1112.
66
See Drossaart Lulofs, p. 9ff.
67
cf. C.A. Brandis, Über Ar.’s Rhetorik u. d. gr. Ausleger derselben’, Philologus 4, 1849, 34ff; F. Solmsen,
‘The Aristotelian tradition in ancient rhetoric’, AJP 62, 1941, 35–50, 169–90 = id., Kleine Schriften II ,
Hildesheim 1968, 178–215; R.B. Todd, Alex. of Aphr. on Stoic Physics, Leiden 1976, 12ff. Some of these
omissions were not made good until the twelfth century, when Michael of Ephesus wrote
 The earliest Aristotelian commentators 75

of the Ethics is curious. Ptolemy’s catalogue (nos. 35–6) lists the Eudemian Ethics
in eight books (i.e. including the three books common to it and the EN ) and the
Magna Moralia, but omits the Nicomachean Ethics. This may be due to an
accident of transmission; Arius Didymus refers to the tenth book of the EN and
Atticus in the second century ad enumerates all three versions,68 and from this
time the Aristotelian commentators concentrate on the EN (to which the
common books are now assigned) to the exclusion of the others. But a recent
study has shown that until then the Eudemian Ethics was quoted more often and
seems to have enjoyed higher esteem than the Nicomachean.69
Besides this the Organon, the physical works, including the de Anima and at
least some of the Parva Naturalia, and the Metaphysics were studied systematically.
But above all it was the Categories which attracted most attention. Simplicius
enumerates five ‘early commentators’ who wrote about this tract and the problems
it raised in the years immediately following its publication by Andronicus; they
include Eudorus the Academic and Athenodorus the Stoic as well as the Peripatetics
Andronicus, Boethus and Ariston.70 Not much later an unknown hack compiled a
version of the Categories in the Dorian dialect which he attributed to Archytas of
Tarentum, a Pythagorean active in the early fourth century bc .71 Subsequently
almost every Aristotelian of note wrote a commentary or interpretative essay;
some Platonists incorporated Aristotle’s doctrine into their philosophy,72 others,

commentaries on the PA, IA, MA, GA, Politics and Rhetoric, as well as parts of the Organon, Ethics
and Parva Naturalia, apparently in a deliberate attempt to fill the gap left by earlier commentators; see
his commentary on the Parva Naturalia (CAG 22.1) 149. R. Browning, ‘An unpublished funeral
oration on Anna Comnena’, Proc. Cambr. Philol. Soc. 188, 1962, 6ff (Chapter 17 below), has shown that
this was ‘part of a cooperative scholarly undertaking conceived and guided by Anna Comnena’. Cf.
A. Preus in the introduction to his translation of Michael’s commentary on MA and IA, Hildesheim
1981, p. 8ff; H. Hunger, Die hochsprachliche profane Literatur der Byzantiner I, Munich 1978, 34f.
68
Arius Did. ap. Stob. 2 p. 52.10 (referring to 1172b9); Atticus fr. 2.9 Edouard des Places (ed.) Paris
1977, cf. Moraux, Listes, 307. The paraphrase of the EN ascribed to Andronicus belongs to a much
later period; see Moraux, Arist., 136ff.
69
Kenny, Ar. Ethics, ch. 1; his conclusions have been qualified in a review by D. Charles, JHS 100, 1980,
224f. According to Kenny (29ff ) the change was due to Aspasius, who believed that the EE,
including the common books, was the work of Eudemus, but transferred the common books to the
EN because they seemed necessary to fill a gap whose existence Aspasius inferred from a cross-
reference at EN 1155b15 (Kenny gives the reference wrongly as 1135b13–16). But Kenny appears to
have misread Aspasius (see ANRW 36.2. 1158 and n. 375) and as we have seen, Arius already knew
the ten-book form of the EN. Kenny’s attempt (p. 21f) to discredit this evidence by attributing the
introduction of Arius’ Epitome to Stobaeus himself runs counter to everything we know about
Stobaeus’ method; cf. C.H. Kahn, ‘Arius as a doxographer’, in W.W. Fortenbaugh (ed.), On Stoic and
Peripatetic Ethics: the work of Arius Didymus, New Brunswick 1983 (a collection of papers presented
at a symposium in 1981), 4 n. 1. – See also R. Bodeus, ‘Contribution à l’histoire des oeuvres morales
d’Aristote: les Testimonia’, Rev. Philos. de Louvain 71, 1973, 451ff.
70
Simpl. in Cat. 159,32; cf. Brandis, Abh. Akad., Berlin 1833, 272ff.
71
See ANRW 36.2.1131ff.
72
e.g. Albinus (Alcinous) Didascalicus 6.10 p. 159 Hermann; Anon. in Plat. Theaet. 20,34; 40,10;
68,10ff. Cf. Philo de Decal. 30; Dörrie, Platonica Minora, 300.
76 Aristotle Transformed

notably Plotinus, attacked it.73 All these efforts were summed up in the great
commentaries of Alexander of Aphrodisias and Porphyry;74 through the latter this
work entered the mainstream of the Neoplatonic tradition and went on to become
one of the basic philosophical texts of the Middle Ages.
What is remarkable about this is that Aristotle’s earliest followers seem to
have attached much less importance to this tract. We are told, on late and
somewhat dubious authority, that Theophrastus and Eudemus wrote Categories
in imitation of their master.75 But since no quotations have come down to us, it
seems that any writings of theirs on this subject were regarded as relatively
unimportant, and that they did not develop this aspect of Aristotle’s teaching to
the same extent as other parts of his logic.76 There is also the fact that the
Categories is not mentioned in the original versions of the Hellenistic catalogue
of Aristotle’s writings, unless it is hidden under the mysterious title Ta pro tôn
topôn á.77 It would appear to have fallen into even deeper oblivion than most of
the other books of the Aristotelian corpus, until Andronicus brought it into
prominence by placing at at the head of his edition.
It would be interesting to know why the Categories came to exercise so much
fascination, but there is no evidence. What we can do is to see what kind of
question the early commentators asked and how they set about answering
them. Many of their queries concerned points of detail and some were trivial
and even perverse, but they also raised the problems which later generations of
interpreters regarded as important and anticipated many of the solutions they

73
Plot. Enn. 6.1–3; cf. Simpl. in Cat. 2,3. See H.J. Blumenthal, ‘Plotinus in the light of twenty years’
scholarship, 1951–1971’, ANRW II , 36.1.547ff; K. Corrigan and P. O’Cleiright, ‘The course of
Plotinian scholarship from 1971 to 1986’, ib. 579ff and S.K. Strange, ‘Plotinus, Porphyry and the
Neoplatonic interpretation of the “Categories” ’, ANRW II 36.2.955ff.
74
See Simpl. in Cat. 1,13ff, who goes on to say something about later commentators down to his own
day. The commentary of Porphyry here referred to is the one dedicated to Gedalius, now lost; his
surviving commentary is a much shorter work intended for beginners (cf. A. Smith, ‘Porphyrian
studies since 1913’, ANRW II 36.2.755 and S.K. Strange, ‘Plotinus, Porphyry and the Neoplatonic
interpretation of the “Categories” , ib. 956f).
75
Philop. in Cat. 7,20ff; Olympiod. in Cat. 13,24ff; David in Isag. 102,4ff = Theophrastus fr. 1–2 Repici,
Eudemus fr. 7–8 Wehrli (the second passage cited here has been omitted in both editions). Cf.
L. Repici, La logica di Teofrasto, Bologna 1977, 179f; M. Frede, ‘Titel, Einheit und Echtheit der
aristotelischen Kategorienschrift’, in P. Moraux and J. Wiesner (eds), Zweifelhaftes im Corpus
Aristotelicum, Peripatoi 14, Berlin 1983, 22ff; H.B. Gottschalk, ‘Did Theophrastus write a Categories?’,
Philologus 181, 1987, 245–53.
76
cf. Brandis, 1833, 268ff. I.M. Bochenski, La logique de Théophraste, Fribourg 1947, 31f, 37 thinks that
Theophrastus did not write a Categories and this is an argument against the authenticity of the
Aristotelian work. He is probably right on the first point, but the second does not necessarily follow.
77
Diog. Laert. no. 59 = Hesychius no. 57; cf. Moraux, Listes, 58ff, id., Arist., 99ff. Its occurrence among
the collections of material near the end of the list is the result of a late interpolation, cf. Moraux,
Listes, 131, Düring, AB p. 69.
 The earliest Aristotelian commentators 77

proposed.78 Perhaps the most fundamental is the one with which the student is
faced at the outset: what is the subject-matter of the Categories? Is Aristotle
saying something about terms, or the things signified by terms? There are good
arguments on both sides: Aristotle’s language is ambiguous, and while most of
the distinctions he makes are more appropriate to things than verbal expressions,
he arrives at his classification by analysing the expressions which can be
meaningfully applied to things.79 A minority of ancient interpreters came down
decisively on one side or the other; thus Plotinus insisted that the Categories
deals with the classification of real things, and his critique was given the title On
the classes of being.80 On the other hand the Stoics, Athenodorus and Cornutus
seem to have treated the essay as an attempt to analyse language and criticised
Aristotle for not including some parts of speech, such as conjunctions, and
grammatical distinctions in his discussion.81 This is in accord with their
formalistic view of logic82 as well as their nominalist interpretation of Plato’s
Forms,83 but implies that they regarded Aristotle’s system of categories as
something very different from their own, which they certainly looked upon as a
classification of real things.84 These are the extreme positions. Most interpreters
accepted a compromise which seemed to do justice to all the arguments: the
Categories is about simple terms considered not as terms (i.e. in a way appropriate
to grammatical science) but as signifying things. This solution was given its
classical formulation by Porphyry, but was already hit upon by Boethus and
accepted by such eminent Peripatetics as Herminus, Alexander of Aigai and
Alexander of Aphrodisias.85

78
The main source of our knowledge is Simplicius’ commentary on the Categories, occasionally
supplemented by those of Porphyry, Dexippus and Ammonius. On their inter-relationship see
Praechter, ‘Nikostratos der Platoniker’, in H. Dörrie (ed.), Kleine Schriften, Collectanea 7, Hildesheim
and New York 1973, 123–8 (first published in Hermes 57, 1922, 481–517).
79
The controversy is summarised by Simpl. in Cat. 9–13.
80
Peri tôn genôn tou ontos = Enn. 6.1–3; the title was devised by Porphyry, but is amply justified by the
opening words of the treatise. Cf. Simpl. in Cat. 2,3, 16,17ff.
81
Porph. in Cat. 59,10ff, 86,20ff; Simpl. 18,26ff.
82
e.g. Galen Inst. Log. 3.5, 4.6.
83
Zeno fr. 65 (SVF 1 p. 19.27); cf. Syrian. in Metaph. 105,21ff = SVF 2.364.
84
SVF 2.369ff; cf. J.M. Rist, ‘Categories and their uses’, in A.A. Long (ed.), Problems in Stoicism,
London 1971, 38ff, who refers to other discussions, and below, p. 78.
85
Porph. in Cat. 58,5ff, 59,5ff; Simpl. in Cat. 10,20ff, 11,23ff, 13,13ff, 41,28ff; cf. Moraux, Arist., 148ff.
Different aspects of the problem are discussed by K. v. Fritz, ‘Der Ursprung der aristotelischen
Kategorienlehre’, Arch. Gesch. Philos. 40, 1931, 449–96 (revised: id., Schriften zur griechischen Logik,
II : Logik, Ontologie und Mathematik, Problemata 71, Stuttgart 1978, 9–52) and A.C. Lloyd,
‘Neoplatonic logic and Aristotelian logic’, Phronesis 1, 1956, 58–79, 146–60. It is instructive to
compare the ancient discussions with their counterparts in modern textbooks and commentaries;
while opinions differ on many points, the arguments used and the range of possible views have not
changed significantly.
78 Aristotle Transformed

Connected with this was another controversy, as to how many categories were
to be postulated and the order in which they were to be placed.86 In general, those
who favoured a verbalist interpretation tended to regard Aristotle’s analysis as
incomplete, because it made no provision for grammatical inflections and such
words as conjunctions. Besides Athenodorus and Cornutus, this criticism is
attributed to the Platonists Lucius and Nicostratus.87 But although it persisted
into the second century, it was already refuted by Boethus, who pointed out that
the parts of speech omitted by Aristotle were irrelevant because they only function
in sentences and do not signify anything by themselves. Presumably his objections
were directed at Athenodorus; we do not know if Cornutus and the rest modified
his position in order to meet Boethus’ arguments.88 Others wanted to reduce the
number of categories. Thus Andronicus is said to have held that the ten categories
could be reduced to two, the self-subsistent and the relative.89 According to
Simplicius, Andronicus was here following the lead of Xenocrates, but whereas
his classification is connected with his derivation of existing things from the
One and Indefinite Dyad,90 it is impossible that it should have had the same
metaphysical significance for Andronicus. It is not even certain that Andronicus
was directly influenced by Xenocrates at this point, for the connection between

86
Simpl. in Cat. 18,22ff, 62,24ff.
87
Simpl. in Cat. 18,28ff, 62,24ff, 64,13ff. For Nicostratus, see K. Praechter, Kleine Schriften, Hildesheim
1973, 101ff; Lucius is only known through his association with him. See below, p. 80.
88
Simpl. in Cat. 11,23ff. Moraux, Arist., 150 thinks that Athenodorus wrote after Boethus and Boethus’
criticism is directed against an earlier Stoic or even Andronicus. But if Athenodorus is to be
identified with the son of Sandon and teacher of Octavian, as is generally believed, he was an old
man when he returned to his native Tarsus and entered local politics there shortly after the battle of
Actium (Strabo 14.674f); so his philosophical activity must have fallen into the years 60–30 bc and
be contemporaneous with that of Boethus. Moraux’s reasons for thinking that Andronicus favoured
a verbalist interpretation are quite insufficient; see below.
89
‘For the followers of Xenocrates and Andronicus seem to comprehend everything under the in-
itself and the relative, so that according to them such a crowd of categories is excessive’, Simpl. in
Cat. 63,22 = Xenocr. fr. 12 Heinze = 95 Isnardi Parente; cf. Moraux, Arist., 103f. The last clause (hôste
. . . plêthos) seems to be Simplicius’ own conclusion. J. Dillon, The Middle Platonists, London 1977,
133 and H. Dörrie, Platonica Minora, Munich 1976, 300 claim that Eudoros took the same view, but
the passages they adduce (Simpl. in Cat. 174,14ff, Dillon, 256,16ff, Dörrie) do not bear them out,
and elsewhere Eudoros writes as if he accepted all ten categories; see below, p. 77f. The same division
is used in the anonymous commentary on Plato’s Theaetetus, col. 20.34 and 40.10; contrast 68.10ff.
90
In their commentaries Heinze (p. 37ff.) and Isnardi Parente (p. 327ff, 439ff ) refer to Hermodorus
ap. Simpl. in Phys. 248,2ff (= Hermodorus fr. 7 Isn. Parente) and Sext. Emp. Adv. Math. 10.263ff. Cf.
K. Gaiser, ‘Platons Menon und die Akademie’, Arch. Gesch. Philos. 46, 1964, 243f (reprinted in
J. Wippern (ed.), Das Problem der ungeschriebenen Lehre Platons, Wege der Forschung 186, Darmstadt
1972, 329–93). The twofold division of onta is also found in the Divisiones Aristoteleae (Diog. Laert.
3.108f = Cod. Marc. no. 67, p. 39f Mutschmann); here substances are said to differ from relatives
inasmusch as the terms denoting the latter, but not the former, ‘require some explanation’ (deitai
tinos hermêneias) to make them meaningful, e.g. ‘larger’ must be larger than something. Tarán, rev.
of Moraux, Arist., Gnomon 53, 1981, 741 thinks that Andronicus was influenced by the Stoics as well
as Xenocrates and offers a different interpretation of his doctrine, but the balance of the evidence
does not support his view.
 The earliest Aristotelian commentators 79

them may have been made by Simplicius himself or an intermediate source,

perhaps Porphyry. In another part of his commentary (p. 342,24), Simplicius
admits that Andronicus ‘maintained the tenfold number of genera’, and he has
much to say about his detailed modifications of Aristotle’s teaching about
particular categories. So it appears that Andronicus did little more than point up
the fundamental distinction, clearly drawn by Aristotle himself,91 between the
category of Substance, whose members subsist in their own right, and the
remaining categories, which depend on substances for their being, without
wishing to obliterate the differences between the latter. Indeed, he seems to have
made an attempt to order them according to the degree to which they were
essential to substances, and to have placed Relation last because its connection
with Substance was loosest.92 Here his treatment could have been influenced by
Stoic notions93 and seems to imply a ‘realist’ interpretation of Aristotle’s doctrine.
The other fragments of Andronicus’ commentary are concerned with details.94
In his paraphrase he tried to clarify Aristotle’s thought by expanding his text
where it seemed unduly compressed or by slight changes of wording; occasionally
he replaced a loose expression by a more formally correct one.95 On occasion,
however, he can interpret Aristotle more freely. Two instances deserve mention
here. One concerns the difference between substance and attribute. Aristotle
distinguished between the two kinds of predication of which ‘Socrates is a man’
(or ‘animal’) and ‘Socrates is a philosopher’ are instances: the first tells us the
species or genus to which Socrates belongs, and the universals, ‘man’ or ‘animal’

91
Arist. Cat. 2a34, etc., cf. An. Post. 83a25ff. The same point is made very clearly by ps-Archytas Peri
tô katholou logô p. 31,6ff. Thesleff, cf. p. 32,7f. (cf. Simpl. in Cat. 76,1ff ) and Boethus de Trin. ch. 4.
Cf. K. v. Fritz, ‘Der Ursprung der aristotelischen Kategorienlehre’, Arch. Gesch. Philos. 40, 1931, 462,
478ff, 487ff; revised: id., Schriften zur griechischen Logik, II : Logik, Ontologie und Mathematik,
Problemata 71, Stuttgart 1978, 20, 35ff, 43ff; H.W.B. Joseph, An Introduction to Logic, Oxford 1916
etc., 49ff; W.K.C. Guthrie, A History of Greek Philosophy VI , Cambridge 1982, 142; Prantl I 564
claims to find a similar doctrine in Galen Meth. Med. 2.7 (X 129 Kuhn), but Galen is not concerned
with the doctrine of categories at this point; in the Institutio Logica he refers to all ten categories and
even claims to have discovered an eleventh; see ANRW II 36.2.1170.
92
Simpl. in Cat. 157,18ff; his reasoning is based on Ar. EN 1096a21ff, cf. Metaph. 1088a23; Joseph 51
n. 1, Moraux, Arist., 107ff. Similar reasoning was used by Eudoros ap. Simpl. 206,10ff and ps-
Archytas p. 23,17ff; cf. T.A. Szlezák, ‘Ps.-Archytas über die Kategorien’, Peripatoi 4, Berlin 1972, 110.
93
cf. Simpl. in Cat. 166,8ff = SVF 2.403.
94
Littig’s collection (Andronikos III , 14ff ) is almost complete; additional passages are Simpl. in Cat.
154,3ff and 332,14f. For a full discussion see Moraux, Arist., 101ff; cf. Prantl I, 537ff; Plezia, Andron.,
6ff, 36ff.
95
e.g. at 8a31f, where Aristotle defines relatives (ta pros ti) as things whose being consists in standing
in some relation (pros ti) to something, Andronicus objected that his definition contains the
definiendum and proposed to substitute pros heteron for pros ti, using the term employed by
Aristotle at 6a37 (there is no need to invoke Stoic influence here, with Tarán, 742). He was followed
by Boethus and Ariston; Achaicus, a Peripatetic probably of the second century ad, tried to show
that the change was unnecessary, but formally Andronicus is right, although the point is trivial. The
source for all this is Simpl. in Cat. 201,34ff.
80 Aristotle Transformed

themselves belong to the category Substance; the second states one of his
attributes and belongs to Quality. Aristotle tries to describe the difference by
saying that the first kind of universal, the ‘second substance’, is predicated of
(kata) the concrete individual while the second inheres in (en) the individual.96
But according to Simplicius, Andronicus claimed that some attributes are
predicated of their subject, e.g. ‘philosopher’ of Socrates.97 We are not told the
reasoning which led Andronicus to adopt this position or the criteria by which
he decided which attributes fell into this class,98 but it is clear that it entailed
some blurring of the distinction between ‘second substances’ and qualities. We
shall find a similar tendency in Boethus; it may be the result of Stoic influence.
The second innovation of Andronicus I shall consider here concerns the
categories which modern writers generally call Time and Place. Aristotle did not
use nouns to designate them, but the adverbs ‘When’ (pote) and ‘Where’ (pou)
and treated time and place as kinds of extension belonging to the category
Quantity (Cat. 5a6ff ). Andronicus, however, decided that time and place exist in
their own right and ‘when’ and ‘where’ are their attributes; he therefore renamed
the categories ‘Time’ (khronos) and ‘Place’ (topos) and assigned ‘when’ and ‘where’
subordinate places within them.99 His view is discussed and criticised by
Simplicius (pp. 342–64) at considerable length, but it is not certain how much of
his argumentation goes back to Andronicus himself. One point, however, is
clearly made by Simplicius: here as elsewhere Andronicus inclined towards a
realist interpretation of the categories; he and ps-Archytas differed from Aristotle
because they were thinking primarily of the relationships between things falling
under the categories, whereas Aristotle concentrated on the meanings of the
terms used to denote them.

96
cf. G.E.L. Owen, ‘Inherence’, Phronesis 10, 1965, 97ff.
97
Arist. Cat. 1a20ff, cf. 2a18ff; Simpl. 54,8ff; cf. Moraux, Arist., 104f.
98
Simplicius seems to say that it included all attributes of which we can say that e.g. ‘Socrates is x’,
where ‘x’ can be construed as a noun in apposition to ‘Socrates’, but this is clearly too wide, as it
would cover relative terms like ‘husband of Xanthippe’.
99
Simpl. in Cat. 347,6: ‘But Archytas and Andronicus, supposing that time exists independently (i.e.
not as an aspect of quantity), placed When in the same category, as existing within the ambit of
time. For it is his rule to take as principal the leading first genera in each <kind of> existence, in
whose ambit the others exist; since therefore Time precedes When and Place precedes Where in
respect of <their> very being, he rightly treats them as genera in a leading sense. So it has been
shown above that Aristotle puts When in a different category <from time> because he concentrates
on the differences in the meanings <of the terms>, but Archytas placed When in the same category
as Time because he paid more attention to the kinship of the things <they signify>.’ 357,28: ‘Archytas
and Andronicus placed When with Time and Where with Place and established the two categories
in this way. But Aristotle, for the reasons stated above, made Time and Place belong to Quantity and
set up two different categories of When and Where.’ ‘Archytas’ here means ps-Archytas Peri tô
katholou logô p. 26,10ff Thesleff; cf. Szlezák, ad loc., p. 126f; Littig III , 25ff, Moraux, Arist., 111ff. The
same terminology is used by Philo, Decal. 30; cf. Eudoros ap. Simpl. in Cat. 206,13ff.
 The earliest Aristotelian commentators 81

The most elaborate of the early commentaries was the one written by Boethus
of Sidon which, as well as line-by-line exegesis and textual criticism, included
in-depth examinations of some of the problems Aristotle raised; Boethus
devoted a whole book to the category of Relation, but we are not told whether
this was part of his commentary or a separate essay.100 Apparently his intention
was to set the Categories in the context of Aristotle’s philosophy as a whole and
of the Academic-Peripatetic tradition. He quoted extensively from Aristotle’s
writings and those of other Academics and Peripatetics.101 Where Aristotle’s
treatment was or appeared to be incomplete, he tried to fill the gap with material
drawn from other parts of the corpus; thus his commentary on the categories
Time, Action and Passion, which Aristotle did not discuss in detail, includes a
great deal of matter based on the Physics and Metaphysics.102 In general his
attitude was conservative. He rejected many of Andronicus’ innovations and
refuted many of the objections which had been brought against Aristotle,
especially by Stoic writers.103 His counter-arguments were not always accepted
by Aristotle’s opponents, and many of the criticisms he refuted were repeated by
later writers, including Plotinus, to be refuted again by others reusing many of
Boethus’ arguments.104 We cannot usually tell what modifications, if any, the
arguments on both sides underwent in the course of this process, but there can
be no doubt that Boethus was largely responsible for the scholasticism which is
such a marked feature of subsequent Aristotelian philosophy.
Work of this kind does not lend itself to detailed discussion in a survey such
as the present one, but two examples may serve to illustrate Boethus’ way of
thinking. One is his extraordinarily subtle treatment of time and its relationship
to both movement and rest. Aristotle’s dictum that the contrary of movement as
such is rest, but the contrary of any particular movement is movement in the
opposite direction, gave rise to a fair amount of controversy, some points of
which were taken up later by Plotinus. Boethus tried to explain it by introducing

100
Simpl. in Cat. 1,17, 163,6. For a full discussion see Moraux, Arist., 147–64; Prantl I, 540ff. Textual
variants or conjectures are mentioned by Simpl. 29,29ff, 58,27ff; cf. Moraux, 150, 153. A full edition
of Boethus’ fragments is a desideratum of scholarship.
101
Aristotle; ap. Simpl. in Cat. 65,21 (An. Post.); 50,2ff, 302,15ff (Phys.); 78,6, 302,15 (Metaphysics); 42,1
(de Anima). Plato: 159,14, 163,6. Speusippus: 36,28, 38,19 (cf. L. Tarán, ‘Speusippus and Aristotle on
homonymy and synonymy’, Hermes 106, 1978, 75ff ). He also used the writings of Theophrastus and
Strato. Besides this it is of little significance that he missed some passages in Aristotle which might
have been relevant to his argument: see Porphyry ap. Simpl. in Cat. 36,27 (cf. Moraux, 151). A
misquotation of Plato Rep. 438A noted by Simpl. 159,16 may be due to his quoting from memory.
102
ap. Simpl. in Cat. 348,22ff, 433,28ff.
103
See Simpl. in Cat. 167,22, 373,7, and above n. 88.
104
See, besides the passages quoted in the last note, Simpl. in Cat. 25,10ff, 58,15ff, 187,24ff, 302,5ff,
338,21ff, 433,20ff; cf. Moraux, Arist., 147f, 150, 160f, 163f.
82 Aristotle Transformed

the notion of time as a coordinate of both movement and rest. Time, he argued,
being in constant flux, is a concomitant (sumparêkei) of both movement and rest;
when objects are in movement, they stand in the same relationship to time as to
the space they traverse (or, if the movement is one of qualitative change, to the
spectrum of qualities within which the change takes place), but when they are at
rest, their relationship to time is the contrary of their relationship to space etc., in
that the former is still changing while the latter is not. Thus movement and rest
can be seen as contrary relationships to time.105 This implies a certain view of
time: not as the ‘number’ or ‘measure’ of movement, but as the quantitative or
extensional aspect of movement and rest – a view similar to that of Theophrastus’
successor Strato or of the Stoics.106 It follows that time is an objective reality,
independent of the existence of a subject to perceive it,107 and belongs, together
with place, in the category Quantity. Boethus agreed with Aristotle, against
Andronicus, in making When and, by analogy, Where, into separate categories.108
Even more revealing is his discussion of Substance. He began by denying the
relevance of a question raised by a previous writer, perhaps Eudorus,109 whether
the category Substance was meant to embrace intelligible as well as sensible
substances, but then went on to raise a problem of his own. In the Categories
(2a11ff ) Aristotle distinguished two kinds of substance, concrete individuals and

105
Arist. Cat. 15b1ff, Simpl. ad loc. p. 432,24ff; Plotinus’ view (Enn. 6.3.27) is quoted at p. 433,20ff,
Boethus at 433,28ff. Cf. Moraux, Arist., 163f.
106
Strato fr. 76–8 F. Wehrli (Die Schule d. Aristoteles V, Basel 1952 etc.), and ap. Simpl. in Cat. 346,14
(omitted by Wehrli), ‘Strato, saying that time is the quantity of movement, supposed that it is
something inseparable from movement’. Stoics: SVF 1.93, 2.509–16.
107
Themistius in Phys. 160,26f, 163,6ff = Simpl. in Phys. 759,18ff, 766,18ff. Here Boethus contradicted
Arist. Phys. 223a21ff; later Alexander of Aphrodisias tried to defend Aristotle’s position (ap. Simpl.
in Phys. 759,20ff ). Cf. Moraux, Arist., 170f.
108
Simpl. in Cat. 348,2ff. A similar doctrine is found in an essay Peri tês tou pote katêgorias preserved
in a Florentine manuscript (Laur. 71.32) and printed by Waitz in the introduction of his edition of
the Organon, Leipzig 1844, I, 19–23 and by P.M. Huby, ‘Boethus of Sidon’s Commentary’, CQ n.s. 31,
1981, 398ff, who claims that it is nothing less than an excerpt from Boethus’ commentary on the
Categories. This is an attractive hypothesis, but there are difficulties. Simplicius in Cat. 347,22–36, in
a passage apparently borrowed from Iamblichus, defines the category ‘When’ as the relation between
time and the things in time. The same definition is found in the anonymous work (p. 22 init. Waitz),
where it is illustrated by means of an example which Simplicius ascribes to Boethus a few lines later
(348,2); but since he introduces his quotation with the words ‘Boethus also . . . supposes . . .’, it does
not look as if he found the definition and the illustration together in his source. Furthermore the
anonymous writer tries to reconcile his theory with the notion that time is the number of movement,
but our authorities seem to imply that Boethus refused to do this; see Simpl. in Cat. 434,18f (Huby,
407 takes a different view of this passage) and Themistius in Phys. 163,6 = Simpl. in Phys. 766,18.
While the anonymous tract contains much that can be traced back to Boethus, it would not be safe
to ascribe anything to him which is not attributed to him by name in an ancient source.
109
Simpl. in Cat. 78,4ff, cf. T.A Szlezák, Pseudo-Archytas über die Kategorien, Berlin and New York
1972, 105. The problem was taken up again by Nicostratus and Plotinus (Enn. 6.1.2, cf. Simpl. in Cat.
76,13ff ) and is presupposed by ps-Archytas Peri tô katholou logô p. 22,31ff Thesleff, cf. Simpl.
76,19ff, 77,8ff.
 The earliest Aristotelian commentators 83

the species or genera to which they belong; the second kind he refers to here, but
nowhere else, as ‘second substances’ (2a14). In the Metaphysics (1028b33ff etc.),
however, he says that ‘substance’ can mean one of three things: matter, form or
universal, or the concrete individual composed of both. Boethus asked how
these classifications are related and pointed out that the definition of substance
given in the Categories (2a11), ‘That which is neither in something else nor
predicated of something else’, fits matter and concrete individuals but would
exclude the form or universal, because it is predicated of the individual. Therefore,
he concludes, forms or ‘second substances’ do not belong to Substance but to a
different category, such as Quality or Quantity.110 Consistently with this view he
held that the unity of individuals is due to their form ‘in as much as the form is
limited and made numerically one’ by being included in matter.111 In this way
Boethus, as Dexippus says, affirmed the priority of the concrete individual over
the universal. He could have found support for his opinion in such passages as
Cat. 3b15 or Metaph. 1033b22 and his interpretation points forward to that of
Alexander, but one cannot avoid the suspicion that it was influenced by Stoic
doctrines.112 But Boethus took care to base his argument on a painstaking
analysis of Aristotle’s own words. His intention was not to mediate or compromise,
and in other contexts he shows himself a true Aristotelian. For example, he
refuted those who wanted to subsume Action and Passion under a single category
Movement by pointing out that the distinction is required by Aristotle’s doctrine
of a first mover who is active but not moved;113 the view he refuted was taken up
again later by Plotinus (Enn. 6.1.15), but originated among Stoic opponents of
Aristotle.114 Boethus wanted nothing more than to be a faithful exponent of
Aristotelianism. He tried to keep the discussion within an Aristotelian framework
and saw his main task as explaining the real or apparent discrepancies he found
in Aristotle’s writings. But he could not altogether escape the influences which
had established themselves in the centuries after Aristotle’s death.

110
Simpl. in Cat. 78,5ff, Dexippus in Cat. 45,27ff; cf. ps-Archytas Peri tô katholou logô p. 24,17ff. Th. See
further Prantl I, 541ff, Moraux, Arist. 155ff, G. Movia, Anima e Intelletto, Padua 1968, 194ff. The
question whether the specific differentia qualifies the genus or species (ap. Simpl. in Cat. 97,28ff ),
which Prantl connects with this doctrine, belongs to a different discussion.
111
Simpl. in Cat. 104,26f, cf. Dexippus 45,22ff.
112
SVF 2.369ff, cf. Prantl I, 430ff, Zeller III 1,95ff, Rist (above, n. 84), 43ff.
113
Simpl. in Cat. 302,12ff, cf. 306,13ff; cf. Szlezák, Pseudo-Archytas über die Kat., 126; Moraux, Arist.,
160. Note the implication: in this case, at least, Boethus applied the doctrine of categories to an
intelligible substance.
114
However Aristotle himself names Kinêsis as a category at Metaph. 1029b25, and at EE 1217b29 has
two categories kinein and kineisthai. Cf. v. Fritz, ‘Der Ursprung der aristotelischen Kategorienlehre’,
Arch. Gesch. Philos. 40, 1931, 451, 465, 484 (revised: id., Schriften zur griechischen Logik, II : Logik,
Ontologie und Mathematik, Problemata 71, Stuttgart 1978, 11, 24, 40).
84 Aristotle Transformed

The other ‘early commentators’ on the Categories are less important, at least
for their contribution to the knowledge of Aristotle. Only one, Ariston of
Alexandria, was a professed Aristotelian. Two of his comments are recorded,
both concerned with Relation; one echoes a remark of Andronicus and
Boethus,115 the other raises the question whether ‘cosmos’ is a relative term and
what its correlate might be. It has been variously interpreted and may have been
distorted in the course of transmission, but the problem seems to hinge on the
ambiguity of the expressions kosmos and to en kosmôi.116 Eudorus, his fellow
Alexandrian and rival,117 tried to determine the correct order in which the
categories should be placed. Arguing that every substance must have some
properties and extension, he placed Quality and Quantity immediately after
Substance; after these came Time and Place which, though extrinsic to substances,
are necessary for their existence.118 This is something to which Aristotle seems to
have attached little importance; the order in which the categories are enumerated
at Cat. 1b26f differs from that in which they are discussed in the body of that
work, and different arrangements are found in his other writings. It is
characteristic both of Eudorus and the intellectual climate in which he worked
that he should have devoted so much effort and ingenuity to this task. His
decisions may have been influenced by Aristotelian and perhaps even Platonic
precedents,119 or by metaphysical considerations like those used by ps-Archytas

115
See above n. 95.
116
Simpl. in Cat. 188,31ff = fr. 2 Mariotti. See Zeller III 1, 649; Prantl I, 546; Moraux, Arist., 183–5;
I. Mariotti, Aristone d’Alessandria, Bologna 1966, 48–57 tries to explain it by reference to the Stoic
distinction between to holon and to pan, but this is far-fetched. The difficulty arises from the fact
that the terms employed by Ariston in this context include three neologisms whose exact meaning
is not clear (kosmôtos, geôtos, aerôtos), as well as the ambiguous to en kosmôi.
117
According to Strabo 17.790C, both published books about the source of the Nile whose content was
so similar that Eudorus accused Ariston of plagiarism. See further Zeller III 1, 633ff; Dillon, Middle
Platonists, London 1977, 115ff (on the Categories, 133ff ); H. Dörrie, ‘Der Platoniker Eudoros von
Alexandreia’, Hermes 79, 1944, 25–38 = id., Plat. Min., 297–309; Praechter, Kl. Schr., 130f (see n. 78);
Fraser, Ptolem. Alexandria I,489; II , 708f; Moraux, Arist. II , 509–27.
118
Simpl. in Cat. 206,10ff; cf. ps-Archytas Peri tô katholou logô p. 23,17–24,16 Thesleff, with Szlezák’s
notes, pp. 108–18; this is the earliest extant attempt to fix the order of the categories and clearly
owes much to Eudorus, although their conclusions differ in some respects. W. Theiler, Unters. zur
antiken Lit., Berlin 1970, 489f, followed by Szlezák 116, claims that Eudoros also influenced the lists
of categories given by Philo Alexandrinus, de Decalogo 30 and Boethius de Trinitate ch. 4 init,
although both, like Arist. 1b26, place Relation fourth, after Substance, Quality and Quantity, Philo
puts Time and Place last, and Boethius retains the terms Ubi and Quando (Where and When).
Theiler’s assertion that Boethius placed Relation last is wrong; his order is Substantia, Qualitas,
Quantitas, Ad aliquid, Ubi, Quando, Habere, Situm esse, Facere, Pati, i.e. essentially the same as in
Aristotle 1b26.
119
Szlezák, 110 thinks that Eudorus placed Quality next to Substance because he was influenced by the
Platonic distinction between ousia or ti estin and poion ti (on which see K. Gaiser, Arch. Gesch.
Philos., 243f), but this would at most account for the placing of one category, and the same order is
found at Metaph. 1017a25.
 The earliest Aristotelian commentators 85

for the same purpose,120 but that the question was raised at all may have been due
to the Stoics, in whose system it was more important;121 indeed Aristotle’s
seeming neglect of it may have been one of their reasons for interpreting his
doctrine as a classification of terms rather than things. – The other comments on
the Categories ascribed to Eudorus are petty and pedantic to the point of
captiousness; most of them are concerned with the sub-divisions of the category
Quality, one with a detail in Relation.122
The last of the ‘old commentators’ was a Stoic, Athenodorus, who wrote a
book entitled Against Aristotle’s Categories; he is generally quoted together with
Cornutus, another Stoic active in the middle of the first century ad.123 As we
have seen already, they thought that the Categories is concerned with language
and criticised Aristotle for unduly restricting the scope of his analysis (above,
p. 70). Their remaining comments were less far-reaching. Athenodorus held that
weight should be regarded as a third subdivision of the category Quantity,
together with number and extension; Cornutus disagreed, preferring to place
weight, i.e. heaviness and lightness, in the category Quality.124 Both agreed in
denying that objects such as heads or wings could fall into the category Relation,
although they gave different reasons for this.125 Probably it was Athenodorus
who suggested substituting a single category Movement for the Aristotelian
Action and Passion, the idea refuted by Boethus.126 Some remarks about Time
and Place are attributed to Cornutus alone.127
These writers did little for the elucidation of Aristotle, but their attitude raises
an interesting problem. The Stoics appear to have been hostile to Aristotle’s
doctrine from the start and, unlike the Platonists, made no attempt later to
incorporate it in their system; apparently their chief concern was to emphasise the

120
Szlezák, 116 has some interesting speculations on this point.
121
Simpl. in Cat. 165,32ff = SVF 2.403; cf. Rist (above, n. 84) 54f.
122
See the index to Simpl. in Cat.; Prantl I, 539f, Zeller III 1, 633.
123
See the index to Kalbfleisch’s edition of Simpl. in Cat. The title of Athenodorus’ book Pros tas
Aristotelous katêgorias, is given by Simpl., 62,25; Cornutus’ book was entitled Pros Athênodôron kai
Aristotelên (ibid. 62,28, Porph. in Cat. 86,23), but we only hear of one substantial disagreement;
Cornutus also criticised Aristotle’s Categories in a Tekhnê Rhêtorikê. See further Brandis (1833),
275, Prantl I, 538f; Zeller III 1, 607f; Praechter, Kl. Schr., 128 (see n. 78); for Athenodorus’ date see
above, n. 88.
124
For Athenodorus see Simpl. 128,7; he is associated with ps-Archytas (p. 25,1–3 Th.) and Ptolemy
the astronomer; cf. Szlezák, 120f, Sambursky, The Physical World of Late Antiquity, London 1962, 83.
For Cornutus see Simpl. 129,1ff; he is joined with Porphyry. The statement of Brandis and Prantl
(539 n. 20) that Cornutus agreed with Athenodorus on this point is based on a wrong reading in the
old editions of Simplicius’ commentary.
125
Simpl. in Cat. 187,28ff.
126
See above, p. 76. Since Aristotle himself seems to do this at Metaph. 1029b25, this may be evidence
that Athenodorus had read the Metaphysics.
127
Simpl. in Cat. 351,21ff, 359,1ff, 15ff.
86 Aristotle Transformed

differences between Aristotle’s teaching and their own. Yet they took an active part
in the debate sparked off by Andronicus and seem to have exercised some
influence on him and the other early commentators. All this suggests that they
were resisting an attempt to find Aristotelian ideas in their teaching. All the ancient
authorities who write about a Stoic doctrine of ‘Categories’ as if it were a mere
variant of the Aristotelian one are late commentators on Aristotle,128 and there is
no good evidence that the early Stoics used the term ‘category’ in anything like its
Aristotelian (and modern) sense. Thus it would appear that the ‘Stoic doctrine of
Categories’, about which so much has been written, originated in an attempt by
Andronicus, or possibly Eudorus, to read Aristotelian notions into Stoicism.129 If
this is correct, Athenodorus was right to object: the assimilation of the Stoic
classification to the Aristotelian Categories has caused nothing but confusion.130
The burst of activity following the publication of Andronicus’ edition was
succeeded by a lull. We know the names of many Peripatetic commentators
active in the first century ad, but the paucity of quotations suggests that their
work was not highly regarded by later authorities, and we hear of no commentaries
on Aristotle by members of other schools. It was not until the middle of the
second century ad that there was a revival, associated with the names of
Aspasius, Adrastus, Herminus and Sosigenes, the last two being teachers of
Alexander of Aphrodisias. The work of Aspasius, Adrastus and Alexander will be
discussed elsewhere in this book; it remains for me to mention what little we
know of commentators active in this period who did not belong to the Peripatetic
school. Only three men have to be considered, all of them Platonists: Atticus,
Lucius and Nicostratus. Atticus, who lived under the emperor Marcus Aurelius,
is chiefly known for a series of intemperate attacks on Aristotle and all his works
preserved by Eusebius,131 but other sources attribute two remarks to him
concerning Aristotle’s definitions of homonyms and synonyms at the beginning
of the Categories which his latest editor, des Places, has taken as evidence that he
wrote a commentary on that work.132 However, no comments on the doctrine of

128
Plotinus, whose critique of the Stoic doctrine (Enn. 6.1.25ff ) is the most comprehensive account of
it extant today, leaves us in no doubt that it differed fundamentally from Aristotle’s.
129
The doctrine of ‘Body going through body’, on which see Todd (op. cit. in n. 67 above) 73ff, would
be an analogous case.
130
In passing, let me point out one difference which has generally been overlooked. The Stoic terms
hupokeimenon, poion, pôs ekhon and pros ti pôs ekhon are adjectives, while the names of the
Aristotelian categories are nouns or adverbs construed as nouns.
131
Eusebius PE 11.1.15.4–9, 12–13. The fragments have been collected (with some omissions) by E. des
Places, Paris 1977. See further Dillon, 247–58; Düring AB p. 325ff; M. Baltes in Jahrb. für Antike u.
Christentum, Ergänzungsband 10, Münster 1983, 38–57; C. Moreschini, ANRW II 36.1.477–91.
132
Simpl. in Cat. 30,16ff; 32,19ff ∼ Porph. in Cat. 66,34ff, incompletely reported as fr. 41–2 des Places.
Cf. Brandis, Abh. Akad., Berlin 1833, 279.
 The earliest Aristotelian commentators 87

categories itself or any other logical question are attributed to him, and there is
nothing in his longer fragments to suggest that he had studied any Aristotelian
treatise at first hand; so des Places’ claim must remain very doubtful. The first of
these remarks is a footling criticism, but the second raises a question about the
relationship of the analogical and metaphorical use of terms which, while only
remotely connected with what Aristotle says, is not without interest. The first one
is said to be an elaboration of an objection by Nicostratus, who seems to have
been about a generation older than Atticus. This philosopher, who is known
from a Delphian honorific inscription and from the pages of Aulus Gellius, is
said by Simplicius to have been the author of a treatise containing objections to
nearly everything said in the Categories, written in a ‘hostile and impudent’
manner; it was based on a similar work by an otherwise unknown Lucius.133
They are quoted together half a dozen times, Lucius alone three times and
Nicostratus alone twelve times; eight of these quotations are in the section
dealing with the Postpraedicamenta, and it looks as if Lucius may not have
included these chapters in his purview. But it is generally assumed that most of
the other criticisms of Aristotle’s doctrine, which are mentioned and refuted by
Simplicius and the other commentators without their authors being named,
were taken from Nicostratus, and that his book was nothing less than a collection
of all possible objections to Aristotle’s teaching on this subject.134 Nicostratus for
his part used older material, including Stoic arguments which he sometimes
misunderstood.135 Like Athenodorus and Cornutus, Lucius and Nicostratus
complained that Aristotle’s list of Categories omitted some terms which should
have been included, and claimed that weight should be counted as a third sub-
division, with number and extension, of the category Quantity (this point was
also made by ps-Archytas).136 Like ps-Archytas again Lucius (and probably
Nicostratus also) argued that Quality should come before Relation, as it does in

133
Simpl. in Cat. 1,18ff; for other quotations see the index of the Berlin ed. s.v. See further, for
Nicostratus, K. Praechter, ‘Nikostratos der Platoniker’, Hermes 57, 1922, 481–517 = id., Kl. Schr.,
101–37; K. v. Fritz, ‘Nikostratos’ no. 26, RE XVII 1, 1936, 547–51; Dillon, 233–6. For Lucius, W.
Capelle ‘Lukios’ no. 1, RE XIII 2, 1927, 1791–7; Praechter, op. cit., 502 n. 1. Cf. L. Deitz, ‘Bibliographie
du platonisme impérial antérieur à Plotin: 1926–1986’, ANRW II 36.1.154.
134
Many of the arguments attributed to Lucius and/or Nicostratus by Simplicius are quoted
anonymously by the other commentators, Porphyry, Dexippus and Ammonius, and are also found
in Plotinus (Enn. 6.1); see Praechter, 501 (=121ff ). Porphyry seems to have made a point of refuting
all of them in his lost large commentary on the Categories; cf. Simpl. 21,4; 29,29 etc.
135
See Praechter, 488 (=108ff ) on Simpl. in Cat. 406,6ff.
136
Simpl. 62,28ff, 64,13ff, 128,5ff; ps-Archytas p. 25,2, 28,4 Th. For Athenodorus see above, p. 78.
Nicostratus, ap. Simpl. 368,12ff criticises Aristotle for not providing a discussion of the category
ekhein which would include all the uses of that word enumerated in the Postpraedicamenta
(15b17ff ).
88 Aristotle Transformed

Aristotle’s list of Categories at 1b26, but not in his detailed treatment in the body
of his work (ch. 7–8); here Simplicius admits that they have a case.137 They also
raised a problem which would have appeared fundamental to Platonists, whether
the Categories are found in the intelligible as well as the sensible world; this was
already considered by ps-Archytas and was later taken up by Plotinus, and it is
impossible to tell for certain how much Lucius and Nicostratus contributed to
the discussion.138 The other criticisms attributed to either are minor quibbles
about Aristotle’s use of particular words139 or other details. In one place
Nicostratus pointed out a discrepancy between the accounts of movement given
in the Categories and the Physics, showing that he was acquainted with this
work.140 Nevertheless it seems that his critique was entirely destructive and
neither he nor his predecessor made any positive suggestions; the best that can
be said for them is that they raised difficulties which it was useful to clear up
once and for all. But the trouble they took to refute Aristotle suggests that they
must have regarded his teaching as a serious threat to the Platonist position.

137
Simpl. 156,16ff; cf. ps-Archytas p. 23,21ff Th.; above, p. 77.
138
Simpl. 73,28ff, 76,14ff; cf. Plot. Enn. 6.1.1–2; ps-Archytas p. 22,31ff; 30,19ff Th., with Szlezák, ad loc.;
above, p. 76 (with n. 109).
139
e.g. poson, Simpl. 127,30; diathesis, 231,20.
140
Simpl. 428,3ff, comparing Cat. 15a13 with Phys. 225a25ff.
 4

The school of Alexander?*

Robert W. Sharples

Alexander of Aphrodisias was appointed by the emperors as a public teacher of

Aristotelian philosophy at some time between 198 and 209 ad.1 As a public
teacher it is likely that he had, in some sense, a school.2 But trying to establish
what happened in that school and how it functioned is comparable to the task
we would have if we had to establish what went on in a philosophy department
in a modern university on the basis of a selection of books by the professor and
a confused collection of his papers, the notes from which he lectured and the
essays of his students, with no obvious indication of which were which.
We know a considerable amount about the Neoplatonic schools of the fifth
and sixth centuries ad and the study of Aristotle’s writings in them; we know the
place they had in the curriculum, the order in which they were read, and we can
compare the ways in which different commentators approached the question of
the relation between the works of Aristotle and those of Plato. We can trace
relations between teachers and their pupils, and we are sometimes told that a
particular text is a pupil’s record of his teacher’s utterances. The very organisation
of the commentaries sometimes reflects and makes clear the requirements of the

* Full details of modern works referred to in the notes by author’s name (and in some cases date) only
are given in the References on p. 111. It will be immediately obvious how much the present
discussion owes, particularly where Alexander’s use of his predecessors is concerned, to the work of
Paul Moraux and in particular to the first two volumes of his study Der Aristotelismus bei den
Griechen (Moraux 1973 and 1984); extensive use has also been made of the word-searching facilities
afforded by the Thesaurus Linguae Graecae CD -ROM disc and the Ibycus Scholarly Computer
kindly loaned to the Institute of Classical Studies, London, by Professor David Packard.
1
cf. Todd 1976, 1 n. 3.
2
But not the Aristotelian Lyceum, which had probably ceased to exist in 86 bc ; cf. Lynch, 192–207.
We do not even know whether Alexander taught in Athens, though this seems likely and Todd 1976,
1 n. 2 notes the reference at in Metaph. 415,29–31 to Aristotle’s statue there.

89
90 Aristotle Transformed

teaching context, in the division of a commentary into separate lectures and the
placing of a general summary of a section of argument before the discussion of
particular points. For the medieval period, too, we have copious information on
the organisation of teaching and study.3
With Alexander matters are very different. We know the names of some of his
teachers,4 and his surviving works give us evidence for his disagreements with
them.5 We also know something of his disagreements with other philosophers of
his own generation or the generation before;6 and we can trace, however
controversially, his influence on later thinkers.7 But we do not know the name of
a single one of his immediate pupils,8 and for all that we can tell the influence of
other writers on him might have been largely, and his influence on other writers
entirely, through the medium of writing rather than personal encounter. We are
after all explicitly told that Alexander’s commentaries were among those read in
Plotinus’ school.9
It is however in principle unlikely that any thinker in the ancient world would
have communicated entirely through the written, rather than the spoken word.
And some of the writings attributed to Alexander are most naturally seen in the
context of his teaching activities or of debates within his circle.
These writings include commentaries on Aristotelian works, treatises or
monographs on particular topics such as those On the Soul10 and On Fate, and
also numerous short discussions.11 Three books of these collected discussions
are entitled phusikai skholikai aporiai kai luseis, ‘School-discussion problems and
solutions on nature’; a fourth is titled ‘Problems on Ethics’ but sub-titled, no

3
cf. especially the account in A. Kenny and J. Pinborg, ‘Medieval philosophical literature’, in
N. Kretzmann, A. Kenny and J. Pinborg (eds), The Cambridge History of Later Medieval Philosophy,
Cambridge 1982, 11–42.
4
Namely Herminus (c. 120–180 or 190 ad, Moraux 1984, 361–3; Alexander ap. Simplicium in Cael.
430,32ff ) and Sosigenes (similar in date; Alexander in Meteor. 143,13); possibly also Aristoteles of
Mytilene (died between 165 and 180 ad, Moraux 1984, 400; cf. Galen Scr. Min. vol. 2 (Müller) 11,4f;
Alexander Mant. 110,4; and below n. 42).
5
cf. below at nn. 20–47.
6
Below at nn. 58–89.
7
cf. Sharples 1987, 1220–4, with references there.
8
H.J. Blumenthal has pointed out that it is wrong to describe Alexander as the last Peripatetic – as
opposed to Neoplatonist – commentator on Aristotle (Chapter 5 below, and ‘Alexander of
Aphrodisias in the later Greek commentators on Aristotle’s De anima’, in J. Wiesner (ed.), Aristoteles
Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin 1987, 90–106, at 100. But Alexander does
seem, as far as our information goes, to stand at the end of a distinctively Peripatetic tradition. On
alleged Peripatetic teachers later than Alexander (Ammonius and Prosenes) cf. Lynch, 213.
9
Porphyry, Life of Plotinus, 14.
10
To be distinguished from Alexander’s commentary, now lost, on Aristotle’s de Anima.
11
The lost and surviving works attributed to Alexander are listed, with bibliography, at Sharples 1987,
1182–94. The authorship of many of the short texts is open to debate; in what follows I will on
occasion use ‘Alexander’ for convenience to refer to the authors of these texts without necessarily
thereby implying any view on the question of authenticity.
 The school of Alexander? 91

doubt in imitation of the preceding three books when it was united with them,12
skholikai êthikai aporiai kai luseis, ‘school-discussion problems and solutions on
ethics’.13 A further collection was transmitted as the second book of Alexander’s
treatise On the Soul, and labelled mantissa or ‘makeweight’ by the Berlin editor
Bruns. Other texts essentially similar to those in these collections survive in
Arabic though not in Greek; and there is evidence to suggest that there were
other collections now lost.14 The circumstances in which these collections were
put together are unclear; it was not always expertly done, and while some of
the titles attached to particular pieces seem to preserve valuable additional
information, others are inept or unhelpful.15 Nor is it clear at what date the
collections were assembled.16
It is not my concern here to give a full enumeration of the works attributed to
Alexander or to classify them in detail. That has been done elsewhere both by
myself and by others. Rather, I will proceed to a discussion of what the works can
tell us about the context in which they arose. It will be helpful to start with
consideration of the relation of Alexander’s works to those of his predecessors,
teachers and contemporaries.

II

Assessment of Alexander’s use of earlier Aristotelian commentators is rendered

more problematic by the fact that the only earlier commentaries which survive
entire are on the Nicomachean Ethics.17 While Alexander was certainly interested

12
So Bruns 1892, v.
13
I am currently preparing translations of the Quaestiones, Ethical Problems and Mantissa for
publication in the series of translations of the Greek commentators on Aristotle edited by Professor
Sorabji.
14
‘Scholia logica’ are referred to in what may be a gloss at Alexander in An. Pr. 250,2; and an
‘Explanation and summary of certain passages from (Aristotle’s) de Sensu’, which Moraux suggests
may have been a similar collection, is referred to by a scholion on quaestio 1.2. Cf. Moraux 1942, 24;
Sharples 1987, 1196.
15
cf. the discussion at Bruns 1892, xi.
16
Alexander’s commentary on the de Sensu cites not only the lost de Anima commentary (167,21) but
also a section of the Mantissa (in Sens. 31,29, citing Mantissa 127–130; cf. P. Wendland, preface to
CAG 3,1, v; Moraux 1978, 297 n. 71), cf. also below, n. 98.
17
Namely the commentary on the Ethics by Aspasius (CAG vol. 19) and the anonymous scholia on
books 2–5 published in CAG vol. 20, which incorporate material by Adrastus of Aphrodisias; cf. P.
Mercken, The Greek Commentaries on the Nicomachean Ethics of Aristotle, Leiden 1973 (Corp. Lat.
Comm. in Arist. Graec. 6.1) 1.14*–22*, and Moraux 1984, 323–30. A. Kenny, The Aristotelian Ethics,
Oxford 1978, 37 n. 3, suggests that the commentary itself is by Adrastus, but this is difficult to
reconcile with the ancient evidence for Adrastus’ work; cf. H.B. Gottschalk, ‘Aristotelian philosophy
in the Roman World’ in W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, Teil II
‘Principat’, Bd. 36.2 ‘Philosophie und Wissenschaften’, Berlin 1987, 1079–174, at 1155 and n. 363.
92 Aristotle Transformed

in that work, as the collection of Ethical Problems shows,18 it is not clear whether
he wrote a complete commentary on it or not.19
Not surprisingly, Alexander refers in several places to his own teachers.
According to Philoponus,20 Alexander in his (lost) monograph On the
disagreement between Aristotle and his followers concerning syllogisms with mixed
premisses21 referred to Sosigenes’ interpretation of Aristotle, Prior Analytics, 1.9,
30a15–23. In his commentary on the Meteorology Alexander cites Sosigenes on
the explanation of the solar and lunar halo, agreeing with Sosigenes in supporting
the position of Aristotle and Posidonius against rival theories.22 Sosigenes’
general interest in the theory of vision is reflected in Alexander’s works,23 and
Moraux has suggested that Alexander was Themistius’ source for Sosigenes’
explanation of phosphoresence.24

18
The Ethical Problems are admittedly chiefly interested in the solution of difficulties in the application
of logical distinctions to ethical subject-matter (for example, the way in which pleasure and pain are
opposites; see below at n. 167, and for another example below at nn. 39–41) and in topics (such as
responsibility for actions) which Alexander dealt with in independent treatises. On the general
character of the Ethical Problems, cf. A.J. Madigan, ‘Alexander of Aphrodisias: the book of Ethical
Problems’, in W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, vol. II .36.2, Berlin 1987.
19
That he did so is suggested by a reference to ‘hupomnêmata on the Ethics’ at Alexander in Top.
187,9–10; but this is the only evidence. Cf. Sharples 1985, at 113.
20
Philoponus in An. Pr. 126,20–3. Cf. also [Ammonius] in An. Pr. 39,24–6; Moraux 1984, 339–44.
Below, n. 25.
21
That is, on syllogisms with mixed modal premisses. Cf. Alexander in An. Pr. 125,30ff; Moraux 1984,
393.
22
Alexander in Meteor. 143,13; Moraux 1984, 335. Sosigenes’ opinion is cited, it may be noted, not
from a commentary but from his ‘eighth book On Sight’.
23
cf. Alexander de Anima 42,4–46,19; Mantissa 127,20–150,18; Quaestiones 1.2 and 1.13. At de Anima
43,16 Alexander refers to ‘enquiries about how we see’, but the points he there mentions cannot all
be related to the Mantissa discussion. Cf. Bruns 1887 ad loc.
24
Themistius in DA 61,22ff, from Sosigenes’ third book On Sight; cf. Philoponus in DA 348,10–19.
Moraux 1984, 359. – When Simplicius in his de Caelo commentary (503,35) cites Alexander and
Porphyry as concerned about a passage of Aristotle where Sosigenes preferred to suppose an error
by a copyist rather than a confusion by Aristotle himself, it might seem reasonable to suppose that
Alexander’s (lost) de Caelo commentary, which Simplicius cites extensively, discussed Sosigenes’
view and was Simplicius’ source for it. However, Th.-H. Martin, ‘Questions connexes sur deux
Sosigène et sur deux péripatéticiens Alexandre’, Annales de la Faculté de lettres de Bordeaux 1, 1879,
174–87, at 176–7, argues that Simplicius, who cites Sosigenes only on the topic of the heavenly
spheres, is drawing directly on a treatise by Sosigenes On the Counteracting Spheres (cf. Proclus
Hypotyp. (Manitius) 4.98, 130,7–23; Moraux 1984, 344–58). It is for the same point, it may be noted,
that pseudo-Alexander in Metaph. cites Sosigenes at CAG vol.1 706,13; it could be the genuine
commentary by Alexander that was the pseudo-Alexander’s source, but it seems equally likely that
pseudo-Alexander derived the information from Simplicius. (Cf., on another similar case, P. Merlan,
‘Ein Simplikios-Zitat bei ps.-Alexandros und ein Plotinus-Zitat bei Simplikios’, Rh. Mus. 89, 1935,
154–60; but also, challenging the generally accepted late dating of the section of the Metaphysics
commentary not by Alexander himself, and its attribution to Michael of Ephesus, L. Tarán at
Gnomon 53, 1981, 750, reviewing Moraux 1973). Elsewhere the pseudo-Alexander uses Sosigenes’
name simply as an example (466,17, 663,2); this may be an attempt to create an impression of
authenticity, as with his use of Alexander himself as an example (cf. M. Hayduck, preface to CAG
vol.1, vi, and the references in his index). – The Sosigenes referred to at Alexander Mixt. 3 216,12 is
a Stoic, not Alexander’s teacher; cf. Martin, op. cit., 177–8.
 The school of Alexander? 93

Alexander argued against his teacher Herminus in the monograph on mixed

modal syllogisms and similarly in the commentary on the Prior Analytics, without
naming him in the latter.25 In his de Caelo commentary, too, Alexander disagreed
with Herminus’ views on the parts played by the nature of the heavenly body on
the one hand, and its soul on the other, in explaining the heavenly motion.26
Simplicius quotes a passage in which Alexander states that he heard Herminus’
explanation of Aristotle’s claim that acceleration of the heavens would imply
unnatural retardation.27 Simplicius himself remarks that Alexander was the first
to understand Aristotle’s argument here correctly,28 and it seems likely that here
too Alexander developed his view in conscious opposition to that of Herminus.
Simplicius refers to Alexander’s statement of Herminus’ position as ‘in
accordance with Aspasius’, and concludes the citation with ‘this is what Herminus
said in accordance with Aspasius’.29 Aspasius was earlier than Herminus;30
Simplicius’ remarks indicate that he took the statement of Herminus’ view from
Alexander – specifically from the lost de Caelo commentary – and that Alexander,
in stating his teacher’s view, remarked on its similarity to that of Aspasius.31
Aspasius and Alexander are cited together several times in Simplicius’ Physics
commentary, sometimes as agreeing,32 sometimes as disagreeing,33 and once
Simplicius tells us that Alexander knew of the reading of the Aristotelian text
adopted by Aspasius, but preferred another.34 It seems likely that Simplicius is

25
The view reported from Herminus by pseudo-Ammonius in An. Pr. 39,31 corresponds to that
criticised by Alexander at in An. Pr. 125,3–6. (While Alexander does not there name Herminus, he
does call his view ‘ridiculous’, 125,16). Moraux 1984, 391–4 notes the connection between the two
passages and suggests that pseudo-Ammonius is using Alexander’s monograph, rather than the
commentary, as his source. Herminus is also referred to at Alexander in An. Pr. 72,27; 89,34; 91,2; in
Top. 569,3; 574,26.
26
For Alexander the nature and the soul of the heavens are identical. Cf. Simplicius in Cael. 380,5,
who notes that Alexander replies to the views of Herminus but not to those of Julianus of Tralles;
Merlan, op. cit. in n. 24, and id., ‘Plotinus Enneads 2.2’, TAPhA 74, 1943, 179–91; Sharples 1987, 1214
n. 153, and references there.
27
Aristotle Cael. 2.6, 288b22; Simplicius in Cael. 430,32ff.
28
Simplicius in Cael. 430,29. At in Int. (ed. sec.) 272,14ff Boethius reports Alexander as supposing a
textual corruption in Aristotle Int. 10,19b29–30, but argues (274,12) that Alexander was wrong and
that the interpretation of Herminus and Porphyry is to be preferred.
29
Simplicius in Cael. 430,32–431,11; cf. Moraux 1984, 240ff.
30
Galen, who was born in 129, was taught by one of Aspasius’ pupils (Galen Scripta Minora
(Marquardt) vol. 1 32,5–7). Aspasius’ philosophical activity is to be placed in the first half of the
second century ad : Moraux 1984, 226.
31
So Moraux 1984, 361 n. 5. The alternative would be to interpret Simplicius as saying that Alexander
cited Herminus’ views according to Aspasius, that is that Aspasius was his source; but even without
the chronological difficulty it would be odd if Alexander had to use Aspasius as a source for the
views of his, Alexander’s, own teacher.
32
Simplicius in Phys. 547,11; 727,35; 752,15.
33
Simplicius in Phys. 131,15ff (where Alexander argued against Aspasius at length; cf. Moraux 1984,
235–6); 558,34; 728,5.
34
Simplicius in Phys. 422,25.
94 Aristotle Transformed

throughout citing Aspasius’ commentary from quotations in Alexander’s.35

Aspasius is cited in Alexander’s Metaphysics commentary36 and in that on the de
Sensu;37 and it seems that Alexander also cited Aspasius extensively in his de
Interpretatione commentary.38
One of the short texts attributed to Alexander, Ethical Problem 11, adopts a
similar solution to Aspasius’, though without referring to him by name, to the
problem that ‘involuntary’ appears to have two senses while its opposite
‘voluntary’ has only one.39 The anonymous commentary on Nicomachean Ethics
Book 3, which has been linked with Adrastus of Aphrodisias, adopts a different
view; unfortunately it is not clear whether Alexander in the Ethical Problem is
replying to Adrastus, or whether this particular material in the anonymous
Ethics commentary is later and is drawing in a confused way on the Ethical
Problem.40 Similar issues are discussed in two passages of Alexander’s Topics
commentary, again without explicit reference to any of his predecessors.41
Alexander’s third teacher apart from Sosigenes and Herminus, Aristoteles of
Mytilene (if Moraux’s view is accepted42) is referred to at Mantissa 110,4, a
section of a text which may or may not be by Alexander himself.43 It seems likely
that both the general account of Aristotle’s doctrine of intellect entering human
beings ‘from outside’ (nous thurathen) at 110,4–112,5, and the following reply to
the objection that intellect cannot be said to ‘come from outside’ if it is incorporeal

35
So Diels in the index nominum to CAG vols 9–10 s.v. ‘Aspasios’, citing Simplicius in Phys. 131,14;
547,11; 558,34 (the last passage is particularly noteworthy; Simplicius says that he thinks that
Aspasius’ position was as he described it, and immediately goes on to cite Alexander). On the other
hand, if all Simplicius’ references to Aspasius by name do come from Alexander, we would have to
suppose that Alexander cited Aspasius (and other predecessors) by name in his lost Physics
commentary rather more than is his practice in the extant commentaries; he is much more sparing
in his references to predecessors by name there than is Simplicius. (The references to Theophrastus
and Eudemus in the Prior Analytics commentary, below nn. 48–9, are perhaps a special case).
Aspasius is referred to in about 20 separate contexts in Simplicius’ Physics commentary, but only
three times in Alexander’s Metaphysics commentary, a difference which cannot entirely be explained
by difference in length between the two works. (Alexander is regularly mentioned in close proximity
to the references to Aspasius in Simplicius’ Physics commentary; there are exceptions, but failure to
mention Alexander cannot prove that he was not the source.)
36
cf. Aspasius ap. Alexander in Metaph. 41,27ff and 379,3ff; in the latter passage Alexander disagrees
with Aspasius, though not explicitly. Cf. Moraux 1984, 246–9; also below n. 62.
37
Alexander, in Sens. 10,2 (disagreement); also (without Aspasius being named explicitly) ibid. 82,16–
17. Moraux 1984, 244–5.
38
So Moraux 1984, 231.
39
Aspasius in EN 59,2–11. Sharples 1985, 110.
40
Anon. in EN (CAG vol. 20) 141,10–20. Sharples (1985) 111. Cf. also Alexander, P. Eth. 27 154,15–
155,5 and 155,29–156,9 with Anon. in EN, CAG, vol. 20, 133,35–134,4 and 134,9–21 respectively.
41
Alexander in Top. 99,2–20; 100,31–101,14; cf. also 181,21–187,9. Sharples 1985, 112.
42
cf. P. Moraux, ‘Aristoteles, Der Lehrer Alexanders von Aphrodisias’, AGPh 49, 1967, 169–82, and id.
1984, 399–401; contra, P. Thillet, Alexandre d’Aphrodise: Traité du Destin, Paris, Budé, 1984, xv–xix.
43
cf. Sharples 1984, 1213–14 and references there.
 The school of Alexander? 95

and hence cannot move (112,5–113,12),44 are to be interpreted as accounts of

Aristotle’s views; it is the latter section that the author of the Mantissa passage
proceeds to attack. Both he and Aristoteles of Mytilene accept as given the
identification of the ‘intellect from outside’ of Aristotle’s Generation of Animals
with the active intellect of de Anima 3.5.45 Alexander also cited Aristoteles of
Mytilene’s views on de Caelo 1.446 and on Metaphysics Book 2, in the latter case
referring to Aristoteles’ interpretation as ‘rather dialectical, though relevant’.47
Of earlier Peripatetics, Alexander refers extensively to Theophrastus48 and
Eudemus.49 He argues against Andronicus for the authenticity of the de
Interpretatione50 and also against Sotion51 and Xenarchus;52 he used the
commentaries of Boethus, apparently,53 and of his namesake Alexander of

44
The objectors are referred to in the third person (epipherousin, 112,6), but not further identified;
however, see below at n. 68. The attribution of both 110,4–112,5 and 112,5–113,12 to the same
author has been disputed; cf. Sharples, op. cit., 1212 and references in nn. 134–6 there.
45
Sharples 1987, 1212 and references in nn. 137–9 there. I am grateful to Victor Caston for emphasising
the importance of this point. At Quaest. 2.3 48,18–22 it is taken as established (ekeito) that divine
providence, acting through the influence of the heavenly bodies on the sublunary, is the cause of
some creatures’ possessing more soul-faculties than others, and in particular of man’s possessing
reason. Moraux 1967, at 163–4 n. 2, compares this passage to Mant. 113,6–12, and argues from this
that Quaest. 2.3 is an early work of Alexander’s.
46
Alexander ap. Simplicium in Cael. 153,16ff; Moraux 1984, 401–2.
47
Alexander in Metaph. 166–19–167,3, referring to ‘our Aristotle’ by contrast with the Stagirite; noted
by P. Accattino, ‘Alessandro di Afrodisia e Aristotele di Mitilene’, Elenchos 6, 1985, 67–74, at 73–4.
48
There are about 50 references to Theophrastus by name in the extant works of Alexander, about half
of these being in the Prior Analytics commentary.
49
There are 11 references to Eudemus in Alexander’s extant works, again mainly from the Prior
Analytics commentary (7 of the total). Simplicius in Phys. 1355,32 cites Alexander as arguing
against Eudemus on the location within the heavenly sphere of the prime mover; cf. 1354,9ff.
Alexander is also referred to as citing Eudemus by Simplicius in Phys. 133,21ff; Simplicius says that
he cannot himself find in Eudemus’ commentary the reading of Aristotle cited by Alexander (ibid.,
133,24–5). Simplicius notes that Alexander seems to be following Eudemus at in Phys. 10,13 (cf.
11,16); a reaction by Alexander to Eudemus is noted at 99,29.
50
Alexander in An. Pr. 161,1.
51
Alexander in Top. 434,2, on whether sleep is a privation.
52
Alexander is cited as replying to Xenarchus by Simplicius, in Cael. 21,33ff (cf. 22,18) and 23,22ff; cf.
Moraux 1973, 199f. Alexander also referred to Xenarchus in his de Anima commentary (Philoponus
in DA (Verbeke) 15,65–9; Moraux 1973, 207). Moraux 1942, 211 n. (e) compares with this passage
Mant. 106,20–3; here too Alexander argues against a literal understanding of material intellect as
matter, though without here naming Xenarchus. The interpretation of the associates of Xenarchus
and Boethus on what is the prôton oikeion according to Aristotle are referred to and criticised at
Mantissa 151,3ff, as are those of Sosicrates and Verginius Rufus; cf. Moraux 1973, 209, and other
references at Sharples 1987, 1190.
53
At in Phys., 211,13ff Simplicius quotes from Alexander, then from Boethus, and then again from
Alexander (below n. 117); cf. Moraux 1973, 170 and n. 1, arguing that Alexander is Simplicius’
source. Alexander is also cited by Simplicius in Phys. 759,20 as replying to Boethus’ claim that
movement could be numerable even in the absence of soul; the issue is one that is also discussed in
the treatise On Time by Alexander preserved in Arabic, and comparison with that suggests that
Simplicius in his Physics commentary may have taken more from Alexander here than he explicitly
acknowledges. Cf. R.W. Sharples, ‘Alexander of Aphrodisias, On Time’, Phronesis, 27 1982, 58–81, at
70–1 and nn. 75–7.
96 Aristotle Transformed

Aegae,54 and also seems to have reported views of Nicolaus of Damascus.55 He

referred to the eclectic Potamo;56 and there is evidence that he knew and made
use of the pseudo-Aristotelian de Mundo.57
Alexander’s interest in the views of earlier and near-contemporary
philosophers was not however confined to Peripatetics. His controversies with
Galen58 and the extent to which he opposed Stoicism, while nevertheless being
influenced by the Stoics both in language and in thought,59 are well known. But
his writings show interest in Platonism and Epicureanism too. It is not, indeed,
surprising that he should discuss Plato’s Theory of Forms in commenting on
Aristotle’s criticisms of it in the Metaphysics; indeed his Metaphysics commentary
is the major source of evidence for Aristotle’s lost de Ideis.60 His arguments
against Xenocrates on the question of the priority of genus to species fall into the
same general area,61 as does his preserving the information, from Aspasius, that
the first-century ad Platonist Eudorus altered the text of Aristotle.62
In quaestio 2.13 Alexander argues that the triangles in Plato’s Timaeus have a
material, rather than a formal role.63 There is no explicit reference by name to

54
At in Cat. 10,19–20 Simplicius cites Alexander of Aphrodisias and then says that Alexander of
Aegae said the same; at in Cat. 13,11–18 he refers to the view of ‘the Alexanders’ (Moraux 1984,
222f). Cf. also Simplicius in Cael. 430,32, stating that Alexander of Aphrodisias cited Alexander of
Aegae on Aristotle’s argument at de Caelo 2.6, 288b22; above, n. 27, and Moraux 1984, 223–5.
55
At in Phys. 23,14–16 Simplicius successively cites interpretations of Xenophanes by Nicolaus and by
Alexander before proceeding to argue against them both; Moraux suggests that Alexander in his
commentary argued against Nicolaus, and that it was through his commentary that Simplicius
knew of Alexander’s views. Cf. Moraux 1973, 451–7, and J.B. McDiarmid, ‘Theophrastus on the
Presocratic causes’, Harvard Studies in Classical Philology 61, 1953, 85–156, at 115–20.
56
Alexander ap. Simplicium in Cael. 652,9ff.
57
So Moraux 1967, 160 n. 2, comparing Alexander ap. Simplicium in Phys. 310,25–311,37 with
[Aristotle] de Mundo 398b16ff.
58
cf. Sharples 1987, 1179 and nn. 18–21, with references there; also now V. Nutton, ‘Galen’s
philosophical testament: “On my own Opinions” ’, in J. Wiesner (ed.), Aristoteles: Werk und Wirkung,
Paul Moraux gewidmet, vol. 2, Berlin 1987, 27–51, at 45–51.
59
cf. Sharples 1987, 1178 and nn. 11–13, with references there. Simplicius in Phys. 671,4 reports
Alexander as commenting that an argument under discussion can also be used against the Stoics;
cf. also below n. 98.
60
Alexander in Metaph. 79,3–98,24; cf. W. Leszl and D. Harlfinger, Il ‘De Ideis’ di Aristotele e la teoria
platonica delle idee, Florence 1975 (Accademia toscana di scienze e lettere La Colombaria, studi, 40).
61
Below, n. 139.
62
Alexander in Metaph. 58,31–59,8, on Metaph. 1.6, 988a10–11. H. Dörrie, ‘Der Platoniker Eudoros
von Alexandria’, Hermes 79, 1944, 25–39, at 34f, 38f, and Dillon, 116 and 128 n. 1. It has generally
been thought that Eudorus changed the text to support his own version of Platonism, deriving
matter from the One; but Moraux argues against this in ‘Eine Korrektur des Mittelplatonikers
Eudoros zum Text der Metaphysik des Aristoteles’, in Beiträge zur alten Geschichte und deren
Nachleben: Festschrift für Franz Altheim, Berlin 1969, vol. 1, 492–504.
63
cf. Plato, Timaeus 53Cff. Alexander is cited by Simplicius in Phys. 26,13, on the reference in Aristotle
Phys. 1.2., 184b19 to those who postulated three principles, as identifying Plato’s first principles as
matter, the efficient cause and the exemplar; Simplicius criticises Alexander for failing to take
account of Plato’s final cause, and himself in this context downgrades matter to an auxiliary cause
(26,5ff ).
 The school of Alexander? 97

Plato or his followers, or to the Timaeus, in the text of the quaestio; the title
describes the text as ‘against the Platonists who say that it is the shapes and forms
of bodies that are composed of the triangles, not the bodies themselves’. The
interpretation advanced by the author of the quaestio, implying that it is the
Receptacle that plays the role of matter in the Timaeus, is like that of Aristotle;64
it would be difficult to interpret the third-person plural references in the text of
the quaestio (as opposed to the title) as referring solely to non-Platonic
interpreters of the Timaeus, but the discussion may have as much to do with
Peripatetic doubts about how to interpret Plato as with Platonist exegesis.
Quaestio 2.21 is concerned, as Merlan showed,65 to defend an Aristotelian
theory of providence against Platonist critics, and notably the middle-Platonist
Atticus, who themselves maintain that the sublunary is a primary concern of the
divine, and criticise the Peripatetics for denying divine providence or making it
entirely accidental; though in attacking the notion that the sublunary is a primary
concern of the divine Alexander takes the opportunity to attack the Stoics too.66
At 66,23 there may be an ironic allusion to the Craftsman of Plato’s Timaeus as
lacking concern for the world himself; this is hardly fair, for as fr. 2 Vitelli
(discussed below) shows, Alexander was perfectly well aware that the Craftsman
of the Timaeus delegates much of his concern for the worlds to lesser gods.
Atticus may however not have emphasised this point in the way other
contemporary Platonists did.67 Alexander’s, or his predecessors’, views on intellect
also seem to involve a reaction to Atticus;68 and Alexander, in attacking Plato’s
view that the universe is perishable in its own nature but preserved by the divine
will,69 was upholding an Aristotelian position which Atticus had attacked.70
Alexander appeals in support of his own view to the principle that what is
impossible is impossible even for the gods, and cites Plato, Theaetetus 176A in

64
Aristotle, Physics 4.2, 209b11ff; cf. F.M. Cornford, Plato’s Cosmology, London 1937, 181 and 187.
65
Merlan 1967, at 90–1. It is true that the issue could arise within a purely Peripatetic context; that the
influence of the heavens on the sublunary may be purely accidental is already suggested by
Theophrastus, Metaphysics 5b19–26.
66
cf. Quaest. 2.21 70,2–6, against those who say that virtue alone is good, with Alexander Fat. 199,14ff.
67
On Atticus’ treatment of divine providence cf. Dillon, 252.
68
So P.L. Donini, Tre studi sull’ Aristotelismo nel II secolo d.C., Turin 1974, 49–50, comparing Atticus
fr. 7,75 des Places.
69
Alexander ap. Simplicium in Cael. 358,27ff, and Alexander Quaest. 1.18; cf. M. Baltes, Die
Weltentstehung des platonischen Timaios nach den antiken Interpreten, Leiden 1976, 76–81; Sharples
1983, at 99–102 and nn. Cf. also Alexander in Metaph. 212,15.
70
Atticus, fr. 4,8–17 des Places. On the issue cf. R.R.K. Sorabji, Time, Creation and the Continuum,
London 1983, 304 n. 47 and references there. Alexander agrees with Atticus that Plato in the
Timaeus held the ordered cosmos to have a real beginning in time, and attacks Taurus on the issue;
cf. Alexander ap. Simplicium in Cael. 297,9ff and ap. Philoponum contra Proclum (Rabe) 213–16
(from Alexander’s de Caelo commentary, ibid., 212,14); Baltes op. cit., 71–6, Dillon, 242–4 and 253–
4, and K. Praechter, ‘Tauros’, RE 2 Reihe 5.1, 1934, 68.
98 Aristotle Transformed

support of this. Other Platonic works referred to by Alexander – often, again, in

discussions prompted by Aristotle’s references to Plato – include the Sophist,
Timaeus, Laws 10 and the Second Epistle.71
In fr. 2 Vitelli, a short text preserved as part of a miscellany of extracts from
Alexander in a Florentine MS ,72 Alexander replies to a Stoic philosopher who has
criticised Aristotle for disagreeing with Plato on the immortality of the soul and
divine providence. Exactly the same criticisms had been made by Atticus. Against
the Stoic, Alexander argues that Aristotle did at least make the soul incorporeal,
like Plato, even if not imperishable, while for the Stoics it is neither,73 and also that
Plato’s view of divine providence is more acceptable than Stoic pantheism, because
it does not involve the supreme god directly in mundane affairs (a feature of Stoic
doctrine that he objects to elsewhere too74). In support of the latter point he
alludes, without naming the works in question, to Timaeus 42E and Politicus 272E.
Alexander’s Stoic opponent is not named in the text of the fragment itself.
There is however a prefatory note in the MS , which begins as follows.

By the same Alexander, from the argument against Heraclides; a consideration

of what was said by Aristotle about the fifth element. In this (Alexander) argues
against a certain Stoic philosopher . . .

At first sight the natural way to interpret this is that, in the course of an argument
against someone named Heraclides75 on the topic of the fifth heavenly element,

71
There are numerous references to Plato in Alexander’s Metaphysics commentary (cf. Hayduck’s
index); notable is in Metaph. 59,28ff, citing Timaeus 28C and the Second Epistle, 312E. Plato is also
cited by Alexander ap. Simplicium in Phys. 355,13 (citing Laws 10); 420,13; 420,18; 454,19 (Plato’s
On the good); 700,19; 705,5; 894,12; 1351,27; by Alexander ap. Simplicium in Cael. 276,14 (referring
to Timaeus 31AB , a passage already discussed by Theophrastus, ap. Proclum in Timaeum (Diehl)
1.456.16–18); and by Alexander ap. Philoponum in Phys. 81,25 (on the Sophist as evidence for
Parmenides; cf. Alexander ap. Simplicium in Phys. 135,16; 136,10).
72
Noted by G. Vitelli, ‘Frammenti di Alessandro di Afrodisia nel cod. Riccard. 63’, Studi Italiani di
Filologia Classica 3, 1895, 379–91, an elaboration of his account in ‘Indice de’ codici greci Riccardiani,
Magliabechiani e Marucelliani’, ibid., 2, 1894, at 515. The fragment was then published by Vitelli in
‘Due frammenti di Alessandro di Afrodisia’, Festschrift Theodor Gomperz, Vienna 1902, 90–3.
(Citations according to this article.) The miscellany includes another otherwise unknown text on
the sôrites argument and a number of extracts from Alexander’s de Anima with one passage each
from the de Mixtione and the Quaestiones. On the value of the MS for the de Mixtione, cf.
E. Montanari, ‘Per un’edizione del Peri kraseos di Alessandro di Afrodisia’, Atti e memorie dell’
Accademia toscana di scienze e lettere La Colombaria 36, 1971, 17–58.
73
‘(Plato) also says that the soul is an incorporeal substance and that it is imperishable; and one of
these points about it is also made by Aristotle’ (93,13–15; my emphasis).
74
Alexander, de Prov. 19,3ff; 25,1ff; 31,16; 53,1ff; 63,7ff in H.-J. Ruland, Die arabischen Fassungen
zweier Schriften des Alexander von Aphrodisias, diss. Saarbrücken 1976; Quaest. 2.21 68,9ff; Mant.
113,12ff; Mixt. 226,24ff. Cf. [Aristotle] de Mundo 6, 397b20ff, 398a5ff; Todd 1976, 226–7.
75
One might suppose that the reference is to Heraclides of Pontus. However, Heraclides’ views on the
fifth element are unclear (cf. H.B. Gottschalk, Heraclides of Pontus, Oxford 1980, 106–7); and,
significantly, he is not cited on this topic in Simplicius’ de Caelo commentary, which draws
extensively on the commentary by Alexander now lost.
 The school of Alexander? 99

Alexander digressed in order to reply to the attacks of an unnamed Stoic. The

argument against Heraclides would have to be either an independent treatise of
Alexander’s of which no other trace has survived, or else an excursus in one of
his commentaries – probably that on the de Caelo. Such an excursus would have
to be on a very considerable scale for there to be room within it for a further
digression in reply to the Stoic; and this makes it all the more odd that no trace
of it has survived in Simplicius’ de Caelo commentary, which cites Alexander’s
extensively.76 (The fragment amounts to nearly a page of printed Greek text, and
appears to refer back at its outset to preceding exposition of the Stoic’s arguments).
The alternative is to suppose that the ‘Heraclides’ in question is in fact the
Stoic referred to in the second sentence of the introduction and in the text itself;
for a Stoic might naturally attack the Aristotelian theory of the fifth element.77
This possibility is the more interesting because in the text itself Alexander says
that his adversary is ‘leader’ (proïstatai) of the Stoic school (p. 93,3).78 The term
can indeed just mean ‘champion’, and is so used later in this text (p. 93,6); but it
is the same word proïstamai that Alexander uses in speaking of his own
appointment as a public teacher of Peripatetic philosophy.79 Conceivably
therefore in Heraclides we have the name of a contemporary holder of the
corresponding Stoic position.
It may, moreover, be possible to identify the Stoic in question. Aurelius
Heraclides Eupyrides is referred to as a Stoic teacher in one, and possibly two,
second-century Athenian inscriptions.80 Lynch dates him to the reign of Hadrian

76
It may be instructive to compare the treatment of Alexander’s two digressions in commenting on
An. Pr. 1.15, 34a10–12, at 177,19–182,8 (on Chrysippus’ claim against Diodorus Cronus – and
Aristotle – that the impossible can follow from the possible) and on 34a12–16, at 183,34–184,18 (on
the Philonian and Diodorean definitions of the possible; cf. below, n. 98). The content of the latter
digression is very briefly summarised by Philoponus in An. Pr. 169,17–23; it does not appear in
[Ammonius] or [Themistius]’ Prior Analytics commentaries, though similar distinctions are drawn
in Simplicius’ commentary on the Categories (195,31ff. cf. Sharples 1982, 93–4). The topic of the
former digression is at least mentioned in all the later commentaries, albeit much more briefly than
in Alexander (cf. [Ammonius] in An. Pr. 50,13–21; [Themistius] in An. Pr. 26,31–29,28; Philoponus
in An. Pr. 165,27–167,30).
77
Atticus too wrote on this; fr. 5 des Places. It is true that even if Heraclides is identified with the Stoic
we would still be faced either with an independent treatise of which no other record has survived
or with a lengthy excursus in a commentary which has similarly left no other traces; but at least in
the later case it would not be a matter of a digression within an excursus.
78
The rhetorical ploy in ‘it is surprising that someone who claims to be a philosopher is not ashamed
(sc. of arguing like this)’ is reminiscent of some of Alexander’s remarks against those whose views
he attacks in the de Fato; cf. especially 171,22ff.
79
Fat. 164,14.
80
IG II 2 3801, and also perhaps IG II 2 3989. S. Follet, Athènes au II e et au III e siècle après J.-C.,
Paris 1976, 88 points out that the apparent peregrine status of the dedicator Symmachus in IG II 2
3801 places that inscription before the granting of universal citizenship in 212. I am grateful to
Dr A.J.S. Spawforth for drawing my attention to Follet’s discussion.
100 Aristotle Transformed

(117–138),81 but this may be much too early. The name Aurelius suggests that
this Heraclides received Roman citizenship not before the reign of Marcus
Aurelius (161–180),82 and indeed the receipt of the citizenship may have been
the occasion of the setting up of one or both of the honorific inscriptions. The
dedicator of the first inscription was a certain Symmachus of Phlya who set up
the inscription to Heraclides as his teacher and who may have been the father of
two youths who were ephebes between 182 and 191.83 That suggests that
Symmachus was born in around 135, and that he may have been a pupil of
Heraclides in about 155–160;84 Heraclides himself might then have been born in
about 115. This would make him approximately the same age as Alexander’s
teachers.85 If Alexander was born in about 145–50 it does not seem inconceivable
that he might have written our text in about 180–185 while Heraclides was still
the principal Stoic teacher; its tone is hardly deferential from a young man to one
30 years or more his senior, but it is not clear that Alexander would feel any
obligation to be deferential towards a member of a rival school.
Quaestio 1.13 is a criticism of a claim that Epicurus’ account of colour is
similar to that of the other schools (sects, haireseis). The person who had claimed
this is named in the text only as ‘someone’, tis, but the title supplied by the editor
– who presumably had independent information86 – identifies him as ‘Censorinus
the Academic’, apparently not the author of de Die natali, who lived in the first
half of the third century ad, but an otherwise unknown philosopher.87 Von
Arnim suggests that he is to be dated to the second half of the second century,
that is to the time of Alexander or the generation before;88 if ‘Academic’ (as
opposed to ‘Platonist’) indicates an adherent of the sceptical Academy, as it

81
I am grateful to Dr Spawforth for pointing this out.
82
Lynch, 190. Ueberweg-Praechter, Grundriss der griechischen Philosophie (12th ed., 665) dates our
Heraclides rather to the middle of the second century ad. A terminus ante quem may perhaps
be given by another Stoic, Julius Zosimianus, whose funerary inscription is preserved in IG II 2
11551 and whom both Ueberweg-Praechter and Lynch (loc. cit.) agree in putting later than
Heraclides; he seems to have died at the end of the second century ad. Two other philosophers
named Heraclides can be excluded: Heraclides (40) in RE is a Stoic of the second century bc ,
but Alexander refers to his adversary in the present tense; Heraclides (41) is a Pyrrhonist of the
first century bc .
83
The suggestion is made by the editors of IG II 2 ad loc. (IG II 2 2111/2). Follet 81 dates the inscription
more precisely to 185/6.
84
It would be natural, as Dr Spawforth points out, to suppose that Symmachus put up the dedication
to his teacher when he ceased to be a pupil, or soon after; however, he might have done so years after
he had ceased to be a pupil, to mark the grant of citizenship or (perhaps more plausibly) Heraclides’
becoming diadokhos, if he was not already so when he taught Symmachus.
85
cf. above, n. 4.
86
See above, n. 15, and Bruns’ discussion cited there which draws attention to the present title.
87
No. 8 in Pauly-Wissowa, RE 3.2, 1899, col. 1910.
88
In RE loc. cit. But this is simply conjecture based on the present text.
 The school of Alexander? 101

regularly does by Alexander’s time, the lateness of this Censorinus’ date is

striking.89
Alexander’s discussions of pleasure, a major concern of the Ethical Problems,
are concerned with problems raised by Aristotelian doctrine rather than with
specifically anti-Epicurean polemic.90 Alexander does however criticise the
Epicurean view that the universe is infinite;91 in his commentary on the Physics,
or perhaps in an independent treatise now lost, he argued against the views of an
Epicurean Zenobius on the related question whether an infinite body can be in
place.92
Other writers referred to by Alexander include Aratus, whose views on which
heavenly pole is uppermost he criticised;93 Euclid, whose first theorem is not, he
argued, incompatible with the Aristotelian theory of a finite universe;94
Hipparchus, against whom he defended the Aristotelian account of the
acceleration of falling bodies;95 and Geminus, whose epitome of Posidonius’
Meteorology he cited to illustrate the difference between physical explanation
and astronomical hypothesis.96

III

Alexander’s commentaries are not formally arranged, as the Alexandrian

Neoplatonist ones were to be, on the basis of a general discussion of each section
(theôria) followed by comments on details of the text (lexeis); he proceeds
through Aristotle’s text paragraph by paragraph and often sentence by sentence.97

89
The last major figures to whom the term ‘Academic’ seems to be applied are Plutarch and Favorinus;
cf. H. Tarrant, Scepticism or Platonism: the history of the Fourth Academy, Cambridge 1985, 97 and
129–35, id. at Class. Quart. 33, 1983, 182–3, and J. Glucker, Antiochus and the Late Academy,
Göttingen 1978 (Hypomnemata 56), 213–15, 281.
90
cf. the article by Madigan cited in n. 18 above.
91
Quaest. 3.12. This text upholds the Peripatetic view, that the universe is finite with nothing outside
it, both against the Epicurean view that the universe is infinite and against the Stoic claim that there
is void outside the finite world-system; cf. R.B. Todd, ‘Alexander of Aphrodisias and the case for the
infinite universe’, Eranos 82, 1984, 185–93.
92
Simplicius in Phys. 249,37; 489,20ff. Cf. Todd 1976, 17 n. 80. This is Zenobius no. 3 in RE vol. 10A
col. 12; no date for him is there suggested. Other references to Epicureanism by Alexander are cited
by Simplicius in Phys. 372,11; 467,1; 679,12; 679,32, and by Philoponus in GC 12,6.
93
Simplicius in Cael. 391,12.
94
Simplicius in Phys. 511,30.
95
Simplicius in Cael. 265,29; cf. 266,29. Cf. S. Sambursky, The Physical World of Late Antiquity, London
1962, 78.
96
Simplicius in Phys. 291,21. Cf. Gottschalk, above n. 75, 63–9.
97
On the arrangement of the Alexandrian Neoplatonist commentaries, cf. L.G. Westerink, The Greek
Commentaries on Plato’s Phaedo, Amsterdam 1976, 25, and references in n. 42 there. It is true that
Alexander does sometimes follow a general summary by a fuller discussion (cf. e.g. in Top. 181,21ff )
102 Aristotle Transformed

There are indeed occasions when a particular point prompts an extended

discussion, in effect a self-contained excursus;98 and Alexander does sometimes
refer to fuller discussion in another treatise.99
The later Neoplatonic commentaries have introductions describing Aristotle’s
works in the context of a formal curriculum;100 Alexander’s commentary on the
Prior Analytics does begin with a discussion of whether logic is a part of
philosophy or a tool (organon),101 and that on the Meteorologica discusses the
nature of the subject and the place of the treatise within a general classification
of Aristotle’s physical works.102 On the other hand the commentary on the
Metaphysics plunges immediately into a discussion of the opening sentences of
the Aristotelian text, with no preliminary consideration of the nature of
metaphysics as a subject, the composition of the Aristotelian treatise, or the
authenticity of the books of which it is composed – though it is true that these
topics are discussed later.103 Alexander did discuss the classification of Aristotle’s
writings104 and the authenticity of the de Interpretatione;105 he was also influential
in establishing Andronicus’ title for the work which thus became known as the

or by detailed comments on particular words within a section (e.g. in Metaph. 289,29,33). But this
is not a general principle of organisation as it is in the later commentaries. Nor is there a division
into daily lectures; proteraia in the context of a citation of Alexander ap. [Philoponum] in DA
470,18 is [Philoponus]’ term, not Alexander’s.
98
e.g. in An. Pr. 177,19–182,8, on Chrysippus’ claim that the impossible can follow from the possible;
ibid., 183,34–184,18, on Philonian, Diodorean and Aristotelian definitions of possibility. The
example in the latter passage of what is possible for Philo although prevented (it is still possible for
chaff divided into atoms, or at the bottom of the sea, to burn) seems to conflate two examples which
are distinct in Quaest. 1.18 31,9ff (chaff divided into atoms cannot burn, and a pebble at the bottom
of the sea cannot be seen). It does not however follow that the discussion in the commentary is later
than and derives from that in the quaestio, for there could be a common source. Cf. Sharples 1983,
99 and n. 5.
99
Above, nn. 19, 21.
100
e.g. Simplicius in Cat. 3,18. Cf. Westerink, above n. 97, 28.
101
Alexander in An. Pr. 1,1–6,12. Cf. also the discussion of the subject-matter of the Categories by
Alexander ap. Simplicium in Cat. 10,8–20 (above, n. 54); Moraux 1984, 609.
102
Alexander in Meteor. 1,1–4,11.
103
The relation between books 1, 2 and 3 is discussed at in Metaph. 136,13–138,23 and 196,19–26, and
this in itself involves some discussion of the subject of the whole treatise (138,6ff ); similarly the
discussion of 3 as in a sense the start of the enquiry, 1 being prefatory, leads to discussion of subject
of the whole (170,5–171,22). It may be noted, though, that the discussion in the commentary on 3
is attached to the opening section of that book as its lemma, while in the commentary on 2 there is
a separate opening section. Alexander is aware both of those who rejected 1 because of 3’s appearing
like an opening book (172,21–2) and, by implication, of those who found the first-person-plural
references to Platonism problematic; cf. 196,20, and W. Jaeger, Aristotle: fundamentals of the history
of his development, tr. R. Robinson, Oxford 1934, 175 and n.2.
104
Simplicius in Cat. 4,13; Moraux 1973, 75. (However, on this cf. L. Tarán’s review of Moraux 1973 at
Gnomon 53, 1981, 739–40, arguing that the material from Alexander here is heavily overlaid by
Simplicius’ own later concerns and terminology.)
105
Alexander in An. Pr. 160,32 (against Andronicus); cf. Boethius, in Int. (ed. sec.) 11,13–15. Moraux
1973, 117 and n. 2.
 The school of Alexander? 103

Categories, against the rival title Before the Topics (ta pro tôn topôn) (supported
by Adrastus and Herminus).106 Nevertheless, the general impression is that
Alexander’s concerns in this area are scholarly rather than a matter of formal
pedagogy.
Alexander frequently gives alternative explanations of the same point without
indicating which he prefers; Moraux has given a list of passages in Simplicius
which show that this was the case in Alexander’s lost de Caelo commentary,107 and
there are examples in the surviving commentaries too.108 This hardly suggests that
the commentaries formed part of a programme of teaching, at least not at any
introductory level. There are occasional remarks in the commentaries about the
process of teaching, but these are prompted by similar remarks in the Aristotelian
text itself and do not give any clear indication of Alexander’s own view.109
Where there is some evidence of school activity in the form of Alexander’s
writings is in some of the minor works.110 In addition to (i) ‘problems’ in the
strict sense with their solutions, these include (ii) expositions (exêgêseis) of
particularly problematic Aristotelian texts, (iii) short expositions of Aristotelian
doctrine on a particular topic, and (iv), straightforward and sometimes tedious
paraphrases of passages in Aristotle’s writings; both (iii) and (iv) alike seem to be
described as epidromai or ‘summaries’. There are also (v) collections, one might
almost say batteries, of arguments for a particular Aristotelian position.111 It
seems likely that (iii), (iv) and (v), in particular, reflect teaching activity; it is
difficult in some cases to see why else they might have been written. Some of
these texts may be Alexander’s own expositions of particular topics, while some
may be more in the nature of exercises by his students.112

106
M. Frede, Essays in Ancient Philosophy, Minneapolis 1987, 18–20. Cf. Alexander ap. Boethium in Int.
(ed. sec.) 10,14, on whether the de Interpretatione is concerned with philosophy or rhetoric.
107
Moraux 1967, 169 n. 1.
108
cf. (e.g.) in An Pr. 156,11–157,10; 161,3–26 (cf. Sharples 1982 at 97–9); in Metaph. 141,21ff; 159,9;
162,6,10; 164,24; 165,4; 165,21; 169,11; 220,24ff; 337,29.
109
At in Metaph. 167,5ff, on 2.3, 994b32, it is observed that people react to lectures in the light of what
they already know; and at ibid., 168,7ff, on 995a10, there is a warning against the danger of giving
an impression of hair-splitting.
110
The following summary account (which is not exhaustive) of the various types of texts to be found
among the minor works may be supplemented by the discussion at Bruns 1892, v–xiv; Moraux
1942, 19–28; Sharples 1987, 1194–5.
111
As Bruns 1892, xii–xiii notes, these are characteristic of the Mantissa rather than of the Quaestiones;
they also occur in the Ethical Problems. (Quaestio 2.28, however, approaches close to the type.)
112
cf. Bruns 1892, ix. In the ancient world an ‘exercise’ would almost certainly be a piece for oral
delivery before the teacher. It could in principle be written out beforehand or written up afterwards;
but some of the passages do not in themselves seem interesting enough to justify records of them
being made for general use after the discussion of which they might have formed part. (There may
be exceptions to this; cf. below at n. 174.) Alexander might on the other hand have preserved his
own teaching materials for future use even when they were somewhat pedestrian.
104 Aristotle Transformed

There are certainly, as one might expect, connections between the short texts
and the commentaries both in general thought and in particular detail. Aristotle’s
analogy in his de Anima of a point uniting two lines, to illustrate the way in
which the ‘common sense’ perceives the objects of different special senses, is
picked up by Alexander in his de Sensu commentary, and elaborated into the
analogy of the centre of a circle uniting the radii in the Quaestiones and de
Anima.113 Quaestio 3.11 may be connected with a possible report of Alexander in
Ammonius’ de Interpretatione commentary,114 and there are links between
quaestio 3.12 and Simplicius’ reports of Alexander’s de Caelo commentary.115 The
idea of nature as a force from the gods at Alexander, in Metaph. 104,8, should be
compared with the discussion in quaestio 2.3 of what the divine force (theia
dunamis) is that comes from the heavens to the sublunary.116 Simplicius in his
Physics commentary117 cites Alexander as saying that matter in itself (as opposed
to matter as matter for some thing, i.e. an instance of some particular type of
thing) is characterised not by the privation (sterêsis) of quality, for privation is in
itself a quality, but by the negation (apophasis) of it; the same argument is
advanced at quaestio 2.7 53,9–19.118

113
Aristotle DA 3.2, 427a9ff; Alexander in Sens. 164,13ff; Quaest. 3.9 96,10ff; de Anima 63,8ff. In
referring to Aristotle’s de Anima Alexander’s de Sensu commentary is picking up Aristotle’s own
cross-reference at Sens. 7, 449a9–10; Quaest. 3.9 is a commentary on 427a2–14. Cf. P. Henry, ‘Une
comparaison chez Aristote, Alexandre et Plotin’, in Les Sources de Plotin (Entretiens Hardt 5),
Geneva 1960, 427–49, at 436–8.
114
R.B. Todd, ‘Alexander of Aphrodisias on de Interpretatione 16a26–9’, Hermes 104, 1976, 140–6,
relates Quaest. 3.11 to Ammonius in Int. 39,17–32, arguing that this comes from Alexander, as does
explicitly the syllogism at 39,13–17 to which it is a reply.
115
Simplicius in Cael. 284,25–285,5; 285,21–5; 286,25–7, noted by I. Bruns, Interpretationes Variae,
Kiel 1893, 4 and 10, and R.B. Todd, ‘Alexander of Aphrodisias and the case for the infinite universe’,
Eranos 82, 1984, 185–93.
116
cf. Sharples 1987, 1188; I am grateful to Professor William E. Dooley for drawing my attention to
the in Metaph. passage. Such a force plays a central part in Alexander’s doctrine of providence, and
one might be tempted to think that the quaestio, discussing exactly what is due to the force, was a
development of the commentary, but in fact Moraux has argued that this quaestio is early (cf. n. 45).
117
Simplicius in Phys. 211,20ff (above, n. 53).
118
For other links between the commentaries and the minor works cf. also nn. 41, 53, 69 and 132. A
contrast between the privation, which per se is not but per accidens is, and matter which per se is but
per accidens is not, is drawn in a similar way in commenting on Aristotle Physics 1.8 both by
Alexander Quaest. 1.24 38,16–19 and by Philoponus in Phys. 173,21–6; but Philoponus does not
cite Alexander here, and it is not clear whether there is any connection with his lost Physics
commentary or whether the similarity is simply the result of coincidence. Cf. also Simplicius in
Phys. 240,19. At in Phys. 238,8 Simplicius reports Alexander as presenting alternative interpretations
of Aristotle’s solution to the problem of coming-to-be, depending in one case on a contrast between
per se and accidental not-being and in the other on one between per se and accidental coming-to-
be; Simplicius himself argues against the latter. See above, n. 107. Quaest 1.8,1.17 and 1.26 argue that
a form, and in particular the soul of a living creature, cannot be in matter or in a body as in a subject
(hupokeimenon; see below, n. 178). At de Caelo 1.9, 278b1–2, however, Aristotle refers to things that
have their being (ousia) in some underlying matter (en hupokeimenêitini hulêi); Simplicius, in Cael.
279,5ff records Alexander’s attempts to explain away this phrase.
 The school of Alexander? 105

Sometimes the relation between different works is more problematic. Quaestio

2.17 argues that even though fire as we experience it, an effervescence (zesis) of
fire, is hotter than elemental fire, identified as hupekkauma, it does not therefore
follow that elemental fire is less hot through admixture of the opposite, coldness.
Aristotle in the Meteorologica119 regards the effervescence and excess of heat as
real fire, by contrast with what occupies the region above air but below the
heavens, which ‘we are accustomed to call fire’. Alexander in his commentary
on the Meteorologica follows this doctrine, but appears to suggest that the
hupekkauma below the heavens, though not in fact fire, is nevertheless an
element (whereas the effervescence of heat, presumably, is not).120 The quaestio
too, in denying that the hupekkauma contains an admixture of coldness, is
safeguarding its elemental status.
Aristotle says, at de Caelo 2.7, 289a20, that the heat and light of the heavenly
bodies are caused by the effect they have by friction on the underlying air; this is
odd, because it should be fire that is next below the heavenly sphere. Simplicius
reports that Alexander interpreted this ‘air’ as the hupekkauma, though – he adds –
Alexander in putting forward this interpretation also observed that Aristotle
elsewhere called this hupekkauma elemental fire.121 This corresponds with the
quaestio, but not, it would seem, with Aristotle’s Meteorology itself or with
Alexander’s commentary on it. Perhaps, though, Alexander’s point is that Aristotle
did elsewhere identify what lay directly beneath the heavens as fire of a sort rather
than air, whether it was fire in the strict sense or not; Alexander himself is
consistent in interpreting the hupekkauma as an element whether fire in the strict
sense is to be identified with it or not. Presumably too for Alexander Aristotle is
describing the same substance as occupying this region in the Meteorology and in
the de Caelo; it is simply Aristotle’s terminology that differs.
In quaestio 2.23 it is argued that the attraction of the magnet for the iron is to
be likened not to the way in which the womb attracts the seed and the veins
nourishment, by drawing in intervening air or moisture, or to the way in which
the sun seems to attract water by turning it to vapour which naturally rises, but
rather to the way in which an object of desire attracts a living creature, the magnet

119
Aristotle Meteor. 1.3, 340b23.
120
Alexander in Meteor. 14,25ff, also citing Aristotle GC 2.3, 330b25. Alexander here refers to the
hupekkauma as to pur tôn stoikheiôn kaloumenon, ‘the one of the elements which is called fire’ –
implying that it is an element, though it is strictly wrong to call it fire.
121
Simplicius in Cael. 439,14. Simplicius’ own answer is that the circular movement of the heavens heats
the air as well as the hupekkauma above it (439,27–32); an explanation might perhaps be sought in
the idea that there is an admixture of fire and air even in the celestial sphere (cf. Meteor. 1.3, 340b8,
and Guthrie’s note on Cael. 2.7; W.K.C. Guthrie, Aristotle: On the Heavens, Loeb ed. 1939, 176–9).
106 Aristotle Transformed

possessing some feature which the iron lacks.122 Simplicius however reports
Alexander in his Physics commentary123 as first asserting that the magnet attracts
things by an incorporeal force – which Simplicius rejects as contrary to Aristotle’s
intention of showing that there is nothing intervening between what moves and
what is moved – and then as saying either that there is after all some corporeal
effluence or else that, since the explanation of magnetism is unclear, Aristotle may
not be referring to magnetism at all but to cases like that of flames which are
attracted by the burning wood and so do not rise up to their natural place.124 (This,
rather than the magnet, is in fact the example which Aristotle himself gives).125
Some of the short texts may indeed be extracts from the commentaries.126
However, if the editor who supplied the titles for the Quaestiones can be believed,
quaestio 2.22 differed from the account in Alexander’s lost commentary on
de Generatione et Corruptione.127 Quaestiones 3.10 and 3.14, summaries of
sections of Aristotle’s Meteorologica, can be compared with Alexander’s extant
commentary on that text, which we have.128 Some of the short texts relate to
issues discussed in Alexander’s treatises, rather than directly to Aristotelian
texts,129 but this division is not a hard-and-fast one.

122
The magnet is described as like iron but lacking moisture (74,26–8), so perhaps it is dryness that the
iron desires? Oddly, too, the quaestio speaks of the living creature attracting the nourishment that
it desires, rather than of the nourishment attracting the living creature (74,20–1); but it is clear that
the magnet possesses what the iron lacks, rather than vice versa (74,26).
123
That the citation is from Alexander’s Physics commentary is not explicitly stated, but is prima facie
likely and rendered more so by the fact that Alexander’s comments are directed towards the
elucidation of a phrase in Alexander’s text (Simplicius in Phys. 1055,15).
124
Simplicius in Phys. 1055,24–1056,6, on 7.2, 244a14.
125
Aristotle Phys. 7.2, 244a12–13.
126
cf. Sharples cited in n. 53 above, at 67–8. Quaestio 1.1 indeed corresponds to a section of the
commentary on the Metaphysics attributed to Alexander, but this commentary is not, at least in its
present form, by Alexander himself (above, n. 24); either there is a common source – which could
be the genuine commentary – or the compiler of the commentary we have has incorporated the
passage from the quaestio.
127
On these titles generally cf. n. 15 above, and on the relation between quaestio 1.11 and the de Anima
commentary see below, n. 142.
128
3.10 is very summary in form; its two pages correspond to pages 66–89 of the Meteorologica
commentary, and while it reports the explanations of the saltiness of the sea that appear in the
Meteorology itself and in the commentary (and in P.Hibeh 16 = Democritus 68A99a Diels-Kranz),
it omits the names of the proponents of the various theories as given in the commentary, though not
by Aristotle himself. Quaestio 3.14 is a summary of Aristotle (?) Meteorology 4, finishing at the end
of ch. 7; Alexander’s commentary itself devotes 34 pages to 4.1–7 and only 16 to the rest of the book.
The quaestio has a comment at the end saying that some things are missing ‘as he wrote before’; this
may be a later addition simply referring to the title at the beginning of the quaestio, which describes
it as ‘incomplete’. The quaestio and the commentary are sufficiently close for the former to be
emended from the latter (cf. quaestio 3.14 109,18 with Bruns ad loc.). Cf. V.C. Coutant, Alexander of
Aphrodisias: Commentary on Book IV of Aristotle’s Meteorology, diss. Columbia 1936, 89 n. 33.
129
Quaestio 3.3 is a turgid and repetitive summary of Aristotle de Anima 2.5, which it is difficult to
believe can have been written by Alexander himself; strangely, it enjoyed great subsequent
popularity, versions surviving both in Arabic and (from the Arabic) in Latin.
 The school of Alexander? 107

The interest and quality of argument of these texts varies considerably. Some
of them are so pedestrian and even turgid that it is difficult to believe that they
could have been produced for anything other than a pedagogic purpose.130
Others show considerable philosophical liveliness and acumen. For our present
purpose there are two interconnected questions that are of particular interest.
One is whether there are discrepancies in doctrine so great that it can be shown
that the texts cannot all reflect the viewpoint of a single author; the other is
whether any of these texts can be shown to reflect a context of live discussion
within Alexander’s school.

IV

There are certainly discrepancies both within particular texts and between
different texts so great that scholars have been reluctant to accept that all of the
views put forward are Alexander’s own. The problem is partly that of knowing
how systematic a thinker Alexander was, but there is the further problem, where
more than one text is concerned, that we are generally ill-informed about the
relative chronology of the works attributed to him,131 and that his views may
have changed over time.
Quaestio 1.11 is concerned to explain Aristotle’s statement in de Anima 1.1,
402b7 that ‘the universal living creature is either nothing, or posterior’. The text
is transmitted in two versions: the second and longer gives two solutions in full,
while the shorter version gives the second solution only and refers for the
first – as does also the longer text – to the author’s commentary on Aristotle’s
de Anima. The first solution (22,23–23,21) turns on the claim that there is no
prior genus of things which form an ordered sequence, as do the faculties of the
soul which different creatures possess;132 the second (23,21–24,22 = 21,19–22,20)

130
cf. Sharples 1987, 1181 and n. 32, with references. The form of many of Alexander’s writings –
commentaries on works by another author, and short discussions of individual problems – is such
as to make a piecemeal approach to problems more likely; but there is also the problem that
Alexander characteristically puts forward alternative solutions without indicating a preference
between them (cf. nn. 107–8).
131
For a summary of what is known cf. Sharples 1987, 1181 and nn. 35–40 with references there.
132
For this argument cf. also Alexander de Anima 16,18–17,8; 40,4–10; in Metaph. 152,6ff; 208,31ff.
Moraux 1942, 51–2; R.W. Sharples, ‘ “If what is earlier, then of necessity what is later?”: some ancient
discussions of Aristotle, De generatione et corruptione 2.11’, Bulletin of the Institute of Classical
Studies 26, 1979, 27–44, at 37 and n. 142. But as Moraux 1942, 57–8 points out, Alexander does argue
that there can be a hierarchy within a genus, and gives the example of different types of creature, at
in Metaph. 210,6–8, against Platonists who deny the existence of a generic idea of number.
108 Aristotle Transformed

argues that the universal is posterior to what falls under it in the sense that, if
only one instance of a kind exists, the nature of the kind does indeed exist but
need not exist as a universal, that is in more than one instantiation,133 while on
the other hand the universal will not (for a Peripatetic) exist if there are no
instantiations of it at all.
Moraux in 1942 suggested that the first solution was the genuine solution of
Alexander, as expressed in his commentary, and that the second solution was an
attempt by a pupil to improve on it; though he insisted on the merits of the
second solution as an interpretation of Aristotle’s thought concerning universals
in general.134 (This need not, I think, conflict with Tweedale’s view that the first
solution is more convincing as an interpretation of Aristotle’s thoughts in this
particular context135). The second solution can however be paralleled elsewhere
in the writings attributed to Alexander.136
There is, moreover, a difficulty within the second solution itself. A.C. Lloyd,
while regarding it as a statement of Alexander’s general position, rejected the
final part (24,16ff Bruns), which in effect asserts that the universal, while
posterior to the nature that happens to be universal, is nevertheless prior to each
individual particular (sc. in the cases where there is a plurality of particulars).137
This seemed to give the universal nature an unacceptable priority over the
particular.138 However, there are other texts attributed to Alexander where
similar points are made about both the genus and the species;139 and there does

133
23,25–24,1 = 21,23–30. The point is here put in terms of ‘animal’ – a generic nature – still existing if
there were only one individual animal, but it would certainly also apply to a species with only one
individual member, and to a genus with only one species (see below). If there were only one type of
animal, but many instances of that type, the nature ‘animal’ would presumably possess the accidental
property of being a universal with several instances, but not that of being a genus common to
several species. – With the following discussion cf. Sharples 1987, 1199–202, and references there.
134
Moraux 1942, 50–62.
135
M.M. Tweedale, ‘Alexander of Aphrodisias’ views on universals’, Phronesis 29, 1984, 279–303, at 284.
136
Quaest. 1.3 8,12ff.
137
The point is actually stated in terms of the relation between genus and species, but seems to apply
to that between species (or genus) and individual too; cf. nn. 133 and 139.
138
A.C. Lloyd, Form and Universal in Aristotle, Liverpool 1980, 51.
139
Alexander de Providentia 89,5 Ruland (above n. 74; cf. P. Thillet, ‘Un traité inconnu d’Alexandre
d’Aphrodise sur la providence en version arabe’, in L’homme et son destin d’après les penseurs du
moyen âge: Actes du 1er congrès international de philosophie médiévale, Louvain 1960, 313–24, at
321); cf. also Alexander’s Refutation of the assertion of Xenocrates that the species is prior to the genus
(S. Pines, ‘A new fragment of Xenocrates and its implications’, Transactions of the American
Philosophical Society 51, 1961, no. 2.) Complaints in authors like Simplicius (in Cat. 82,22–8;
85,5–9) and Dexippus (in Cat. 45,12–31) that Alexander makes universals posterior to particulars
do not conflict with this, for they simply reflect the fact that he did not give the universal form the
priority that would satisfy a Platonist, that is regarding the universal as existing even if it has
no individual instantiations at all. Cf. P.L. Donini, Le scuole, l’anima, l’impero: la filosofia antica da
Antioco a Plotino, Turin 1982, 222.
 The school of Alexander? 109

not seem anything paradoxical in saying (i) that the specific nature of man would
still exist even if it were only instantiated in one individual and hence was not a
universal (and similarly that the generic nature of animal would still exist,
though not as a genus, even if only one species of animal actually existed), and
yet (ii) that the nature of man as a species is prior to individuals like Socrates and
Callias with their individual accidents, snub-nosedness and the rest. (ii) may in
itself be an unacceptable and un-Aristotelian position (though I do not myself
believe so), but that is a different point; it does not in itself seem an impossible
one for Alexander to have held sincerely as an interpretation of Aristotle, and it
seems to agree not only with other texts attributed to Alexander but also with his
treatment of providence, where it is the preservation of the species that is the
primary concern.140
Unfortunately, though, there are yet other texts attributed to Alexander which
lay emphasis on the creation of universals by the minds that distinguish the
specific or generic nature from the individual accidents;141 it seems difficult to
reconcile these with the idea of a specific or generic nature that may or may not
be instantiated in several individuals or species, but is at any rate in some sense
prior to the individual members of a plurality with their accidents. The question
really is, whether we can believe that Alexander’s own views can have varied so
much as to allow both approaches to be his, or whether rival views were being
advanced by different members of his school.142
Quaestio 2.9 argues that the Aristotelian definition of soul as the entelechy of
a body of a certain sort143 does not involve making the soul relative to something.
For, while the soul is ‘of ’ the body, it is already something in itself first (which
seems to contradict Alexander’s definition of the soul as the product of the
mixture of the bodily elements144) just as a head is first of all a thing and (only)

140
Quaest. 1.3 8,22–4; 1.25 41,8ff; 2.19 63,15ff; de Providentia 33,1ff; 59,6ff; 87,5ff. Sharples 1987, 1216
and n. 170.
141
Alexander de Anima 90,6–7; Quaest. 2.28 78,18–20.
142
The citation of what must surely be Alexander’s de Anima commentary, now lost, as the author’s
own (eirêtai moi, 21,18 = 22,26) suggests one of three things: (a) the whole text with both solutions
is by Alexander himself; (b) the part with the first solution was composed by him (but why, if the
point was already discussed in the commentary?) and the second solution added later by someone
else; (c) the whole text is by a single author who is not Alexander but, in citing the commentary as
his own, wishes to pass himself off as so being (so Moraux 1942, 53, 61). A case where multiple
authorship, or at least an addition by the original author but at a later date, does seem likely is the
last section of quaestio 1.4; cf. R.W. Sharples, ‘An ancient dialogue on possibility: Alexander of
Aphrodisias, Quaestio 1.4’, Archiv für Geschichte der Philosophie 64, 1982, 23–38, at 37–8, but also
M. Mignucci, ‘Pseudo-Alexandre, critique des stoïciens (Quaestiones 1.4)’, in Proceedings of the
World Congress on Aristotle, Thessaloniki August 7–14 1978, Athens 1981, 198–204.
143
Aristotle DA 2.1, 412a27–8.
144
Alexander de Anima 24,21–3.
110 Aristotle Transformed

then the head of something else (which seems to contradict Aristotle’s view that
a hand, or any other bodily organ, is only a hand equivocally if it is not the hand
of a living creature, in other words that the part has its being in the context of the
whole145). These remarks in themselves are odd enough; when however the
quaestio goes on to argue that both soul and head indicate a proper nature and
substance which happens (sumbebêken) to belong to something else, it is difficult
to believe that this text can be by Alexander expressing his own views.

Since we know so little about the way in which the collections of short texts
attributed to Alexander were assembled,146 there is the possibility that a text
whose attribution to Alexander himself is questionable on doctrinal grounds
might be the work of a later scholar who never met him and knew his views only
through his written works. A surer sign of the origin of a text in a school
discussion may be given by phrases like ‘it was said’, ‘the question was asked’, ‘the
difficulty was raised’, and past tense first person plural endings. Such phrases
may indeed simply be cross-references within the commentaries themselves; the
first person plurals may just express an editorial ‘we’. And even when there does
seem to be reference to discussion, it may be a discussion carried on in earlier
tradition rather than in Alexander’s own school.
At in An. Pr. 191,19ff Alexander is discussing the proof by reductio at 1.15,
34a34ff. He first observes that certain people said certain things (elegeto oun tina
hupo tinôn, 191,24) to the effect that the conclusion of the reductio is false but
not impossible, and so does not generate a contradiction. To this Alexander
responds (legomen, 192,4) by appealing to Aristotle’s observation at 34b7ff that
assertoric premisses need to be understood without temporal restriction. A
further objection (‘to this it was said in turn’, pros touto elegeto palin, 193,13)

145
cf. Aristotle Meteor. 4.12, 389b31; 390a12; DA 2.1, 412b21; PA 1.1, 641a4; GA 1.19, 726b24; 2.1,
734b25, 735a8; Metaph. 7.10, 1035b24ff.; Politics 1.1, 1253a21. It is true that a head is less obviously
an organ used for a certain purpose than is a hand. Admittedly, Aristotle does in Categories 7 argue
that particular heads and hands, as primary substances, are not relative at all (8a20), and that ‘head’
and ‘hand’ as universal, secondary substances (which are the concern of our quaestio; 54,23) are
relative in the sense that what they belong to enters into their definition, but not in the narrower
sense that their very being consists in being related to something else in a particular way (8a27ff.).
This might perhaps be explained by saying that a bodily organ such as a head or a hand is to be
defined in terms of its function, rather than (e.g.) of its position in relation to the whole body. But
it still hardly provides a parallel to the use of sumbebêken in the quaestio.
146
cf. however above, nn. 14–16.
 The school of Alexander? 111

turns on the minor premiss of the reductio, arguing that it is wrong to create a
contradiction by supposing that what was only stated as possible does in fact
apply; for, it is claimed, ‘can’ is only said of what is not in fact the case.147 Alexander
answers this objection (luoito, 193,20) with the observation that there is a
difference between (i) supposing, when something is not in fact realised, that it
is realised, and (ii) supposing that something is realised when it is not, in the
sense of supposing the contradiction that it is simultaneously realised and not.148
It seems difficult not to take this as the record of an actual discussion.
At in An. Pr. 218,7ff Alexander is discussing the argument at 1.17, 36b26 that
two contingent negative premises in the second figure yield no conclusion,
because with contingent premises negative and positive are convertible, and two
positive premises yield no conclusion in the second figure.149 Alexander observes
that someone might ask (epizêtesai d’an tis) why a similar argument is not
applied, with reverse effect, to positive contingent premises that do apparently
yield a conclusion in other moods. The first reply, introduced by the words ‘it was
said first of all’ (elegeto prôton men, 218,13) is that such conversion is natural
only in the case of premises that would not yield a conclusion without it (which
seems to beg the question rather). Alexander then continues with the argument
(eti de, 218,20) that the contingent premise is in fact positive rather than negative
in character. Again, it seems natural to interpret ‘it was said first of all’ as a record
of an actual discussion. However, the topic of mixed modal syllogisms was one
on which Alexander disagreed with two of his teachers, Sosigenes and
Herminus;150 so in both these passages the discussion referred to could be one in
which Alexander participated as a pupil, not one in his own school.151
At in Top. 133,24, on 2.1, 109a27ff, Alexander says that there was an enquiry
(ezêtêthe) as to how a problem could be false. His reply is that what is false is the
actual statement that someone supports and the dialectician tries to refute.152

147
On this cf. Sharples 1982, at 97 and nn. 91–2.
148
cf. Aristotle, SE 4, 166a24; Cael. 1.12, 281b8ff.; Metaph. 9.4, 1047b13.
149
So Alexander. Aristotle’s own argument is not that (e.g.) Q(aEb) and Q(aAb) convert, but that
Q(aEb) and Q(bEa) do not; cf. 36b35, and Alexander at 220,3ff.
150
Above, nn. 20, 25.
151
At Quaest. 2.3 48,18–22 (cf. above, n. 45) the statement that the derivation of human reason from
divine providence is established is followed by the remark that for this reason ‘it was said’ (elegeto)
that man derives from divine providence everything he has through reason. It is not clear whether
‘it was said’ refers to the context of a particular discussion, of which this quaestio is the only part
recorded, or to a more general tradition.
152
Appropriately enough, Alexander’s solution to this general problem about problems is itself
introduced by a phrase used to introduce solutions to particular problems, ‘may it not be the case
that’ (mêpot’ oun, 133,26; cf. Alexander in Top. 250,26; Quaest. 1.26 42,25; 2.3 48,22; 3.2 81,29; Mant.
119,31; 177,5). What would be false in this case, according to his solution, would be not the question
‘how can a problem be false’ but the assertion ‘a problem cannot be false’.
112 Aristotle Transformed

There is no ‘enquiry’ into the point in question in Aristotle’s own text; and it
seems natural to take ‘there was an enquiry’ as referring to some discussion at
which Alexander was present, rather than to the commentary tradition.153
Other quaestiones beginning with the statement ‘there was an enquiry’
(ezêtêthe)154 include 1.9, on the question how, if opposites are the things furthest
removed from each other in the same genus, it is not disease, rather than poor
physical condition, that is the opposite to good physical condition, since the later
is an intensification of mere healthiness,155 and Ethical Problem 17, which
considers under which type of good pleasure should be classified.156 In neither of
these passages, again, is there any obvious indication that the enquiry referred to
is one in earlier written commentaries, rather than in a discussion in the school;
equally, there is no indication that the development of the problem or its solution

153
A rather similar case occurs in Quaest. 3.2. After citing Aristotle, DA 2.5, 417b5, this continues,
‘Aristotle having said this, there was a general enquiry by the commentary (exêgêsis) whether the
transition to actuality is a change’. (The enquiry was ‘general’ (katholou) because it related to
transitions to actuality in general, and not just to the actualisation of knowledge in the activity of
theôria.) ‘By the commentary’, hardly natural in English, could in principle refer to one particular
commentary either as the actions of commenting or as the resultant written work, or, perhaps more
naturally, to the tradition of commentary on the passage extending over several generations.
However, ‘by (hupo) the commentary’ is only an emendation by Bruns; the MSS have ‘from (apo)
the commentary’. With this reading it seems more natural to interpret the past tense as referring to
consultation of a commentary on a particular occasion, which might or might not be in the context
of a discussion actually taking place in the school. But the point at issue is precisely the sort on
which, in reading Aristotle’s de Anima, one would go and look up commentaries on other
Aristotelian works, if one did not indeed already have to hand a commentary on the de Anima that
itself discussed the issue in general terms; and in this case the singular ‘from the commentary’ seems
odd, unless exêgêsis were to have the sense of ‘the tradition of commentary on all Aristotle’s works
in general’.
154
Such references do not indeed always imply discussion. When in P. Eth. 25 the author states that ‘we
enquired into what happiness is’ (ezêtêsamen ti pote estin hê eudaimonia), the first person plural
verb is most naturally taken as including both the writer and his readers – or the speaker and his
audience – as students of the Nicomachean Ethics in general, even if the text never actually formed
part of an ordered sequence of expositions of the whole of that work. (I am grateful to Professor
Sorabji for a suggestion that prompted this line of thought.) What follows certainly seems more like
a teaching summary based on Aristotelian texts than anything that had its origin in a philosophical
debate.
155
The attempt to reconcile a general Aristotelian principle and an awkward example is characteristic
of the Quaestiones. The question might indeed arise, in discussion of an Aristotelian text such as
Top. 8.2, 153b17ff; Alexander’s commentary on this passage (in Top. 507,10ff ) does not contain
anything relating closely to the point of the quaestio, though the extant version of this section of
Alexander’s Topics commentary may be an abridgement.
156
The conclusion reached is that, just as pleasures are good only if they accompany activities that
deserve to be chosen (cf. P. Eth. 2 120,11ff; 16 137,2ff; 19 139,23ff; Aristotle EN 10.5, 1175a21ff ), so
the type of good they fall under will be that under which the activity to which they relate also falls.
The classification of goods from which the quaestio starts is a familiar one in the Peripatetic
tradition; cf. Arius Didymus ap. Stobaeum Ecl. 2.7.19 (p. 134,20–1 Wachsmuth); [Aristotle] MM
1.2, 1183b19; Aspasius in EN 32,10; Moraux 1973, 370; R.W. Sharples, ‘The Peripatetic classification
of goods’, in W.W. Fortenbaugh (ed.), On Stoic and Peripatetic Ethics, the work of Arius Didymus,
New Brunswick 1983 (Rutgers University Studies in Classical Humanities 1), 139–59, at 143 and 151.
Indeed, the classification occurs in Alexander’s Topics commentary (Alexander in Top. 242,4 =
Aristotle fr. 113 Rose), though without any reference to pleasure.
 The school of Alexander? 113

is itself the record of an actual discussion, ‘there was an enquiry’ simply being the
starting point from which the statement of the problem and its solution develops.
In quaestio 1.16 citation of Aristotle, Physics 1.5, 188a26–30, is followed, in the
primary MS , V, and in Bruns’ edition, by ‘He raised the problem of what was
meant by this statement’ (inferior MSS and the earlier editions have the passive,
‘the problem was raised’). It does not seem easy to interpret ‘he raised the
problem’ as a reference to some other part of Aristotle’s own discussion in the
Physics;157 it could be a reference to a speaker in a discussion or to some earlier
commentator, but in either case the lack of any explicit indication as to who is
meant is odd.
Elsewhere too there are problematic references to unidentified persons which
could reflect a context of discussion in which the identity of the person meant
would be clear.158 Ethical Problem 4 begins by stating an argument the conclusion
of which is considered paradoxical: ‘there is no opposite to what is an instrument,
but there is something opposite to wealth, so wealth is not an instrument’, and
then continues, ‘if he supposes that generally no instrument has an opposite . . .’,
where ‘he’ must be the proposer of the objection.159 However, when in Ethical
Problem 14 an extended statement of the problem of how there can be an
intermediate state between pleasure and distress on Aristotle’s account160 is
followed by the bald statement ‘yet they think that there is some condition
intermediate between pleasure and distress’, ‘they’ is probably a reference to
Aristotelians and other right-thinking people in general – Alexander tending to
identify Aristotelian doctrine and common sense161 – for the problem is that of

157
Aristotle does go on (188a30–1) ‘it is also necessary to consider by reference to argument how this
comes about’, that is (W.D. Ross, Aristotle’s Physics, 488–9) by argument as well as by appeal to
authority. But the discussion in the quaestio is an attempt to interpret difficulties in Aristotle’s
position, not a statement of his argument. ‘He raised the problem’ in 28,8 appears to be picked up by
‘someone might raise the problem’ in 28,17; and while autos in 28,11 refers to Aristotle, it is
Aristotle’s apparently conflicting statements, that opposites come from one another (Phys. 1.5,
188b21–6 being so interpreted at 28,10–16) and that the principles are opposites but cannot come
from each other (188a27–8) that have raised a difficulty in the mind of someone else. The difficulty
is noted, without reference to Alexander by name or explicit citation of 188b21–6, by Simplicius (in
Phys. 182,19–21); his solution (182,21–6) is similar to 28,25–29,2 of the quaestio. Cf. also Philoponus
in Phys. 111,19ff.
158
cf. also Mant. 112,6, above n. 44.
159
122,34. Bruns 1892, xiv used the fact that the objector is simply referred to in passing in this way to
argue that this text was an incomplete fragment of some larger whole; but it is not clear that this is
a safe inference. He similarly argues that P. Eth. 3 is a fragment, because the Stoics are there referred
to simply as ‘them’, but it is not clear how far ‘fragment’ is a relevant concept in the case of these short
texts whose original context is unclear. Certainly P. Eth. 3 and 4 are not fragments in the same sense
as Quaest. 2.21 (below, n. 166) or Quaest. 3.14 (above, n. 128).
160
Alexander cites Aristotle, EN 7.12, 1153a14 and 10.4, 1174b14.
161
cf. R.W. Sharples, Alexander of Aphrodisias: On Fate, London 1983, 18.
114 Aristotle Transformed

reconciling Aristotelian theory with accepted opinion, rather than an attempt to

find a damaging inconsistency in some opponent’s particular position.

VI

Quaestio 2.21, already discussed in the context of Alexander’s relations with

Platonism, is in the form of a narrated dialogue. The opening deserves extended
quotation, as the nearest we have to what at any rate purports to be an account
of a discussion between Alexander and his associates:

Recently I was162 discussing providence with my companions, and I tried to

show both that there is some concern and forethought from the divine beings
concerning mortal ones, according to Aristotle, and what it is. I was ready to say
in what way it comes about, when one of those who were present said that it was
worth learning first what response one should give to those who ask whether
one should say that the divinities exercise providence over these things here in a
primary way according to us too, or only accidentally . . .163

The narrator responds by challenging the assumption that primary providence

and purely accidental ‘providence’ are the only two possibilities, arguing that
there are several ways in which providence or forethought can be neither primary
nor accidental, without indicating which of them actually applies to Aristotelian
providence as he understands it.164 The respondent then asks him to give an
account of the Aristotelian position and to say which of the solutions suggested
does in fact apply (70,6–11).165 The narrator protests that this goes beyond his
original undertaking (70,11–17), but eventually agrees (70,24–71,2); and at this
point the text breaks off. In fact, the narrator had intended to give an account of
the Aristotelian position even before ‘one of those present’ intervened; the point
of the exchange at 70,6–17 is presumably to emphasise that the solution will be
in terms of the ways found in the intervening discussion of escaping this
dilemma, and also to emphasise the difference between finding a number of
dialectical solutions to an opponent’s objections and giving a positive account of
one’s own position.

162
The Greek has the plural ‘we’, presumably the editorial ‘we’.
163
Quaest. 2.21 65,18–23.
164
66,3–70,6.
165
On this cf. R.W. Sharples, ‘Alexander of Aphrodisias on divine providence: two problems’, Class.
Quart. 32, 1982, 198–211, at 204–8.
 The school of Alexander? 115

The strategy of the dialogue, in so far as it can be discerned in its incomplete

state, was similar to that of the treatise de Providentia surviving in Arabic, and
the quaestio may be an unfinished attempt to develop the ideas of the de
Providentia in a dialogue form.166 The dialogue is a literary construction rather
than the record of an actual discussion. The introductory section may be purely
based on literary models – though Alexander shows little interest in Aristotle’s
exoteric dialogues.
In Ethical Problem 7 the initial statement of the problem how some pleasure
can be good and other pleasure bad, if all pain is bad, is in the past tense: ‘it was
necessary either for pleasure too to be a bad thing . . . or for all [pleasure] to be a
good thing’, because of the apparent paradox in some pleasure being opposed to
pain as good to bad, other pleasure as one evil to another. The past ‘it was
necessary’ can be read, and is perhaps most naturally read, as a report of a
discussion.167 What is more striking, however, is that the second and longer of
the two solutions offered168 is introduced not by the customary ê (‘or rather: . . .’),
but by ‘something of the following sort was also said’ (elegeto de ti kai toiouton).
‘Was said’ seems to apply not just to the first sentence that follows it (127,20–3),
which could hardly stand on its own as a contribution, but to the whole view
developed in the following section;169 and it seems natural to take this as being a
discussion at which the author of the Problem was present. The suggestion
introduced in this way is after all one of the two main suggested solutions to the
problem, and the one stated at greater length; in other words, it is part of the
reason for this text existing at all in its present form.
A similar expression occurs in Mantissa 169–72. This text, entitled by the editor
‘[From] the [teachings] of Aristotle concerning what depends on us’, is an attempt
to resolve the problem of free human agency in an orderly universe, a problem

166
ibid., 199 and nn. 13–14.
167
Not all past tenses have this significance, of course. At P. Eth. 16 137,9 ‘the opposite of what is bad
was either good or bad’ is probably best taken, not as a reference to a discussion, still less as a
reference back to other texts (5 125,2ff; 7 127,3ff ) in a collection which probably did not exist as
such at the time of writing, but rather as a statement of generally accepted doctrine. Cf. also perhaps
P. Eth. 5 124,29.
168
The first suggested solution (127,8–20) is that not all pain is in fact bad; pain relating to shameful
activities is something deserving to be chosen (in the sense, of course, that we should positively
want to find such activities painful, not in the sense that we should want both them and the pain
that attaches to them). Another suggestion, which occupies the rest of this text, is then made
(127,20ff ); essentially, it is that, even if all pain is bad, it is opposed to some pleasures as one evil to
another, to others as good to bad. This is really just a restatement of what was found difficult at the
beginning of the text (127,5–7), but now at 127,20ff an account of how this can be so is provided by
the distinction of different types of pleasures.
169
The next sentence, 127,23–4, begins ‘For’ (gar).
116 Aristotle Transformed

which the de Fato itself does not really deal with in a satisfactory way.170 Its author
takes the bold step of linking responsible human choice with the presence of not-
being in the sublunary realm, which risks making it an imperfection in the
universe as a whole. (It might be argued that it is nevertheless the highest aspect
of sublunary nature; but our text does not in fact make this point.171)
The opening words of this discussion are ‘concerning what depends on us, an
opinion like the following was also stated (elegeto tis kai toiade doxa)’. The
argument that neither what depends on nature nor what depends on education
can depend on us (169,34–170,2) is then linked with the problem with which the
subsequent discussion is chiefly concerned, that of the apparent dilemma of
determinism or uncaused motion (170,2–3: ‘and this would be a still greater
difficulty, if it were the case that nothing comes about without a cause’). The text
then comments, ‘and this, too, seemed so to everyone (kai auto hapasin edokei)’,
before proceeding to its solution. It is not immediately clear whether ‘this seemed
so to everyone’ refers to a specific occasion or to the opinions of philosophers in
general, and whether ‘this’ is the fact that there would be a difficulty, or the view
that nothing comes about without a cause.172 Nor is it clear whether the initial ‘an
opinion like the following was also stated’ applies just to the indications of the
problems, or also to the following solution.173 If the former, the reference could
be to a particular discussion – as ‘was also stated’ perhaps suggests – or to a
general tradition; the difficulties raised at the start of the passage are after all
ones which would occur naturally enough in any discussion of the problem of
free will and determinism. If on the other hand ‘an opinion like the following was
also stated’ refers to the solution as well, the question arises whether the author
introduces the solution in this way because he himself does not accept it. This
has indeed been the most common interpretation, that Alexander himself
reports, because of its interest, a view that he does not himself share.174

170
cf. R.W. Sharples, ‘Responsibility, chance and not-being (Alexander of Aphrodisias Mantissa 169–
172)’, Bulletin of the Institute of Classical Studies 22, 1975, 37–64, at 42.
171
Sharples, op. cit. 52 and n. 170; cf. id. ‘Responsibility and the possibility of more than one course of
action: a note on Aristotle de Caelo 2.12’, Bulletin of the Institute of Classical Studies 23, 1976, 69–72.
172
It is true that most Greek thinkers had held that nothing came about without a cause, except
Epicurus; but his atomic swerve is, oddly, nowhere mentioned in the writings attributed to
Alexander. Bruns indeed suggests reading tisin for hapasin; ‘and this, too, seemed so to some’.
173
However, Professor G.E.R. Lloyd suggested to me that it might be more natural to refer to the
statement of the problems alone by elegeto tis toiade aporia, rather than doxa.
174
So Bruns 1892, xiii; Merlan 1967, 88. Later ancient and medieval practice might suggest that
discussions would have been recorded by a pupil rather than by Alexander himself; but that on its
own would not explain the contrast between occasional remarks like ‘the following opinion was
stated’ and solutions which on the face of it seem to be put forward, even if tentatively, by the writer
in propria persona.
 The school of Alexander? 117

Ethical Problems 8 and 28 present two different solutions to the same problem,
that of the relation of virtue to the various virtues. A cross-reference from one
Problem to the other has apparently been inserted in the text.175 There is a
temptation to see these two texts as, perhaps, papers prepared for the same
seminar; but there could on the other hand have been a considerable lapse of
time between their writing. Alternative answers to a single problem occur both
in the Problems attributed to Aristotle176 and elsewhere in the texts attributed to
Alexander.177
Another case in which one text seems to be developing a discussion from
another has been brought to my attention by John Ellis. Quaestio 1.8 begins
by stating the argument that a form cannot be in matter, or a soul in a body,
as in a subject (hupokeimenon); for a subject must be something that exists in
actuality in its own right, but matter cannot exist apart from form (17,8–12).
This argument is introduced by ‘it was said that’ (ên legomenon hoti. 17,9).
The objection is then made that the argument could equally well be applied
to accidents of bodies, since a body cannot exist without shape and colour
(17,12–17); and the quaestio continues with a reply to this objection (17,17ff ).
The initial argument at 17,8–12 occurs also in the Mantissa,178 but without the
particular objection at 17,12–17 being brought against it. It therefore looks very
much as if the quaestio may be referring back to, and building on, the discussion
in the Mantissa.

VII

The impression given by the writings attributed to Alexander is one of lively

philosophical discussion. The opinions of his predecessors and of his near-
contemporaries from other philosophical schools play a part in this; but it is

175
P. Eth. 8 128,22, bracketed by Bruns as an editorial addition.
176
cf. e.g. [Aristotle] Probl. 10.18 and 10.54, 10.48 and 34.1, 11.20 and 11.47, 38.1 and 38.11.
177
The subject matter of P. Eth. 13, on differentiation among pleasures, is to some extent similar to that
of P. Eth 2, 17 and 19, but the argument in 13 that we choose between pleasures and activities in
themselves, rather than as means to some further end, seems a distinctive contribution.
178
At Mantissa 119,32–120,9; I am grateful to Mr Ellis for drawing to my attention the significance of
the absence of the counter-argument found in quaestio 1.8. The text at Mantissa 120,1 is corrupt
and difficult to interpret, but the general sense is clear. Similar arguments occur also at Quaest. 1.17
30,3ff and Quaest. 1.26 42,14ff; cf., on Aristotle, G.E.M. Anscombe and P.T. Geach, Three Philosophers:
Aristotle, Aquinas, Frege, Oxford 1961, 10f, and above n. 118. But the mention of the soul suggests
that the reference in Quaest. 1.8 is to the Mantissa rather than to Quaest. 1.17 (where the soul is only
mentioned at the end), and the emphasis of Quaest. 1.26 is rather different.
118 Aristotle Transformed

difficult to believe that some at least of the texts here considered do not reflect
the activity of Alexander’s own philosophical school. This makes it all the more
odd that we know so little about any pupils of Alexander, and that as far as our
information goes he seems to mark the end of a distinctive and continuous
Peripatetic tradition.

References

This is a list for the reader’s convenience of some works referred to repeatedly in the
notes to this article; it is not intended as a comprehensive bibliography of the subject.

I. Bruns (ed.), Supplementum Aristotelicum 2: 2.1, Berlin 1887, 2.2, Berlin 1892.
J.M. Dillon, The Middle Platonists, London 1977.
J.P. Lynch, Aristotle’s School, Berkeley 1976.
P. Merlan, ‘Zwei Untersuchungen zu Alexander von Aphrodisias’, Philologus 113, 1969,
85–91.
P. Moraux, Alexandre d‘Aphrodise: Exégète de la noétique d’Aristote, Liège and Paris
1942.
P. Moraux, Der Aristotelismus bei den Griechen von Andronikos bis Alexander von
Aphrodisias: I, Berlin 1973; II , Berlin 1984.
P. Moraux, ‘Alexander von Aphrodisias Quaest. 2.3’, Hermes 95, 1967, 159–69.
P. Moraux, ‘Le De anima dans la tradition grecque’, in G.E.R. Lloyd and G.E.L. Owen,
(eds), Aristotle on Mind and the Senses: Proceedings of the Seventh Symposium
Aristotelicum, Cambridge 1978.
R.W. Sharples, ‘Alexander of Aphrodisias, Problems about Possibility I-II ’, Bulletin of the
Institute of Classical Studies 29, 1982, 91–108; 30, 1983, 99–110.
R.W. Sharples, ‘Ambiguity and opposition: Alexander of Aphrodisias, Ethical Problems
11’, Bulletin of the Institute of Classical Studies, 32, 1985, 109–16.
R.W. Sharples, ‘Alexander of Aphrodisias: scholasticism and innovation’, in W. Haase
(ed.), Aufstieg und Niedergang der römischen Welt II .36.2, Berlin 1987, 1176–243.
R.B. Todd, Alexander of Aphrodisias on Stoic Physics, Leiden 1976.
 5

Themistius: the last Peripatetic commentator

on Aristotle?1
Henry J. Blumenthal

It may be thought old fashioned to begin with dates, but the fact that Themistius’
active career ran from the late 340s to 384 or 5, and that his paraphrases of
Aristotle were probably compiled during the earlier part of the period 337–357
is not unimportant.2 By the end of his life Plutarch the son of Nestorius, arguably
the man who restored the study of Platonic philosophy at Athens,3 will already
have been teaching there. By the close of the fifth century the study of Aristotle
had become inextricably intertwined with the exposition, as it purported to be,
of Plato’s thought, and the fifth- and sixth-century commentators whose work
we now have were Neoplatonists to a man. The last major commentator before
Themistius, over a century before him, was Alexander, who may claim to be a
genuine Peripatetic without too much fear of contradiction. Notwithstanding
recent attempts to establish that Alexander was influenced by Plato or Platonists
on certain important issues,4 the core of his thought, and so his explanation
of Aristotle, is overridingly Peripatetic. That is not to say that there are no
ideas in Alexander which are the product of meditation on and argument with

1
References to the Aristotelian commentators are to page and line of the Berlin Academy edition:
where unspecified they are to commentaries on the de Anima.
2
For these dates cf. O. Seeck, Die Briefe des Libanius, Leipzig 1906, 292f and 306; G. Dagron, L’Empire
Romain d’Orient au IVe Siècle et les Traditions Politiques de l’Hellénisme: Le Témoignage de
Thémistius, Travaux et Memoirs du Centre de Recherche d’Hist. et Civ. Byz. 3, Paris 1968, 5 n. 3 and
12; A.H.M. Jones, J.R. Martindale and J. Morris, Prosopography of the Later Roman Empire I,
Cambridge 1971, 892 and 889.
3
cf. É. Évrard, ‘Le maître de Plutarque d’Athènes et les origines du néoplatonisme athénien’, Antiquité
Classique 29, 1960, 404–6 and my ‘529 and its sequel: what happened to the Academy?’, Byzantion
48, 1978, 373–5; for another view see H.-D. Saffrey and L.G. Westerink (eds), Proclus, Théologie
Platonicienne 1, Paris 1968, xxxv–xlviii.
4
cf. P. Merlan, Monopsychism Mysticism Metaconsciousness. Problems of the soul in the Neoaristotelian
and Neoplatonic Tradition, The Hague 1963, ch. 2; P.L. Donini, Tre Studi sull’Aristotelismo nel II
Secolo d.C., Turin 1974, passim.

119
120 Aristotle Transformed

other traditions, but he clearly stood apart from the Middle Platonist amalgam
of his time.
In between came Plotinus, Porphyry and Iamblichus. By the time of
Philoponus and Simplicius Neoplatonism had been dominant for some 250
years, and independent commentary was a thing of the past. Aristotle was by
then almost always read as a thinker who was expressing the Platonic truth,
sometimes inadequately or obscurely.5 These later commentators owe a great
deal to Porphyry and Iamblichus, as well as to their successors, and were not
ashamed to confess their debt. Since Porphyry and Iamblichus were active before
Themistius, it would not be surprising if he too were strongly influenced by
Platonism, and yet one’s first impression on reading his philosophical work is
that he was not. Nevertheless the Orations make a number of pronouncements
which might point in the opposite direction. So occasionally do the paraphrases.
The purpose of this chapter is to look at the evidence on a few points that may be
taken as a test of philosophical orientation, in the hope of being able to determine
which of these contrary impressions is correct.
Three kinds of evidence are available to help us make this decision. First,
Themistius’ own statements of his philosophical attitudes. Second, and most
important, the substance of his interpretations of Aristotle: does he give what we
should regard as a straight reading of Aristotle, or does he prefer the kind of
explanation that makes Aristotle a Platonist? Theoretically, we should be able to
derive further information on this question from Themistius’ comments on
previous interpretations, but the paraphrase style of commentary does not
normally allow him to discuss them, though it is sometimes clear that he is
following Alexander.6 On the other hand we do have some references to
Themistius in the later commentators, who will occasionally take issue with him
on whether or not a passage should be read Platonically.
The main source for Themistius’ own statements of his orientation is not
his philosophical oeuvre, but rather his Orations. Their reader cannot fail to
notice the honorific terms in which he repeatedly refers to Plato. He is divine,
all-wise, reverend, most magnificent and, most often, marvellous. In addition,
Themistius’ admiration for Plato seems to have impressed his correspondents.
Gregory of Nazianzus speaks of ‘your Plato’ (Ep. 24), and Libanius attributes
to Themistius, as a compliment, the possession of both Plato’s faculties, noble

5
cf. Chapter 13 below.
6
cf. esp. Themistius’ comments on nous dunamei and nous kath’ hexin and Alexander de Anima
80.24ff.
 Themistius: the last Peripatetic commentator on Aristotle? 121

teaching and eloquence (Ep. 793). Further, Themistius clearly approved of

the metaphysical foundation which Plato gave to his politics. In Or. 34 he
comments on how Plato progressed beyond Socrates and his band, who were
concerned with virtue, vice and happiness on the human plane, to attach human
to divine goodness, and to model the human polity on that of the universe
as a whole; with that disregard of progress in Plato’s thought characteristic
of antiquity, he sees this as the point of the Gorgias as well as the Republic,
Phaedrus and Laws. Themistius seems to prefer this grandiose design to
Aristotle’s more modest treatment of virtue on the human plane alone (447,29ff
Dindorf). Yet at other times he will say that Aristotle’s view of philosophy and
politics is better than that of Plato, who was too naive in saying that the world’s
wrongs will not end before philosophers become kings or kings philosophers
(Or. 8 = 128,19ff D).
One point of particular interest that emerges from the Orations is the
frequency with which Themistius mentions Plato’s views on ethics and politics.
Themistius’ Plato is that of the philosopher-king and the Cave, not that of the
Good beyond being, or the perfect being of the Sophist. His Platonic texts – they
tend to recur – are predominantly those that might be described as in the
Socratic rather than the Pythagorean and metaphysical tradition, and this
immediately marks his Plato as very different from that of the Neoplatonists.
Plotinus, as W. Theiler once pointed out, gives us a Plato dimidiatus:7 the missing
half was that concerned with moral virtue in a city (polis) and the organisation
and structure of that city. Themistius, who was, after all, a politician and man of
affairs, was interested in this other half, in Plato the would-be man of action who
went three times to the West (cf. e.g. Or. 8 = 125,19–32D), thereby showing at
least his independence of the Neoplatonic tradition which concentrated on Plato
the metaphysician.
Themistius did not abandon his high regard for Plato when he wrote the
paraphrases on Aristotle – he is alleged by Photius in a report, which in other
respects is not reliable, to have written paraphrases on Plato too.8 There, for
reasons connected with the type of exposition, explicit references are
comparatively rare, but when he does appear Plato is treated no less reverently
than in the Orations (cf. e.g. 107,22). Probably for this reason Themistius has

7
‘Plotin zwischen Plato und Stoa’, in Les Sources de Plotin, Entretiens Hardt 5, Vandoeuvres-Geneva,
1960, 67.
8
cf. pp. 121–2 below and n. 32.
122 Aristotle Transformed

sometimes been described as a Platonist.9 But when it comes to actual doctrines,

and here we pass to our second heading, reverence for Plato is not translated into
that superimposition of Platonism on Aristotelian texts which is the hallmark of
the Neoplatonists.
At this stage we must identify some criteria by which the commentaries can
be used for our present purpose. If we look primarily at the relatively well-
documented tradition stemming from Aristotle’s de Anima – ‘stemming from’ to
include independent treatises like Alexander’s de Anima as well as commentaries,
whether paraphrases or the more conventional kind – we may take the following
points, exempli gratia, as reliable indicators:

(1) Do soul and body form a single unit, related as form to matter, or is soul
separate, and in control of an ontologically inferior subordinate?
(2) Is the soul a continuum, possibly with a division between intellect and
reason, a matter on which Aristotle himself is notoriously hesitant, or does it
divide at one or more points below this level?
(3) A related question: is the soul seen in Aristotle’s way as a structure in
which the lower components are pre-requisites for the existence of the higher
ones, or as a Neoplatonic ‘emanation’ where the higher level is given and the
lower ones derive from it?
(4) Does the human soul relate to higher reality by being a part, or a product,
of it? Here, though as we shall see the answer may not be decisive, we may put
the question which became so important later, ‘is the agent intellect one or
many?’

The most difficult problem for exponents of a Platonist reading of Aristotle

was to explain his treatment of the relation between body and soul. This was a
crucial issue on which the two traditions were clearly opposed. However one
might interpret certain passages from the Eudemus and Protrepticus,10 the de
Anima clearly presents us with a soul indissolubly linked to the body by being its
form. While the intellect may be an exception, there is no doubt about the unity
of the rest. For a Platonist this view was, of course, totally unacceptable.

9
That he defended some Platonic views against Aristotle, or adopted Neoplatonist doctrines, has
been claimed on the basis of texts which will not support it, so. S. Landauer, citing in Metaph. 9,3–11
(in his index thereto, s.v. Plato), Schmid-Staehlin, Gesch. Gr. Lit.6 II .ii, Munich 1924, 1011 and n. 8,
citing Or. 32 = 439,4D and 4 (2 is a misprint) = 61, 28, which are not evidence for Neoplatonism. O.
Hamelin, La Théorie de l’Intellect d’après Aristote et ses Commentateurs, Paris 1953, 38–43, regards
Themistius as an ‘Alexandrin’ manqué, but he insists on seeing Platonism where it does not, or need
not exist, so, e.g., the aüla of 115,5 are not evidence for Platonic Ideas.
10
Eudemus fr. 8 Ross, Protrepticus fr. 10b Ross.
 Themistius: the last Peripatetic commentator on Aristotle? 123

Most of Aristotle’s discussion of the actual operations of soul could be

translated fairly easily into Platonist dualism by a realist reading of texts treating
the activities of its faculties.11 But the explicit discussion of the nature and
definition of soul, and its connection with the body, could not be dealt with so
easily. In fact Platonists seem to have found this sufficiently difficult for some at
least to have read Aristotle as if he were talking about a soul and a body that
already had soul: the latter would be the form that makes matter into body, upon
which the individual soul was superimposed. This is what we find in the Platonist
handbook which appears in Plotinus and Porphyry, and which Themistius too
may have used.12 There this view of soul and body appears as an assumption. The
later commentators, whose Platonising of Aristotle was more thorough, tried to
show that only the soul below the level of the individual soul was united to the
body as form to matter, and argued that this manifestation of soul which gives
formless matter the status of body could be described as form and actuality.
Thus Simplicius has phusis (nature), a standard term for this level, as an
informing cause, eidêtikê aitia, to which soul is added (4,14ff ). When he comes
to the discussion of the soul’s definition as the first actuality of a sôma organikon
he uses, or rather abuses, Aristotle’s notion of a first and second actuality to
distinguish one which merely gives form and so life to the body and another
which uses that body, so informed, as its instrument, Platonically (90,29ff ). Thus
he constructively misinterprets sôma organikon to mean not a body fully
equipped with the necessary parts, but a body which is a tool.13 This is in line
with the Neoplatonist concept of the human soul as a series of strata arranged
hierarchically. These layers, in defiance of Aristotle’s notion of form, were
sometimes described as a series of forms, each being form to the one below,
which constituted the matter for its superior. Philoponus, though he tried to
provide a less un-Aristotelian solution, approached the same problem with
similar preconceptions (cf. 198,22ff ).14 For these men the form was not the
essence of any given kind of life which subsumed all lower kinds.
In Themistius there is, with one arguable exception, no sign of any such
attempt to avoid the clear import of Aristotle’s text, and he gives us both the
Aristotelian structure of faculties and its arrangement (e.g. 44,21ff ). If it is argued

11
cf. my Plotinus’ Psychology, The Hague, 1971, 135ff.
12
cf. Plotinus 4.7,1–8, Porphyry ap. Nemesium de Nat. Hom. ch. 2, Themistius 24,22ff and 41,11ff, and
Plotinus’ Psychology 11–13 with nn. 10 and 13; H. Doerrie, Porphyrios’ ‘Symmikta Zetemata’,
Zetemata 20, Munich 1959, 111ff.
13
For further discussion cf. my ‘Some Platonist readings of Aristotle’, PCPhS, n.s. 27, 1981, 4–6.
14
cf. ‘Body and soul in Philoponus’, The Monist 69, 1986, 378–9.
124 Aristotle Transformed

that he was constrained by his genre, and unable to depart radically from Aristotle,
this argument cannot be used about his treatment of the notorious comparison
of the soul with a sailor (413a8f), where Aristotle’s meaning is sufficiently unclear
to allow even a mere paraphrase scope for a Platonist reading. Aristotle’s hesitation
here, his suggestion that perhaps the soul was an actuality of the body like a sailor
in a ship, could easily be referred to areas of soul other than the intellect to which
Aristotle himself seems in the end to have confined it. This was done by Simplicius
and Philoponus, and doubtless by other Neoplatonists whose comments we do
not have. Simplicius professed to be puzzled by how Aristotle could entertain any
doubt about the separability of the soul: for him it was clearly and indisputably
separate (96,8ff ). Themistius, on the other hand, perplexed though he seems to
be, takes Aristotle’s remark as a hesitation, and explains its presence as a feature of
the early stage in the investigation. If soul were an actuality like a steersman, he
says, and it is clear that this applies to nous, then it could be an actuality, but
separable. So far the whole soul had been characterised in outline, but not
accurately. Up to this point, he continues, the discussion is like a sketch until we
turn our attention to all the soul’s faculties: that is how we would achieve clarity
about whether the whole soul, or some part of it, is a separable form (43,25–34).
Since, as we have mentioned, the soul’s activities as described by Aristotle
could be taken as compatible with a Platonist view of its relation to the body,
there is no special occasion for widespread Neoplatonisation in commenting on
the rest of Book 2 or 3.1–2. Difficulties will again present themselves over the
imagination, a faculty which may not always or necessarily involve the body in
its activity. This could lead to problems both about separability, and about
whether or not the soul comes apart, as it did for some Neoplatonists, at this level
just because the same sort of activity related both to sense-perception, with its
clear links with the corporeal, and reason and intellection, where these links
could be held no longer to apply. This, as I have shown elsewhere, is one point
where there is a great deal of discussion of Aristotle in non-Aristotelian ways.15
In particular we find that the later commentators are much concerned to find
clear demarcations between upper and lower soul, a preoccupation of which
Themistius is innocent, and whose absence caused Stephanus to find fault with
his commentary (508,19–25).16 That, briefly, is the answer to our second question.

15
‘Neoplatonic interpretations of Aristotle on Phantasia’, Rev. of Metaphysics 21, 1977, 242–57; for
Themistius’ deviations from Aristotle see now R.B. Todd, ‘Themistius and the traditional
interpretation of Aristotle’s theory of phantasia’, Acta Classica 24, 1981, 49–59.
16
cf. ‘Neoplatonic interpretations’ (n. 15 above), 253–4.
 Themistius: the last Peripatetic commentator on Aristotle? 125

The third we have answered in passing, and now we must turn to the fourth, the
treatment of Aristotle’s chapter on the active intellect.
Here Themistius departs from the paraphrase style and launches on an
extended discussion (98,12–109,3), aimed primarily at resolving the status of the
active intellect. In particular he is concerned with whether or not it should be
identified with the ultimate divine being, the prôtos theos, or perhaps some other.
This aspect of the problem may be understood in the light of Alexander’s well-
known view that the active intellect, nous poêtikos, was in fact the supreme
principle itself. Themistius rejects it (102,30ff ). His own suggestion that it might
be a god (theos) other than the first may most readily be seen as an attempt to
place it in the hierarchy of unmoved movers deployed in Metaphysics 12.8. By
contrast there is nothing about whether it is part of a hypostasis, be it Intellect
(Nous) or Soul, a question which forms the background to the Neoplatonists’
discussions.17 Though they are of course concerned with Themistius’ question
whether or not each intellect is internal to the individual, the later commentators
‘solve’ this problem by saying that Aristotle’s active intellect must be part of the
individual because the subject, skopos, of the de Anima is the rational soul.18 They
were all nevertheless much interested in its relation either to the simple series of
hypostases of the earlier Neoplatonists, or to the complex system of unparticipated
and participated souls and intellects of later Neoplatonism, a system which, let
us note, was originated, if not fully developed, by Iamblichus,19 and so should
have been known to Themistius. He was certainly aware of some Neoplatonic
thought on these matters, for he refers to a question that was of some importance
to them, whether or not all souls were one,20 as one that was discussed by both
earlier and more recent thinkers (cf. 104,15f).
Themistius’ stress on the unity of the actual and potential intellects at
one point in his discussion seems at first not only to be inconsistent with the
view that it was a god – an inconsistency he does not really resolve21 – but also
to exclude the Platonist patterns. But we must remember that they did discuss
whether the active intellect was or was not part of a fully descended intellect,
that is one that is part of the individual soul rather than, as Plotinus and

17
For further details cf. Chapter 13 below, pp. 312–13.
18
cf. Simplicius 240,2ff, [Philop.] 536,2–5, citing Plutarch.
19
cf. e.g. R.T. Wallis, Neoplatonism, London 1972, 100 and 123–34.
20
cf. esp. Plotinus 4.9.
21
On this, and some other points arising from Themistius’ treatment of 3.5, cf. P. Moraux, ‘Le de
Anima dans la tradition grecque. Quelques aspects de l’interprétation du traité, de Théophraste à
Thémistius’, in G.E.R. Lloyd and G.E.L. Owen (eds), Aristotle on Mind and the Senses, Cambridge
1978, 309–13.
126 Aristotle Transformed

some others maintained, a section of the transcendent intellect itself,22 though

it should be noted that the two ‘parts’ of intellect they discussed were more
often seen as intuitive and discursive intellect. So the type of question Themistius
is asking does not in itself exclude Platonist influence on his approach. What
he says in the course of answering it is, however, so different from the
Platonist treatments as to show that he was not thinking in the same conceptual
schemes.23
It should nevertheless be noted that this view is not universally shared,
and before leaving this matter we should look briefly at the evidence that has
been adduced in support of the thesis that, at least in this area of his thought,
we are in fact dealing with a commentator influenced by Neoplatonism. The
main recent proponent of this view has been E.P. Mahoney. In a 1973 article on
Themistius in James of Viterbo and other mediaeval philosophers he drew
attention to certain features of Themistius’ doctrine of the intellect which could
be described as Neoplatonic.24 These are the apparently hierarchical organisation
of the soul’s faculties in a matter-form relation on the Plotinian model, the
ambiguity about the unicity or otherwise of the agent intellect to which we have
already referred, and the description of a plurality of intellects as illuminated
and illuminating (ellampomenoi kai ellampontes). Several further articles on the
influence of Greek commentators on Renaissance philosophy take Themistius as
holding these Neoplatonic views, which the later writers followed. The position
is restated with further arguments, partly in reply to the original version of this
paper, in a long note to the published version of Mahoney’s contribution to a
1978 conference.25 This gives an excellent survey of the opinions on both sides of

22
For a doxography cf. [Philop.] 535,2ff.
23
A further text from outside the de Anima which would give ample scope for Neoplatonising
distortion is Metaph. 12: Themistius in his paraphrase characterises the human mind as weak and
inefficient, which would not be acceptable to a commentator who thought each human mind
capable of assimilation to the transcendent nous, and of thereby attaining its power, in Metaph.
32,18ff.
24
‘Themistius and the agent intellect in James of Viterbo and other thirteenth-century philosophers’,
Augustiniana 23, 1973, 423–67: on Themistius cf. esp. 428–31. Here, and in ‘Neoplatonism, the
Greek commentators . . .’ (see n. 25 below) Mahoney indicates that he started from suggestions
made by P.O. Kristeller.
25
‘Neoplatonism, the Greek commentators, and Renaissance Aristotelianism’, in D.J. O’Meara (ed.),
Neoplatonism and Christian Thought, Studies in Neoplatonism: Ancient and Modern 3, Albany
NY 1982, 169–77 and 264–82, cf. esp. n. 1 on pp. 264–6. For his subsequent treatments of the use
of the Greek commentators in the Renaissance, cf. ‘Philosophy and science in Nicoletto Vernia
and Agostino Nifo’, in A. Poppi (ed.), Scienza e filosofia all’ Università di Padova nel Quattrocento,
Contributi alla storia dell’ Università di Padova 15, Trieste 1983, 135–202; ‘Marsilio Ficino’s
influence on Nicoletto Vernia, Agostino Nifo and Marcantonio Zimara’, in G.C. Garfagnani (ed.),
Marsilio Ficino e il ritorno del Platonismo, Florence 1986, 509–31: neither, however, contains direct
discussion of the text of Themistius.
 Themistius: the last Peripatetic commentator on Aristotle? 127

this question and recommends a compromise view that, while Themistius’ work
contains some Neoplatonic views, he is not a faithful representative of that
philosophy.26
While Mahoney’s comments have caused me to see that in the original version
of this article, as opposed to some earlier discussions of these matters,27 I had
underemphasised the extent to which Themistius uses Neoplatonic vocabulary,
I would maintain that this is primarily a matter of language, and that even
his noetic, the area in question, is still fundamentally free of Platonism – except,
of course, in so far as Aristotle himself was not. To take the points at issue
in turn:

(1) The most obviously Neoplatonic-looking part of Themistius’ discussion is

that in which he talks about the successive faculties of soul being related to those
below them in each case as form to matter (100,32–5). That is an idea that can, of
course, be found in Plotinus.28 What is less clear is that it is not a possible
interpretation of Aristotle himself: one might think of the relation of wood to
matter, plank to wood, and ship to plank. At the very least it is a possible Peripatetic
solution to the problems, highlighted for example by J.L. Ackrill,29 inherent in
Aristotle’s concept of the soul as form of what already has form, though it might
be argued on the other side that such a solution would give rise to problems
analogous to the notorious difficulty about the plurality of unmoved movers in
Metaphysics 12.
(2) A better candidate for Neoplatonist influence here is the notion that
intellect could be both one and many, a view that is, of course, unproblematical
for a Neoplatonist, but has caused sufficient offence to some Aristotelian scholars
for them to suggest that the text of Themistius is here corrupt.30 That, as Mahoney
rightly says, is not a satisfactory solution. In any case caution is again called for:
what we have could be simply an implication of Alexander’s view that there is
only one active intellect: in so far as that operates in all individuals it must also,
at least conceptually, be plural.

26
Following A. Guzzo and V. Mathieu s.v. ‘Intelletto’ in vol. 3 of the Enciclopedia Filosofica2, Florence
1967, 963 (Mahoney’s reference).
27
cf. the articles cited in nn. 5 and 15 above.
28
cf. p. 116 above and the selection of references given by Mahoney, ‘Neoplatonism, the Greek
commentators . . .’ (n. 25 above), 266.
29
‘Aristotle’s definitions of psuchê’, Proc. Aristotelian Soc. 73, 1972–3, 122–7, repr. in J. Barnes,
M. Schofield and R. Sorabji (eds), Articles on Aristotle 4. Psychology and Aesthetics, London 1979,
67–70.
30
So P. Merlan, Monopsychism (see n. 4), 50 n. 3, following A. Kurfess, Zur Geschichte der aristotelischen
Lehre vom sog. NOUS POIÊTIKOS und PATHÊTIKOS, Tübingen 1911 (repr. in A. Kurfess and
I. Düring, Aristotle and His Influence: two studies, New York 1985), 23.
128 Aristotle Transformed

(3) When we come to language, Mahoney rightly points to what is apparently

the strongest indication of Neoplatonic ways of thinking, namely the reference
to illuminated and illuminating intellects. But on closer examination this is less
clearly Neoplatonic. In the first place Themistius ties it to Aristotle’s comparison
of intellect with light in de Anima 3.5, and uses the characteristics of illumination
to explain Aristotle’s choice of light rather than Plato’s of the sun in the Republic
(103,34–5). It might be worth adding that, even though ‘illumination’ (ellampsis)
is a Neoplatonic rather than an Aristotelian term, the word does occur once in
the Aristotelian corpus.31 In the work On Plants pseudo-Aristotle writes that
sense-perception causes the illumination of life: hê gar aisthêsis aitia estin
ellampseôs zôês (815b33). Though there is no evidence, as far as I know, that this
work was familiar to the tradition, it is interesting to note that this is just the sort
of remark about the relation of one level of soul to the next that one normally
expects to find in a Neoplatonist: we cannot, however, be sure that ellampsis was
in the original text.

A further item of language, which Mahoney might have added, but perhaps
did not because it is not necessarily significant, is the use of the common
Neoplatonic terms for the operation of discursive reason in its transition from
one object to another, namely metaballô and metabainô (100,7 and 11). They are
to be found over and over again in the later commentators’ discussions of
reasoning, but not in Aristotle’s; what they describe, however, is not in any way
un-Aristotelian.
All this shows that such Neoplatonic influence as there may have been was
marginal in matters of doctrine, and only a little greater in features of language.
These conclusions are reinforced by a wider look at Themistius’ language. Any
reader familiar with Neoplatonist writing must notice the almost complete
absence of typically Platonic and Neoplatonic terminology. While one should
not expect the fully developed vocabulary of fifth-century Neoplatonism, there
is enough in Plotinus, Porphyry and particularly Iamblichus to obtrude itself if
it appeared. It could be objected that the paraphrase form did not lend itself to
the introduction of un-Aristotelian terms, but while this objection might apply
to the passages on de Anima 2, it cannot apply to the discussion of 3.5, where
Themistius produces ordinary commentary. Given the control afforded by this
chapter, where, if anywhere, one would expect Platonism and Platonist language,
we may safely say that their absence is neither insignificant nor fortuitous.

31
In a work whose original was probably by Nicolaus of Damascus: The extant text is a retranslation
from Latin, cf. P. Moraux, Les listes anciennes des ouvrages d’Aristote, Louvain 1951, 109.
 Themistius: the last Peripatetic commentator on Aristotle? 129

Before turning to the later commentators we should notice a problem which

their references to Themistius have been taken to present. Some of them do not
correspond closely, or at all, with the text of the paraphrases. On the basis of this
difficulty, of a report in Photius that there were commentaries as well, and of
remarks by Themistius himself in Or. 23, it has been argued that he wrote a series
of full commentaries, and that the later commentators sometimes refer to these.
If this were correct, it would have important implications, for clearly in such
works Themistius would have had more scope for departures from Aristotelianism
than within the constraints of paraphrase. I think, however, that we may safely
assume that the picture presented by the extant paraphrases is accurate, and that
untraceable references call for other explanations.32
One further preliminary point: the commentators’ procedures can easily
give a misleading impression of their attitudes, since they will most often refer
to earlier opinions only when they wish to show that these are mistaken, or
use them as ammunition in a controversy. Where a matter is uncontroversial,
they will often simply give the standard explanation – a procedure, after all, not
unlike ours.
The largest number of later named references to Themistius is in Simplicius’
commentary on the Physics. Apart from Books 7 and 8, this treatise is not as open
as some to Neoplatonising.33 But it is of interest that only one text in Simplicius’
commentary shows Themistius clearly opposed to Aristotle. That is on the
independent existence of time (765,31ff ). Otherwise Themistius is treated in the
same way as Alexander, with whom he is recorded as agreeing and disagreeing
more or less equally often.34 Simplicius himself agrees with him more often than
not, but since the points at issue are often philological rather than philosophical,
and most of the latter are not relevant to differences between Plato and Aristotle,
one can find no evidence here for Themistius as a Platonist. In the de Caelo
commentary, on the other hand, Simplicius says in so many words that Themistius
generally follows Aristotle, but occasionally, as on movement up and down,
prefers Plato’s account (69,9ff ). The remaining references there are to scientific
matters. In the de Anima commentary, whose different style excludes frequent

32
A case for commentaries is made by C. Steel, ‘Des commentaires d’Aristote par Thémistius’, Rev.
philosophique de Louvain 71, 1973, 669–80; for a critique cf. my ‘Photius on Themistius (Cod. 74):
Did Themistius write commentaries on Aristotle’, Hermes 107, 1979, 168–82. Professor Steel has
since indicated (by letter) that he no longer believes in the existence of separate commentaries.
33
As Simplicius noted in his own commentary, 1036,14–17, Themistius did not pay much attention
to 7.
34
On Simplicius’ use of Alexander see my ‘Alexander of Aphrodisias in the later Greek commentaries
on Aristotle’s De Anima’, in J. Wiesner (ed.), Aristoteles Werk und Wirkung, vol. 2, Berlin 1987,
90–106.
130 Aristotle Transformed

references to others, there is only a single named reference to Themistius, to his

view that the objects of taste are only the painful and pleasant, whereas Simplicius
thinks taste gives other kinds of information too.35
Philoponus’ Physics commentary, like Simplicius’, gives little that is to our
purposes, and the few references in in An. Post. are also primarily technical,
showing no symptoms of Platonising. The part of the de Anima commentary
written by Philoponus himself is again unhelpful. On whether the sense of touch
is one or many, Themistius is reported as disagreeing with Aristotle, and
Philoponus follows him, but this has nothing to do with Platonism. The same is
true of the question whether flesh is an organ or a medium, where Themistius
agrees with Aristotle (418,24f) while Philoponus expresses dissatisfaction with
the whole discussion (433,15ff ).
The third book of the Philoponus commentary, written by Stephanus, gives us
more, and, as we have mentioned, indicates that Themistius’ interpretation was
seen as inadequately Platonic. One report, on his argument for the need for a
soul to have all the irrational faculties before it can have reason, has him giving
an argument which is not Aristotelian, but the position it maintains is (450,8ff ).
All this pertains to the discussion of how many senses we have, and Stephanus
himself goes on to offer a professedly more Platonic explanation of why we have
five (ibid. 27–33). Another citation, difficult to trace, has Themistius talking of
imagination as an active power. This does look like Neoplatonising, but the very
fact that it does not appear in the paraphrases excites suspicion that either
Stephanus, or a source, has introduced the Neoplatonic element himself. It is
interesting that he was not content with Themistius’ account, in so far as he, like
Aristotle, said sensation and imagination happen together, a view not conducive
to the hierarchical picture of the soul favoured by the late commentators.
Stephanus proceeds to criticise Themistius for a view he did not express, namely
that imagination can only operate simultaneously with sense-perception.36
One earlier report should be added. Ammonius, the teacher of both Simplicius
and Philoponus, tells us how Julian gave a ruling against Themistius in a dispute
on whether 2nd and 3rd figure syllogisms are perfect, in favour of Maximus, who

35
The authenticity of this commentary is questioned, and its attribution to Priscian suggested, by
F. Bossier and C. Steel, ‘Priscianus Lydus en de in de Anima van Pseudo(?)-Simplicius’, Tijdschrift
voor Filosofie 34, 1972, 761–822; for another view cf. I. Hadot, Le problème du néoplatonisme
Alexandrin. Hiéroclès et Simplicius, Paris 1978, 193–202 and my ‘The psychology of (?) Simplicius’
commentary on the De Anima’ in H.J. Blumenthal and A.C. Lloyd (eds), Soul and the Structure of
Being in Late Neoplatonism, Liverpool 1982, 73–5. Professors Bossier and Steel have both since
(1985) told me that they are still fully convinced that this commentary is by another hand.
36
cf. 508,19ff and n. 16 above.
 Themistius: the last Peripatetic commentator on Aristotle? 131

had heard Iamblichus, and followed him and Porphyry (in An. Pr. 31,15–22).
Even if this is not a matter where a Platonic view could have been adopted, the
tale is further evidence for Themistius’ independence of the Neoplatonists.
Thus both the content of Themistius’ works, and such evidence as we have of
the commentators’ attitudes to him, show that he was predominantly a Peripatetic.
In this he stood out against the tendencies of his time. His frequently expressed
admiration for Plato does not invalidate this conclusion. Themistius may rightly
claim to have been the last major figure in antiquity who was a genuine follower
of Aristotle. For him, unlike his contemporaries, Plato does not surpass the
master of those who know but he, and Socrates, ‘innanzi agli altri più presso gli
stanno’.
132
 6

The harmony of Plotinus and Aristotle

according to Porphyry*
Pierre Hadot

The objections which Plotinus raised against Aristotle’s theory of categories have
played a major role in the history of logic. They required Neoplatonists wishing
to save a place in their system for Aristotelian logic truly to exert themselves in
the logical and metaphysical interpretation of the Categories.
It is well known that the second and third books of Dexippus’ commentary
on the Categories are almost totally devoted to the resolution of Plotinus’
objections. The solutions which Dexippus proposes to Plotinus’ objections are
borrowed, on the one hand, from the great commentary which Porphyry wrote
on the Categories and addressed to Gedalius (now unfortunately lost), and, on
the other hand, from Iamblichus’ commentary on the same work of Aristotle.
Unfortunately, when quoting from these two sources, Dexippus does not name
them. They can occasionally be identified by comparing his text with that of
Simplicius, who generally uses the same authors in commenting on the
Categories, except that he names them (sometimes, at least). Simplicius tells us
(in Cat. 2,9) that Iamblichus’ commentary often followed Porphyry’s word for
word, but also corrected it, abridged it, and added to it both important theoretical
discussions and, notably, a comparison between Aristotle’s Categories and the
treatise On the Whole of Archytas (whom Iamblichus considered to be Aristotle’s
source). This last feature of Iamblichus’ commentary permits us in certain cases
to trace its influence.
The present essay is confined to several lines of Dexippus (in Cat. 40,13–42,3)
which relate to a Plotinian aporia that is particularly important, since it concerns
the concept of ousia (substance).

* Translated by Victor Caston.

133
134 Aristotle Transformed

Here is the text of Plotinus’ objection (from his treatise ‘On the genera of
being’; 6.1.2.4–8) as Dexippus reports it:

If there are two ousiai (one intelligible, the other sensible), how can they be
traced to a single genus? For what do these two ousiai have in common as regards
being? And, even if we assume that there is something in common, it will still be
necessary to posit something prior and different from the two, which is neither
a body, nor an incorporeal (in order to explain their commonality). It follows
from this that either the incorporeal is a body or body is an incorporeal.
(Dexippus, in Cat. 40,14–18)

To be properly understood, Plotinus’ aporia must be placed within its original

context. The fundamental question Plotinus poses is the following: How can
Aristotelian substance claim to be a single genus? On the one hand, Aristotle does
not speak of intelligible substance, even though that is what is at the highest level
(6.1.1.28). On the other hand, even if one wanted to give intelligible and sensible
things both the name ‘substance’, it would not be possible, for intelligible being is
essentially different from sensible being. Moreover, even if it were agreed that the
intelligible and the sensible were differentiae of a single genus, it would be
necessary to posit something prior to the intelligible and the sensible, which was
both intelligible and sensible, incorporeal and corporeal. Yet, even if restricted to
sensible substance, it is unacceptable that the concept of substance should be truly
common to the realities to which one attributes it. If Aristotle can call form and
matter, as well as the composite of the two, ‘ousia’, it must be admitted that these
realities are not equally substances, but that form is ‘more’ substance (6.1.2.8–12).
The general sense of Plotinus’ objection is the following. If one enumerates all
ousiai – the intelligible ousia (which Aristotle does not mention in the Categories,
yet recognises elsewhere), the ousia which is form or which is matter, the ousia
which is composed of the two (form and matter) – one must admit that the term
‘ousia’ is not employed synonymously (in the same sense), but homonymously (in
different senses). These different ousiai do not, therefore, belong to the same genus.
To this objection, Dexippus’ source gives a preliminary response:

Obviously, it is easy to respond to these objections by saying that the difficulties

Plotinus raises are foreign to the intention of the Categories. For in the present
case, Aristotle proposed to discuss neither the beings nor the genera of first
ousia; he has beginners in mind, who can only understand the most basic lessons.
Thus, as the actual discussion concerns words which can be said of ousia (kat’
ousiôn legesthai), it is unreasonable of Plotinus to introduce a discussion of
beings into a discussion which concerns words. (Dexippus, in Cat. 40,19–25)
 The harmony of Plotinus and Aristotle according to Porphyry 135

In this text we find Porphyry’s customary argument when refuting objections

of this kind: Aristotle’s Categories is addressed to beginners and only treats
meaningful words, not beings (cf. Porphyry in Cat. 56,31; Simplicius in Cat. 10,20).
Yet Dexippus’ source wishes to go further in refuting Plotinus and grasp the
opportunity which this occasion provides to treat the problem of substance in
depth. It does not necessarily follow a priori that the author Dexippus copied could
not be Porphyry merely from the fact that the author he uses wishes to go beyond
arguments based on the intention of the Categories and to devote an important
discussion to substance considered from a metaphysical point of view. Porphyry’s
greater commentary on the Categories contained investigations of this kind (cf.,
for example, Simplicius in Cat. 158,27). In any case, Dexippus’ text continues:

It does not seem right, however, thus to avoid this extension of the argument, an
extension which is well adapted to our present inquiry. On the contrary, it seems
desirable to start from the same point as the philosophy of Plotinus does and
relate the extension of the present arguments to the whole of his doctrine.
Plotinus in fact posits ousia as a genus unique among intelligible realities because
it procures being for incorporeal forms universally and gives being to all the
forms which are sensible and blended with matter. If that is so, the principle of
ousia extends across all things, taking successively the first, second, and third
rank insofar as it gives being to one primarily and to others in another manner.
This is why, if everything leads to this principle of ousia (since everything is
suspended from it), Aristotle’s description of ousia can also provide a glimpse of
the first principle of ousia, from which the ousia has fallen to its lowest degree.
(Dexippus, in Cat. 40,25–41,3)

Here we have a fascinating text. It attempts to justify Aristotle with the help of
the philosophy of Plotinus and to show that, at bottom, Aristotle’s logic can find
room within the ambit of the Plotinian system.
To begin with, notice the appearance of the phrase the ‘philosophy of Plotinus’.
In the Neoplatonic period, it is generally the ‘philosophy of Plato’ or the
‘philosophy of Aristotle’ which is spoken of – that is, this phrase is only used for
the doctrine of the great masters. The formula which Dexippus uses betrays the
enthusiasm of a pupil for his teacher, recalling the title of a work by Amelius in
which he speaks in praise of Plotinus’ doctrine: ‘On the proper character of the
philosophy of Plotinus’ (Porphyry, Vita Plotini 20.101). We can assume that the
pupil whose text Dexippus has preserved is Porphyry rather than Iamblichus;
the latter in fact had a tendency to distance himself from Plotinus.
This assumption appears to be confirmed by an examination of the text of
Simplicius’ commentary (in Cat. 76,25) which corresponds to the discussion of
136 Aristotle Transformed

Dexippus under consideration. In this parallel passage, ‘the philosophy of

Plotinus’ disappears, and the philosopher who posits substance as a genus unique
among intelligible realities is no longer Plotinus, but Plato. This substitution
seems to have been made in fact to eliminate the overly enthusiastic mention of
the ‘philosophy of Plotinus’ and is quite probably the work of Iamblichus. Note
in this regard that the substitution of Plato for Plotinus destroys the force of the
argument. It was very interesting to say that Plotinus’ own philosophy affords a
resolution of Plotinus’ objection against Aristotle. Yet it is far more banal to
remark, as Simplicius does, that Plato considers substance as an intelligible
genus. In wishing to replace Plotinus with Plato, Iamblichus has completely
distorted the sense of the reasoning.
Two questions may now be posed: (1) Does Dexippus give us an exact
summary of Plotinus’ doctrine? and (2) How does this evocation of Plotinus’
philosophy permit a resolution of Plotinus’ aporia?
To answer the first question is already to answer the second in part. In fact,
Plotinus himself had proposed a solution to his own anti-Aristotelian objection,
and this solution agrees with the résumé of the Plotinian doctrine of substance
which Dexippus offers.
In fact, after having refused to accept that ‘ousia’ can be attributed
synonymously (in the same sense) to all the ousiai Aristotle acknowledges,
Plotinus makes the following assumption, representing his attempt to preserve
the unity of the concept of ousia: ‘Yet if we collect the intelligible “substance”, the
matter, the form, and the composite of the two together, is it not necessary to say
that this is a single category? It is in this manner that someone would say that the
genus of the Heraclides is a single thing; this sort of unity would not be a
common predicate attributed to all, but would result from the fact that all take
their origin from a single thing. Intelligible ousia in this case would be ousia in a
primary sense, the others in a derivative and lesser sense’ (6.1.3.1–5).
According to the hypothesis which Plotinus proposes, the unity of the concept
of substance would thus correspond to the type of unity that exists between
terms which are neither totally synonymous nor totally homonymous, but whose
common name is based on a single principle (Metaph. 4.2, 1003a33; 7.4, 1030a35).
The single principle would here therefore be intelligible ousia, which would be
the source of being for the form, the composite, and the matter. Porphyry’s
Isagoge (1,18 Busse) offers, like Plotinus, the example of the Heraclides in order
to define genus: ‘One calls genus the collection of certain terms which have a
determinate relation with respect to some one thing, and each with respect to the
others. It is in this sense that one speaks of the genus of the Heraclides in virtue
 The harmony of Plotinus and Aristotle according to Porphyry 137

of their relation to a first (or primary) principle, I mean Heracles.’ There will,
however, be a certain ambiguity here, since both the group of terms connected
by this relation and their principle will equally be called ‘genus’ (Isag. 1,23ff ). The
genus is therefore both the principle which grounds the community of name and
the group of terms which have this name in common. From Plotinus’ Platonic
perspective, intelligible ousia – which is the principle of the community of name
between the different ousiai – is not just any origin or generator whatsoever; it is
a principle because it is ousia in itself, it grounds being, because it is the idea
itself of ousia. It is a genus in the Platonic sense, that is, an intelligible reality
which exists in and for itself, which is not attributable to subjects lower than it,
but grounds the very possibility of predicates being attributed to subjects in the
beings which participate in them.
This conception of ousia does exist in Plotinus. One encounters a preliminary
version of it in the treatise ‘On the ousia of the soul’ (4.2.1.29ff ). There Plotinus
comments on Timaeus 35A: ‘From the ousia which is indivisible and always
identical with itself and the ousia which becomes divisible in the body, the
Demiurge makes a third kind of ousia by blending them’. Plotinus identifies
indivisible ousia with intelligible reality, which is compared to the undivided and
unmoving centre of a circle, from which the rays shoot forth and derive their
being (4.1.27). Divisible ousia corresponds to material forms; intermediate ousia
is the ousia of the soul. There is a hierarchy of ousiai, therefore, which are
grounded in an original ousia and progressively descend on account of distance
and division. Yet it is above all in the two treatises dedicated to the theme ‘That
what is one and identical can be at the same time in its entirety everywhere’
(6.4–5) that Plotinus develops a doctrine of intelligible ousia the most clearly. If
we remain at the level of logic, Plotinus tells us (probably with Aristotle’s
Categories in mind), the nature of bodies serves as our model and our principles
are taken from there, so that if we come hence to divide ousia, we no longer
believe in its unity. If, on the contrary, we start from principles proper to
intelligible ousia, we understand that it is an indivisible unity, which is not
divided in the subjects that participate in it, but always remains identical with
itself, while being present everywhere in its entirety (6.5.2.1–3, 32). In the same
way, in the treatise ‘On numbers’ (6.6.13.27), Plotinus does not hesitate to affirm
that the name ‘ousia’ applies to sensible things by derivation from intelligible
things. If one applies the category ousia to sensible things and to intelligible
things, it is nevertheless the latter to which the category is attributed in the
highest degree. Ousia and being are, by themselves, intelligible and not sensible,
even if the sensible participates in them. Lastly, the treatise ‘On the genera of
138 Aristotle Transformed

being’ (6.2.7.6) explicitly presents intelligible ousia as a genus along with the
other Platonic genera – movement, rest, identity and difference – by specifying
(6.2.8.25) that what comes after these genera are the particularised genera:
determinate ousia, determinate movement, etc.
One can therefore find in Plotinus elements of the doctrine of ousia which
Dexippus (probably following Porphyry) attributes to him. Yet it must be
recognised that the formulas Dexippus uses are not at all the same as those in
Plotinus. The text of Dexippus systematises the teaching of Plotinus and, in
systematising it, causes it to undergo an important transformation. Intelligible
ousia is there defined as a principle of ousia, that is, as a principle of being in the
three degrees (taxis) of reality: the intelligible, the psychic, and the sensible
levels. To each degree there corresponds a type of being (primary being,
derivative being) and a type of ousia. As regards this participation of the lower
degrees in the principle of ousia, it can be said that the principle of ousia extends
through all things, though this only means that everything has its origin and end
in it. The hierarchy of the degrees of being and of ousia corresponds to a
progressive degradation.
This presentation is completely analogous to the systematisation of Plotinian
theses in the theorems of Porphyry’s Sententiae. It might be worthwhile to
emphasise the correspondence between Dexippus’ text and certain tendencies
that come into relief in the Porphyrian systematisation (in the Sententiae and in
other works, fragments, and testimonia). The idea of a first substance, a principle
of the being of things, appears to be attested under Porphyry’s name in John
Lydus (de Mens. 138,18 Wünsch): ‘Hestia is the substance-source, the cause of
being for all things, situated in the Father.’ The principle of the degradation of
ousia, as it were, recurs very frequently, as for example in Sententiae 13: ‘everything
engendered is worse in its substance than that which engendered it (pan to
gennôn têi ousiâi autou kheiron autou gennâi)’, and recapitulated in ad Gaurum
54.11, 42.18 (Kalbfleisch): ‘The products issued from the substance of a thing are
always lower in rank with respect to power and substance than that which
engendered them; they cannot be of the same ousia as the first generators’. This
degradation of ousia in passing from the generator to the engendered corresponds
to a progressive materialisation, as expressed in the following text of Porphyry,
cited by Proclus, in Tim. 1.439.30 (Diehl): ‘To the extent that intelligible ousia
descends towards the world, it leads to a state of divided, dense, and material
multiplicity, although above it was unified, indivisible, and single.’ In this regard,
Sententiae 11.3.5 (Mommert) speaks of a diminution (huphesis) of power in the
progression towards individuality. The degrees of reality are presented as a
 The harmony of Plotinus and Aristotle according to Porphyry 139

hierarchy of ousiai (Sententiae 5, 10, 17, 22) and great importance is given to the
manner in which each degree of ousia acquires its being (kektêtai to einai;
Sententiae 4.9, 6.2, 27.4, 40.7 Mommert). This ontology corresponds rather well
to the Porphyrian ontology as characterised by W. Theiler in his work Porphyrios
und Augustin (Halle 1933, 11ff ). It can in fact be traced again in Augustine, for
example in the following text, which in a way appears to summarise what we
have just read: ‘For since God is the highest essence, that is to say, he is in the
highest sense and is therefore unchangeable, he gave being to those things which
he created from nothing, though not being in the highest sense (such as his
own), but gave more being to some and to others less, and so ordered natures by
degrees of being (Cum enim deus summa essentia sit, hoc est summe sit, et ideo
immutabilis sit, rebus quas ex nihilo creavit esse dedit, sed non summe esse, sicut
est ipse, at aliis dedit esse amplius, aliis minus, atque ita naturas essentiarum
gradibus ordinavit)’ (de Civ. Dei XII .2).
The second question we were to answer was: how does this evocation of the
Plotinian doctrine of ousia permit the resolution of Plotinus’ aporia concerning
the unity of the concept of ousia? As mentioned above, the Plotinian doctrine of
ousia allows different ousiai to be conceived of as deriving from and referring to
a single principle. This is precisely what Dexippus’ text asserts: ‘If everything
leads back to this principle of ousia because everything is suspended from it, the
description of ousia (by Aristotle) can also provide a glimpse of the first principle
of ousia.’ The expressions ‘leads back’ and ‘suspended’ evoke the relation ‘from
one and to one’ (aph’ henos kai pros hen): the different levels of ousiai come from
a single principle and refer to a single principle. This relation is what makes
knowledge by analogy possible, and, as we shall see in the continuation of
Dexippus’ text, this is in fact the type of knowledge envisaged. This principle of
analogy permits Aristotle’s description of sensible ousia to be used to provide a
glimpse of the original ousia. It is indeed a matter of description and not of
definition (cf. Dexippus, in Cat. 44,11; for a clear distinction between definition
and description, cf. Porphyry, ap. Simplicium in Cat. 30,13–15). The description
of ousia in question is the one which Aristotle gives at Categories 2a11: ‘Ousia in
the most fundamental, first, and principal sense of the term is that which is
neither said of a subject nor in a subject.’ Dexippus in fact refers to this text
several times when speaking of the description of ousia (cf. 42,2; 44,15 and 25).
The Aristotelian description ‘neither said of a subject nor in a subject’ therefore
gives us a glimpse of intelligible ousia. One can assume that, for Porphyry, these
negative phrases take on a positive and fuller sense when referring to intelligible
ousia. ‘Not to be in a subject’ means ‘to be in itself and by itself ’, as he states in his
140 Aristotle Transformed

Sententiae 39.35.2 (Mommert): ‘ “Being always at rest within itself ” is predicated

of the real being which subsists by itself (tou de ontôs ontos kai kath’ heauto
huphestêkotos aülou to einai aei en heautôi hidrumenon)’. One encounters traces
of this transposition again in Simplicius, in Cat. 76,3: ‘If the ousia which subsists
by itself (kath’ hautên huphestôsa) has need of nothing else and if all others have
need of it, if this ousia communicates being absolutely to all the others, it is fitting
that this ousia should be more honoured than the others.’ With regard to ‘not to
be said of a subject’, in the case of intelligible ousia this phrase means that it is
absolutely one and identical, since Categories 1b6 states that what is numerically
one is never said of a subject. Now, according to Plotinus himself, intelligible
ousia is ‘one and the same in number, not partitioned, but whole (hen kai tauton
arithmôi mê memerismenon, alla holon on)’ (6.5.3.19). One can thus say that the
individuality and subsistence of sensible ousia are, as it were, only a reflection
and trace of the indivisibility and the independence of intelligible ousia.
Dexippus’ commentary then continues by proposing a new objection and a
response to that objection (41,4). They are both inseparable from the discussion
just considered. The latter, in fact, showed that from the perspective of Plotinus’
philosophy the unity of the concept of ousia was acceptable. Yet to this solution
Dexippus immediately raises the objection that one would expect, namely, that
Plotinus’ system has different premises from Aristotle’s system. It would be
necessary to begin with Aristotelian, and not Plotinian, premises in order to
examine whether the Aristotelian concept of ousia offers a true unity of content:

Had Aristotle used the same premises as Plotinus, what was just claimed would
appear somewhat reasonable. Yet in actual fact it is by taking [reading prolabôn
with MSS AM ] Plotinus’ dogmas for granted that you defend Aristotle.
(Dexippus, in Cat. 41,4–6)

The answer to this objection will consist in showing that, correctly understood,
Aristotle’s philosophy agrees with Plotinus’ philosophy, and that there is in
Aristotle, as in Plotinus, a concept of an intelligible ousia which grounds both the
being and the concept of sensible ousia. The present discussion, devoted to
Aristotle, has the same structure as the one preceding devoted to Plotinus: (1)
Aristotle accepts the existence of an intelligible ousia; (2) this intelligible ousia is
the principle of unity for the different ousiai which come after it; and (3) this
intelligible ousia, unknowable in itself, is known to us by analogy with sensible
ousia. Here is the portion of Dexippus’ text that corresponds to the first two points:

Now for this question, I will appeal to what is said in the Metaphysics. For there
are two ousiai according to Aristotle, the intelligible ousia and the sensible ousia;
 The harmony of Plotinus and Aristotle according to Porphyry 141

intermediate between these two is physical ousia. The ousia which is composite
is sensible ousia; the ousia reduced to matter and form is physical ousia; the
ousia above these is intellective and incorporeal ousia, which Aristotle often calls
‘unmoved’ on the one hand, and ‘motive’ on the other hand, in so far as it is the
cause of specifically living movement.
This is in fact what Aristotle demonstrated in Book 12 of the Metaphysics
concerning these ousiai. And in this treatise he brought together the several
ousiai into the total ousia. For he arranged them in a single system and traced
them back to a single principle. For otherwise what will participate in the One, if
ousia itself – that which has its being in the One – is deprived of the internal
cohesion which is attributed to the One? (Dexippus, in Cat. 41,7–18)

Let us say straight away that this discussion appears to be the work of Porphyry,
as did the preceding one. In fact, from parallels to be found in Simplicius (77,4ff ),
we can again confirm the distortion which Iamblichus has inflicted on Porphyry’s
text. Dexippus’ source refers to Book 12 of the Metaphysics and tells us explicitly
that what it reports concerning the Aristotelian doctrine of ousia originates from
this treatise. In this book of the Metaphysics (1069a30ff and 1071b3), one does in
fact find the classification of ousiai which Dexippus’ source mentions: a
distinction between an unmoved ousia and a sensible or physical ousia, and then
a distinction within the latter, that is, between sensible, corruptible ousia and
truly physical, eternal ousia. The two sensible ousiai are the subject of physics,
whereas unmoved ousia is the subject of a higher science. Now, in Simplicius the
Aristotelian classification of ousiai distinguishes an intelligible ousia, a
mathematical or psychical ousia, and a sensible ousia. This is the first distortion:
this division is not to be found in book 12, but in Books 6 (1026a6–19) and 11
(1064b1–3). This division, and especially the close association of mathematical
ousia and psychical ousia, is particularly dear to Iamblichus, as P. Merlan has
shown (From Platonism to Neoplatonism, the Hague 1960, 11–33). Indeed, it
appears that Iamblichus might have considered it unworthy of a Platonist to
accept a classification of ousiai which did not make a place for the soul and
mathematica. Porphyry, on the contrary had probably retained the classification
of ousiai into the sensible, the physical, and the intelligible, because, as can be
seen from the text of Dexippus, it allowed the systematisation of the different
kinds of ousia which Plotinus, following Aristotle, had distinguished: composite
ousia corresponded to sensible, corruptible ousia; ousia reduced to matter or to
form corresponded to physical, incorruptible ousia, since matter and form, being
incorporeals, were incorruptible and since, as they composed sensible ousia, they
were the subject par excellence of physical science. Finally, unmoved ousia
142 Aristotle Transformed

corresponded to intelligible and intellective ousia, since Aristotle showed in

Book 12 that this ousia at its highest level was the Thinking of Thinking.
By substituting for the classification of Book 12 a classification which made
room for mathematical ousia, Iamblichus was no longer able to perform the
systematisation which has just been described: ousia considered as matter or
as form could not be assimilated to mathematical ousia. Iamblichus found
in Porphyry’s commentary the systematisation which makes composite ousia
correspond to sensible ousia, ousia as matter or form to physical ousia, and
intelligible ousia to unmoved motive ousia. No longer able to present this
systematisation as Aristotle’s as Porphyry had done, Iamblichus (77,7) did
not hesitate to attribute it to Archytas, who actually had spoken (cf. Simplicius,
in Cat. 76,20) of a sensible ousia and a physical ousia (without, however,
distinguishing one from the other), as well as of an ousia conceived by the mind,
though not, it appears, having said anything about a motive ousia, and still less an
unmoved ousia.1 Here we see the second distortion which Iamblichus has
inflicted on Porphyry’s text, by attributing to Archytas what Porphyry had
presented, not unjustly, as Aristotelian.
To repeat, the distortions Iamblichus inflicted on Porphyry’s text cause
Porphyry’s argument to lose all its vigour. What Porphyry wanted to show was
that the systematisation of ousiai in Book 12 was not incompatible with the
Plotinian doctrine of ousia. For this project it was completely irrelevant to show
that Archytas also traced ousiai back to a single principle: this has nothing to do
with the problem in question.
We know from Simplicius (in Cael. 503,34 and 506,13) that Porphyry had
commented on Book 12 of the Metaphysics. The several lines we have just read in
Dexippus allow us to form a general idea of this commentary. We saw how
Porphyry drew a systematic correlation between sensible ousia and composite
ousia, physical ousia and matter and form, unmoved ousia and intelligible ousia.
This represents a certain systematisation and transformation of Aristotle’s
doctrine. In fact, Aristotle does not speak explicitly of physical ousia, but
distinguishes two degrees of sensible ousia: corruptible sensible ousia and
incorruptible sensible ousia. The first actually corresponds to ousiai composed
of matter and form – Aristotle himself gives as an example plants and animals
(1069a31) – incorruptible, or eternal, sensible ousia corresponds to the ousiai of
the celestial, superlunary world. In Porphyry, on the other hand, the first category
remains unchanged, but the second is called ‘physical ousia’ and is considered as

1
Parallel texts of Simplicius, in Cat. 77, 7–10 and 76,17–22 and Dexippus, in Cat. 41,9–12.
 The harmony of Plotinus and Aristotle according to Porphyry 143

the subject par excellence of the Physics. The latter, however, is no longer a matter
of astronomy, as it is in Aristotle, but of the ultimate principles of the physical
world, that is, matter and form. This is why Porphyry no longer speaks in this
regard of sensible ousia, but of physical ousia, for matter and form are not
sensible. Yet they are incorruptible. On this point, Porphyry might have thought
that there was a short treatise of physics in Book 12 itself (1069b–1070a), since
in that passage Aristotle showed how composite sensible ousiai presuppose
matter and form and that the matter and form are ungenerated (1069b35).
Matter and form, since they are ungenerated and therefore incorruptible, could
thus correspond to incorruptible ousia, which is the subject-matter of the highest
part of physics. With regard to the unmoved ousia of which Aristotle speaks
(1069a34 and 1071b4), Porphyry gives it several names: ‘intelligible’, ‘intellective’,
‘incorporeal’, ‘unmoved’ and ‘motive’. This group of names can already be found
in the description which Alexander of Aphrodisias gives of Aristotle’s first cause
(Quaest. 3,20–4,25). He therefore conforms to the Aristotelian tradition, and one
actually re-encounters the description of Aristotle’s Prime Mover in Book 12.
None of these names poses difficulties for a Platonist except ‘mover’, which could
apply to the soul, the principle of movement for the body, but applies only with
difficulty to an intelligible ousia that is pure thought. Porphyry therefore tries to
explain this phrase in a manner satisfactory to a Platonist by saying that
intelligible ousia is motive to the degree that it is the cause of life, which is
movement: ‘insofar as it is the cause of movement specified as living’. This
interpretation agrees with the definition of the Prime Mover as a perfect, eternal
living being, whose life consists in its own acts of intellection (1072b26).
The last lines of the text of Dexippus we are commenting on reveal in an
extremely interesting manner how Porphyry interpreted the whole of Book 12
from a Plotinian perspective. For a pupil of Plotinus, the intelligible, unmoved,
motive ousia of Aristotle corresponds to the second hypostasis, that is, to Intellect.
Beyond this first ousia, there was the One, the absolutely simple principle of
ousia. In Porphyry’s eyes, the Aristotelian doctrine of ousia presupposes and
ultimately entails the Plotinian One. In fact, Book 12 of Aristotle’s Metaphysics
proposes a system of ousiai, a system whose unity and coherence are assured by
a single and first principle, intelligible ousia. Now, according to a fundamental
axiom of Neoplatonism, which can be found already in Plotinus (5.6.3.1–23;
6.6.13.17ff ), every ordered multiplicity, every ‘system’, presupposes the
transcendent One whose unity grounds the possibility of that order. Hence the
system of ousia implies the transcendence of the One. There must be a system of
ousiai, that is, different ousiai must form an ordered and unified multiplicity,
144 Aristotle Transformed

since, as Porphyry tells us, ousia has its being in the One. The formula probably
alludes to two Plotinian concepts. First of all, it recalls the Plotinian principle: ‘All
beings have their being from the One’ (6.9.1.1; 5.1.7.15; 5.5.6.14ff ). Secondly, it
might be an allusion to the exegesis which Plotinus gives of Timaeus 37D6, when
he says that eternity (here assimilated to intelligible ousia) remained ‘within the
One’ (en heni). Porphyry (Sent. 44.21 and 45.2–3) will later use the expression
again: in accordance with the One, within the One (kath’ hen, en heni). Therefore,
if the ousia which is so close to the One, since it has its being in the One, were not
unified, were not coherent and did not form a system, what then would participate
in the One? The whole would be completely scattered and rent. It may legitimately
be supposed that Porphyry had attempted to rediscover this Plotinian doctrine
in Book 12 of the Metaphysics. The last chapter of Aristotle’s book, in fact, permits
the notion that in the nature of the whole the Good was both something separate,
existing by itself and in itself, and the order of the whole (1074a11). A Plotinian
might have been able to recognise in this Good, existing by itself and in itself, the
One which is the principle of universal order and especially of the system of
ousia. The last chapter of this book insisted precisely on the internal cohesiveness
of ousia: ‘Those who want mathematical number to be the first reality – and thus
always accept one ousia after another and always different principles for each of
these ousiai – render the ousia of the All incoherent epeisodiôdê); for one
substance in no way affects another by its existence or non-existence, and they
accept a multiplicity of principles. Yet the beings do not wish to be badly
governed: “Government by many is not good; One should be sovereign” ’
(1075b38). We can discern this Aristotelian sentence behind Porphyry’s
discussion: ‘He has ordered the ousiai into a single system and has traced them
back to a single principle. For what then will participate in the One, if ousia itself
– that which has its being in the One – is deprived of the internal cohesiveness
which is attributed to the One?’ Thus it can be shown by what reasoning it was
possible for Porphyry to give a Plotinian interpretation of Book 12.
The last part of Dexippus’ text (41,18–42,3) is devoted to the relation –
whether homonymy, synonymy, or analogy – which unites the different ousiai
and most especially to the manner in which we know intelligible ousia; this was
the third point we distinguished above (p. 132):

Most certainly intelligibles are ineffable: he therefore uses the term ousia in
accordance with metaphor and analogy with things knowable through sensation.
For all things which share the same name do so in three ways: homonymously,
synonymously, and in accordance with the relation of the proper sense to the
metaphorical sense–homonymously, as the foot of a tool stands in relation to the
 The harmony of Plotinus and Aristotle according to Porphyry 145

feet of other instruments; synonymously, as when one says that a sword is

sharper than another sword and a voice is sharper than another voice (but not
that a sword is sharper than a voice, since the quality ‘sharp’ is different in one
and in the other); and by the relation ‘proper-metaphorical’, as for example with
our foot and the foot of mountains.
Intelligibles, thus, are ineffable: it is therefore metaphorically that Aristotle
uses the word ousia, bringing us to know imperceptible things with the help of
things that are sensible and perceived by us.
In fact, just as ousia will be homonymous in relation to intelligible ousia
(because it only means intelligible ousia by pure analogy), so too ousia will be
synonymous in relation to physical ousia, because it means physical ousia in
virtue of the fact that it is composite.
For, in the same way that Aristotle in the treatises of the Physics (or:
Metaphysics?) calls a certain incorporeal ousia ‘body’ – calling it incorporeal
because it proceeds from intellective ousia, but asserting further that it possesses
the being of body since it is already in movement towards sensible ousia – so too,
by saying that this ousia is synonymous in relation to physical ousia, though
homonymous in relation to intelligible ousia, he makes the other ousiai known
through sensible ousia and above all makes it known that the description which
relates to sensible ousia – ‘neither said of a subject nor in a subject’ – can also
apply to the other ousiai. (Dexippus, in Cat. 41,18–42,3)

This text develops, within the framework of the general interpretation

previously given of Book 12, the doctrine of the analogy of the concept of ousia,
which, we have seen, concluded the exposition of the Plotinian doctrine of ousia.
Here and there, one re-encounters the claim: it is possible to extend the
Aristotelian description of ousia (‘not being in a subject, not being said of a
subject’) to all ousiai, even to intelligible ousia. This possibility is grounded
precisely in the fact that the different ousiai are unified in intelligible ousia,
which is their principle.
There are, therefore, the following relations between the different ousiai.
Sensible or composite ousia is synonymous in relation to physical ousia, that is,
ousia considered as matter or as form. Sensible or composite ousia is homonymous
in relation to intelligible ousia, this homonymy not being complete, however,
since it is compatible with an analogical relation. This doctrine is implicitly
presented as Aristotelian. It can in fact be justified by the Aristotelian division of
the sciences. The Physics has for its subject form, matter, and the composite of the
two – thus, on the one hand, physical ousia, and on the other hand, composite or
sensible ousiai (Phys. 194a12). First Philosophy, by contrast, has as its subject the
‘separated’ (khôriston), that is, intelligible ousia (Phys. 194b14). The synonymy of
146 Aristotle Transformed

the ousiai that are the subject of the Physics is grounded in the concrete unity of
phusis, the principle of movement and formal principle of natural ousiai (Metaph.
5.4, 1014b35; 12.3, 1070a5ff; Cael. 1.8, 276b1–6). On the contrary, the homonymy
between sensible ousia and intelligible ousia, that is, between the subject of the
Physics and the subject of First Philosophy, corresponds to the total difference of
principles that exists between sensible ousia and intelligible ousia (Metaph. 12.1,
1069b1; 4, 1070b15–20). That this homonymy is not absolute, but permits an
analogical relation, results from the discussions of analogy contained in Book 12
(1070b30–1071b3). What grounds the analogy is the very fact that all the ousiai
refer ultimately to intelligible ousia, from which they originate. The analogical
relation thus tends to be confused with the relation ‘from one and to one’ (aph’
henos kai pros hen), although this relation is not made explicit either in Aristotle
or in Porphyry. In any case, in Porphyry’s eyes, by affirming that there is an
analogical relation between sensible ousia and intelligible ousia, Aristotle is in
profound agreement with Plotinus, who writes in his treatise ‘On the genera of
being’ (6.3.5.3): ‘This is what must be said about ousia here below; does all this
suit ousia in the higher world? Perhaps, but it is by analogy and homonymously
(kat’ analogian kai homônumôs)’.
We have just mentioned analogical relation. Dexippus’ text seems to confuse
this with metaphor, since, before telling us that composite ousia stands in
analogical relation to intelligible ousia, he declares that Aristotle uses the term
ousia ‘in accordance with metaphor and analogy’ in order to speak of intelligible
substance. In fact, in his commentary in question and answer form on the
Categories (66,34), Porphyry points out that Atticus classes analogy and metaphor
together. He does not himself distinguish between the two either, but in the
same commentary (67,4) he takes care to distinguish between metaphor and
homonymy: homonymy occurs when a single term designates different things,
without there being a distinct word to designate these things: for example, with
the ‘foot’ of a bed and the ‘foot’ of a table; metaphor occurs when the same term
designates different things, even though there is a distinct word to name them, as,
for example, with the ‘foot’ of a mountain (since the word hupôreia designates
this) and the ‘foot’ of a ship (since the term pêdalion designates this). As Simplicius
remarks (in Cat. 33,11), Porphyry appears to have given up this kind of distinction
between homonymy and metaphor in his greater commentary addressed to
Gedalius, since in that work he gave the foot of the ship as an example of
homonymy and the foot of the mountain as an example of metaphor. One can
perhaps recognise the traces of the commentary to Gedalius in Dexippus’ text,
where the foot of the mountain is given as an example of metaphor and the feet
 The harmony of Plotinus and Aristotle according to Porphyry 147

of different machines or tools (probably with the foot of the ship contained in
this category) as an example of homonymy. One might even think that Porphyry
had been led to modify his theory of metaphor and homonymy because of the
doctrine of ousia. If sensible ousia and intelligible ousia are homonyms, they are
still not only that, for there also exists a relation of analogy and metaphor between
them. In that case, if the definition of metaphor proposed in the question and
answer commentary were applied, one could not say that there was a metaphorical
relation between the two ousiai, since it would be necessary to assume that there
was a distinct term for designating intelligible ousia. Now, the doctrine expounded
in our text consists solely in the assertion that there is no name properly belonging
to intelligibles: they are ineffable. It is necessary, therefore, to accept that a
metaphor can apply to something which has no distinct name.
Intelligibles are ineffable precisely because they are imperceptible. All of our
language originates from sensation (Porphyry in Cat. 91,8 and 20). The word
‘ousia’, therefore, can only properly designate the sensible individual, and
therefore composite ousia. In applying the term ‘ousia’ to intelligible ousia, one is
only using it as a metaphor.
This doctrine is never stated explicitly by Aristotle: the Greek terms
corresponding to ‘ineffable’ and ‘imperceptible’ do not even belong to his
vocabulary. At the very most, one can note that Aristotle time and again (An.
Post. 1.2, 71b33; Phys. 1.1, 184a16; Metaph. 2.1, 993b9; 7.3, 1029b3–12) asserts
that what is knowable in itself is, in relation to us, the most difficult to know, and
that we must seek to attain it by beginning with that which is more easily
knowable for us, that is, with that which can be grasped by sensation. The
doctrine can nevertheless be developed from Aristotelian principles.
Theophrastus comes very close to it in his Metaphysics (9a18): ‘If there exist
certain things that are only knowable in that they are unknowable, as some
maintain, the manner of inquiry into them would be one peculiar to them, but
needs some care to distinguish it from others; though perhaps, in cases where it
is possible, it is more appropriate to describe them by analogy than by the very
fact that they are unknown – as if one were to describe the invisible by the mere
fact that it is invisible.’ Traces of this doctrine are re-encountered in Albinus
(Didask. 165,17 Hermann). Yet curiously it is in Porphyry himself that we again
find the general system in which this doctrine is integrated most clearly
expressed. For Porphyry, the incorporeals are defined only by a negative method
(Sent. 19, 27, 35 and 38): their name is fixed by saying what they are not, without
our being able to define what they are. One can equally define them by the
method of attributing antithetical predicates (‘everywhere – nowhere’, for
148 Aristotle Transformed

example). There is a complete reversal of perspective once we pass from the

sensible to the intelligible or vice versa: the intelligible is nothing in relation to
the sensible, the sensible is nothing in relation to the intelligible. It is therefore a
genuine negative theology that is applied to the intelligible.2 This is completely
different from Plotinus, who restricts the methods of negative theology to
knowledge of the One and never stretches the opposition between the intelligible
and the sensible to the point of making each the negation of the other.
It remains then to explain Dexippus’ curious allusion to a text of Aristotle’s in
which the latter might have called a certain incorporeal ousia a ‘body’. Even the
sense of the allusion is double. To begin with, in the immediate context, it
illustrates the double relation of sensible ousia to physical ousia and to intelligible
ousia. Sensible ousia is synonymous with physical ousia, but is homonymous to
intelligible ousia, that is, the term ‘ousia’ is transferred from sensible ousia (where
it is employed in a proper manner) to intelligible ousia, because the latter is the
principle of sensible ousia. In the same way, the incorporeal ousia of which
Aristotle is said to speak could be called both ‘body’ and ‘incorporeal’:‘incorporeal’
synonymously with intelligible ousia, from which it originated, but ‘body’
homonymously, since body and sensible ousia originate from it. The term ‘body’,
when applied to this incorporeal ousia, would be metaphorical, just as ‘ousia’ is
when applied to intelligible reality. A second reason motivates this curious
citation of Aristotle. Plotinus had objected to the Aristotelian doctrine of ousia
that if sensible ousia and intelligible ousia had something in common, then the
incorporeal would be body and body incorporeal as a result (cf. p. 126 above).
Porphyry appears, therefore, to be responding here to this objection by saying: it
is possible to call the incorporeal ‘body’ – Aristotle himself has done it – but
using ‘body’ homonymously and metaphorically; it is in precisely the same way
that the word ‘ousia’ is applied to the intelligible.
To which text of Aristotle’s is Porphyry alluding? To judge by the explanation
which Porphyry offers of the text, Aristotle spoke of a reality which was placed
between intelligible, intellective ousia and sensible ousia. In Aristotle’s system,
this substance is very probably that of the Heavens, whose description in the de
Caelo (1.9, 279a19ff ) might cause a Neoplatonist to think that this was an
incorporeal ousia.3 Now this reality was quite clearly presented by Aristotle as a

2
cf. P. Hadot, Porphyre et Victorinus, Paris 1968, 110 and 420.
3
Notably 279a19ff: ‘Hence the upper things are by nature neither in a place nor caused by time to
grow old, nor is there any change in anything beyond the outermost motion, but they spend their
entire duration unchanged and unaffected, having the best and most self-sufficient life. . . . On them
depend the rest – some more clearly, others dimly – for life and existence’.
 The harmony of Plotinus and Aristotle according to Porphyry 149

body (Cael. 1.9, 278b10ff ). Yet according to Dexippus’ text, Porphyry apparently
alluded to a passage in the Metaphysics. In Book 12, there is a single passage
(1074a30) where there is a question of ‘divine bodies’ carried in the heavens. It is
not very likely that it was this passage to which Porphyry alluded: the relation of
these bodies to an incorporeal ousia is hardly sugested by the context. One may
ask, therefore, whether it might not be necessary to read ‘in the Physics’ in place
of ‘in the Metaphysics’. Note moreover that Dexippus always introduces a citation
of the Metaphysics with ‘en têi’ and a citation of the Physics with ‘en tois’. The
expression ‘en tois Phusikois’, in Aristotelian usage could designate the de Caelo
(Metaph. 1.8, 989a24), among other works. Porphyry might therefore be making
an allusion to this treatise in which the first body presents the properties of a
divine being, on which the being and life of each thing depends, and which
moves itself endlessly in a circular motion (279a11–b3). One might even wonder
whether Porphyry had a precise text of the de Caelo in mind, in which a textual
error justified his interpretation. Note that in a discussion devoted to the first
body, Aristotle uses the following formula: ‘that there is naturally some corporeal
ousia which is different from all the other structures here, and which is more
divine and prior to all of these (hoti pephuke tis ousia sômatos allê para tas
entautha sustaseis, theiotera kai protera toutôn hapantôn)’ (269a30). Porphyry
might have read: ‘some incorporeal ousia (tis ousia asômatos)’. Whatever it was,
here there is again a glimpse of a fascinating Neoplatonic exegesis of Aristotle’s
system. The first mobile (Cael. 288b1) proceeds from the incorporeal ousia and,
by its movement, initiates the procession towards sensible ousia. It is in exactly
the same situation as the soul in the Neoplatonic system.
Dexippus appears, therefore, to have preserved in his commentary on the
Categories a precious testimony of the manner in which Porphyry was able to
comment on Book 12 of the Metaphysics, while seeking to interpret Aristotle’s
philosophy in order to set it in harmony with Plotinus’ philosophy. Certain
indications permit a glimpse even of Porphyry’s rediscovery of the doctrine of
the transcendent One in Book 12 of the Metaphysics. If that is true, there is
something of a paradox about a pupil of Plotinus rediscovering in Aristotle the
very doctrine which Plotinus expressly reproaches Aristotle for having ignored.
150
 7

Porphyry’s legacy to logic: a reconstruction

Sten Ebbesen*

1. Introduction

Porphyry’s importance as an Aristotelian commentator has been a commonplace

for a long time. He is generally recognised as the great sifter of the tradition,
so that the history of the commentaries may be represented schematically as
follows

This means that whatever ante-Porphyrian materials are found in later

commentaries came to them via Porphyry. Of course, this cannot be true without
modifications, but it is obvious from a reading of the preserved commentaries
on the Organon that the stemma works in a great many cases.

* In the footnotes, these special abbreviations are used:

CLCAG = Sten Ebbesen, Commentators and Commentaries on Aristotle’s Sophistici Elenchi, 3 vols,
Corpus Latinum Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Comm. 2 = a Byzantine commentary on Aristotle’s Sophistici Elenchi, excerpts from which are
edited in CLCAG 7, vol. 2.
ps.-Alex.-1 in SE = The final version of Michael of Ephesus’ commentary on Arist. SE, published as
Alexandri quod fertur in Aristotelis Sophisticos Elenchos commentarium, ed. Maximilianus
Wallies, Commentaria in Aristotelem Graeca 2.3, Berlin 1898.

151
152 Aristotle Transformed

This much is fairly uncontroversial. The evaluation of Porphyry’s activity

as a scholastic logician is less so. One question which I find interesting is
whether the works quarried by posterity were like much of their offspring
in being jumbles of scholia containing tantalisingly disconnected pieces of
logico-semantical theory. The question has received little attention.1 In my
opinion, his major commentaries, like the great one on the Categories and the
one on de Interpretatione, are likely to have differed from their offspring in this
respect.
Now, suppose (1) that some of the most important and tantalising bits of
doctrine found in later authors should be traced back to Porphyry’s works.
Suppose (2) that he had a coherent – if not entirely consistent – set of views on
questions of logic. Suppose further (3), that his voluminous production contained
not only more details than the less extensive works of later scholars, but also
some better indications of the mutual dependence of the sundry pieces of logical
theory. If all three suppositions are granted, it follows that such later scholars
as had access to Porphyry’s own writings would have known and presumably
found acceptable the systematic context of the fragments of doctrine which
they themselves chose to transmit to posterity. If only the two first suppositions
are granted, it still follows that a systematic understanding of the fragments
presupposes an understanding of Porphyry’s theory – unless, indeed, the later
authors managed to have an essentially different theory while still having a place
for items that fitted into Porphyry’s as well.
The nature of our sources makes it very difficult to produce convincing
support of supposition (3). The greatest obstacle is the loss of all Porphyry’s
commentaries except for the concise Introduction to the Categories (the ‘Isagoge’)
and the equally concise minor commentary on the same work. But at least the
preserved writings and the reports of the lost ones point to a high degree of
consistency. This raises hope that we may successfully reconstruct some main
outlines of his logical theory in accordance with supposition (2). I shall try to do
so and to show that supposition (1) receives support from the fact that several
important items of later commentaries do actually become intelligible when
placed in a Porphyrian framework.

1
The best attempt to reconstruct a coherent Porphyry is by A.C. Lloyd, ‘Neoplatonic and Aristotelian
logic’, Phronesis 1, 1956, 58–72 & 146–60. Porphyry’s logic is also treated in Bent Dalsgaard-Larsen,
Jamblique de Chalcis: exégète et philosophe, Diss., Århus 1972, ch. 5, but his approach is so
fundamentally different from mine that it would make no sense to polemicise against details of his
interpretation. I must leave it to the reader to judge whose reconstruction of Porphyry is the most
convincing.
 Porphyry’s legacy to logic: a reconstruction 153

Since the purpose of this chapter is primarily to elucidate pieces of doctrine

that have to do with fallacies, it is, of course, important to know whether Porphyry
ever wrote a commentary on the Elenchi. There are signs that he did so;2 but the
problem is not quite as acute as it might seem, for the affiliation of the Elenchi
scholia that I am going to discuss is generally such that if they do not stem from
Porphyry’s notes on that work, they certainly count scholia of his on other books
of the Organon among their ancestors. One thing is known with certainty:
Porphyry on several occasions stressed the role of concepts in the semantic
process. He considered vocal discourse as a sign of mental discourse and the
principal terms (words) of a vocal proposition (sentence) as signs of terms
(concepts) of the mental proposition. Words signify things only via the concepts.
It is also clear that ‘concepts’ are somebody’s concepts, not self-subsistent entities
independent of their human possessors. I shall come back to the evidence for
this later, but first some general considerations.
By stressing the role of concepts in the semantic process Porphyry shows that
he had successfully resisted the temptation felt by many logicians to eliminate
the human users of language. Can we use our knowledge of this fact as a clue to
his logic? I think we can.
One natural consequence of saying that signification depends on concepts
and that concepts are somebody’s concepts is that an analysis of language is
something distinct from an analysis of the nature of things. Studying the way
language is used we study the ways in which concepts are expressed but not the
structure of reality. We may reach an insight into the manner in which reality is
conceptualised, and that will be reflected in language, but logic cannot get us
nearer to reality than this. That is, logic does not comprise ontology.
Another consequence is this: vocal language must be public and it must
be possible to identify concepts across the boundaries of individual minds.
Otherwise communication would break down.
Further, and for the same reason, people must be assumed to have a tendency
to link concepts in the same way, so that it is possible to cross-identify their
conceptualised reality. But it must also be possible for some individual to link
concepts in a way that does not correspond to reality as commonly conceptualised.
If this were not the case, concepts would be an otiose element in the semantic
model. But this being so, it is natural to hold that truth and falsity are primarily
properties of mental propositions, wrongly linked concepts constituting false
propositions, whereas vocal sentences are not strictly speaking true or false.

2
See CLCAG 7, ch. 5.4.
154 Aristotle Transformed

They may be called so only in the sense that they are correct expressions,
according to common usage, of true or false mental propositions.
On the other hand, we may expect a Porphyrian logician to hold that
ambiguity is a property of words and sentences, not of mental language. Holding
that a concept could be ambiguously related to things would imply a claim that
we can know more about the things than our concepts give us. Within the
framework of this kind of logic, the claim is hardly acceptable, whereas it is
perfectly possible that one element of vocal language can express several
elements of mental language.
Ambiguity, then, will be a property of expressions and dependent on a
conventional use of one such expression to signify several concepts or
combinations of concepts. – A conventional use, I say, because with signification
depending on the way people understand expression, i.e. the way they relate
them to their concepts, it is reasonable, though not strictly necessary, to consider
vocal language as man-made and usage as a matter of convention. – But it is also
reasonable to assume that many features of this conventional language can be
explained genetically as resulting from the way in which the concepts it is
designed to express are formed.
In the main, I think, the texts bear out the above expectations concerning
Porphyry’s views. I shall try to show that he considered Aristotelian logic as an
instrument designed to deal with the language whose primary function is to
describe the commonly conceptualised sensible world, and that he thought the
instrument could be used for such purposes without clashing with Platonic
ontology. I think his position can be compared to the way many modern
people view Einsteinian and Newtonian physics. They accept the former as the
comprehensive theory and the latter as a useful theory for limited purposes,
which, however, happens to be of particular importance for normal life. And as
long as the less comprehensive theory is applied within the proper confines, there
is nothing wrong about accepting it. In the same way, I suggest, Porphyry assigned
the little corner of the universe that interests most men as a sphere of operation
for Aristotelian logic on the condition that it must not extend its field of operation
to ontology proper which can be investigated only in Platonic terms.
Besides this, I shall try to show that Porphyry had consistent or at least
interconnected theories to explain phenomena falling within the domain
encompassed in his view by Aristotelian logic, and, in particular, that he had a
theory concerning the rise of concepts which he connected with a genetic
account of the rise of vocal language; that he considered ambiguity a property of
expressions and truth as a property of mental propositions, primarily at least.
 Porphyry’s legacy to logic: a reconstruction 155

Finally, I shall try to show how Porphyrian theory and the problems that
interested him lived on – though more or less maimed – constituting an
important factor in the shaping of the Greek scholastic tradition in many fields,
and with important consequences for the treatment of fallacies as well.
If the picture of Porphyry that arises from the reconstruction should remind
anybody of William of Ockham or John Buridan, that will hardly be fortuitous. I
find that on many basic points the three philosophers agree so closely that in
matters where our knowledge of the medieval thinkers surpasses what we know
of their ancient predecessor we may use them as guides to the reconstruction of
his thought. This is not to say that I have deliberately tried to model Porphyry in
the likeness of Buridan or Ockham, but they have no doubt had an influence on
my conception of what logic with such basic tenets must be like. If anybody
should think it incredible that the so-called ‘nominalism’ was to any important
degree a revival of Porphyrian thought, the reply is that this is not so very
incredible: fourteenth-century philosophers were busy readers of such late
pupils of Porphyry’s as Ammonius, Simplicius and Boethius. Porphyrian logic
was as accessible to late medieval men as it is to us, though they knew less about
the provenance of what they read. There is nothing incredible in the idea that
men with the flair for consistency that Buridan possessed should have picked up
from the ancient sources exactly such materials as had belonged in the context
of one coherent logic, namely Porphyry’s.

2. The harmony of Plato and Aristotle

Porphyry’s basically friendly attitude towards Aristotelian logic is beyond doubt.

He wrote commentaries and companions to several parts of the Organon,3 and
the preserved part of his work shows that he was not polemical. That he thought
Platonism and Aristotelism to be compatible, at least to a high degree, is
demonstrated by the fact that he wrote a work in seven books to show that the
two philosophical systems are really one and the same (Peri tou mian einai tên
Platônos kai Aristotelous hairesin).4 To be sure, he also wrote a treatise On What
Separates Plato from Aristotle (Peri diastaseôs Platônos kai Aristotelous),5 but we

3
cf. Ebbesen, ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and Boethius. Ancient
scholasticism and twelfth-century Western Europe’, Cahiers de l’Institut du moyen âge grec et latin
(CIMAGL), University of Copenhagen, 16, 1976, 11.
4
See J. Bidez, Vie de Porphyre le philosophe néo-platonicien, Gand-Leipzig 1913, 68*.
5
ibid.
156 Aristotle Transformed

have no reason to believe it was anti-Aristotelian. It was written for the benefit of
the same Chrysaorius for whom he composed the Introduction to the Categories,
so in all likelihood it was a short treatise designed to show that the two men’s
different approaches to philosophy did not amount to a radical disagreement.
Of course, Porphyry was not blind to the differences between Plato and
Aristotle; he just did not think that the peculiarities of the latter’s philosophy
were such as to make his words worthless to the Platonist. A few passages from
the Isagoge and the minor commentary on the Categories should suffice to
delineate his attitude:
A Platonist need not disagree with Aristotle’s claim that ‘being’ (to on) is
equivocal. Plotinus did not, Pletho – much later – did.6 But surely one does not
expect a Platonist to mention the Aristotelian doctrine without any kind of
comment. Yet, Porphyry quietly remarks:7 ‘As being is equivocal, in Aristotle’s
opinion, [. . .] therefore he talks first of equivocals’ (i.e. therefore the Categories
treat of equivocals before univocals).
Most Platonists have found it shocking that Aristotle ranks abstracts as
secondary substances and sensible entities as primary substances.8 Porphyry’s
remarks on the question reveal much, I think, about the way he viewed
Aristotelian logic.
Of course, he says,9 intelligible entities like the intelligible god and the intellect
and the ideas – if in fact ideas exist (eiper eisin ideai) – are primary. But that is no
real argument against Aristotle’s terminology, which is quite justified in its context.
The theme of the Categories is significative expressions and the classification of
realities they indicate; but men started by giving names to the things they saw and
felt; linguistic occupations with the entities that are naturally prior, but posterior
with respect to the senses was a secondary event. So, in relation to the significative
expressions, individuals and sensible substances are the primary ones, even
though the intelligible ones are so from the perspective of nature.
The main point in this passage is clear: the doctrine of the Categories does not
have universal application, but within its own confines it is valid.
A little bit of extrapolation leads to the conclusion that Porphyry thought of
the whole of Aristotelian logic in that way. There is evidence10 that he (like many
later commentators) considered the Organon as a systematic whole moving

6
Plotinus 6.1.1; Pletho de Diff. ch. 3 PG 160:893–6.
7
Porph. in Cat. 61,10–12.
8
Thus Plotinus 6.3.9; Pletho de Diff. ch. 4 PG 160:896–7.
9
Porph. in Cat. 91,14ff.
10
ibid. 56,23–p.57.
 Porphyry’s legacy to logic: a reconstruction 157

from simple expressions in the Categories to compound expressions (sentences/

propositions/premisses) in the de Interpretatione and to compounds of sentences
(syllogisms) in the Prior Analytics, and thence to the several kinds of syllogisms
dealt with in the Posterior Analytics, the Topics and the Sophistici Elenchi.
So, the whole of logic is a study of the properties of extramental language
considered as an instrument basically used to comment upon the world of
experience.11 This is why there cannot be a clash between Aristotelian logic and
Platonic metaphysics as long as the logician works within the confines of his art.
The same attitude is manifest in the proem of the Isagoge where Porphyry
says that although he is going to write about genus and species, he will not treat
the difficult question of their mode of existence. He just wants to indicate what
older philosophers, and in particular the Peripatetics, have said about them from
a logical point of view.12 The implication must be that as long as you do not
confuse it with metaphysics, Aristotelian logic is both a useful discipline and one
that does not clash with Platonism.
Again, talking about differentiae, Porphyry explains why a genus cannot have
the differentiae of its subordinate species actually, but only potentially. He does
not even mention that Platonists have another opinion. Only the presence of an
unobtrusive ‘as they [i.e. the Aristotelians] claim’ hints at the possibility that their
claim may not have universal validity.13

3. The rise of language and the formation of concepts

Concept-formation and the naming of concepts were closely related processes in

Porphyry’s view. We have rather good information on the question, but I must
begin by mentioning one difficulty. It is this:
Whereas later reports and echoes of his logical writing state in an unequivocal
way that he paid great attention to the role of concepts in the semantic process,
the most important of his extant works for the present purpose does not even
contain the word noêma (‘concept’). Why is it that the minor commentary on the
Categories generally treats of signification as a direct relation of expressions to
things? Does it belong to some phase of his philosophical development other

11
cf. Simpl. in Cat. 9–13.
12
Porph. Isag. 1,15 logikôteron [. . .] dielabon; I do not think Boethius’ translation ‘probabiliter [. . .]
tractaverunt’ hits the mark, though he argues well for it in his in Isag.2 I, ch. 12 p. 168 (Brandt).
Ammonius in Isag. 44–5 takes my view.
13
Porph. Isag. 10,22–11,6. Cf. the judicious remarks in Ammonius in Isag. 101–5.
158 Aristotle Transformed

than the great commentaries on the Categories and de Interpretatione? Hardly so.
A much simpler solution imposes itself. The preserved commentary on the
Categories is of a decidedly compendious character. The author felt it was
important that people – and in particular the individual for whose benefit the
work was written – should not think that the Categories is a treatise ‘On What
There Is’, because interpreted as metaphysics its doctrine cannot be reconciled
with Platonism. But to drive home this point he had no need to expatiate on
concepts as the immediate significates of expressions. To produce a compendium
of a reasonable size he had to sacrifice something, and he chose to sacrifice the
concepts. They do not occur under their own name in the book, though there are
some passages in which they are hinted at. In exactly the same manner he does
not introduce the technical term akatataktos (unranked) which he is known to
have used in the lost major commentary. Furthermore, as there is hardly a place
in which his argument would be spoilt if it were complicated by the insertion of
concepts as a link between expressions and things, I feel convinced that we are
entitled to take what he says in the concept-less commentary as part of the same
theory that he defended in his other writings. The following reconstruction of
Porphyrian semantics presupposes that all statements about words and things
can, and should, be regarded as abbreviated statements about words, concepts
and things.
I shall use the following conventions:

<dog> = a thing which is a dog

/dog/ = a concept of a dog
[dog] = the word ‘dog’

Some central pieces of Porphyrian semantics are describable by means of a

genetic model of the rise of concepts and language. I shall call the theory behind
that model ‘the theory of imposition’. Porphyry gives a short version of it in the
minor commentary on the Categories.14 It can be read as an historical account
implying that at some definite time men introduced all the words that refer to
sensible reality, and at some later time they resolved to introduce the vocabulary
used to speak of the primary language. Some later commentators took this to be
Porphyry’s intention,15 but they were surely wrong. What he meant was rather
that there are certain natural steps in the development of our stock of concepts
and words. Objects must be sensed and conceptualised before anybody can

14
Porph. in Cat. 57–8.
15
See CLCAG 7, ch. 4.4.8.
 Porphyry’s legacy to logic: a reconstruction 159

speak about them. The use of class-names presupposes a process of abstraction.

A meta-language has no sense unless there is an object-language. Whether or
not the historical development of the Greek language has any similarity to the
process described in the model is really irrelevant.
On the following pages I shall reconstruct a fuller version of the theory of
imposition than any that occurs in the sources. It will be a version that Porphyry
perhaps never wrote, but not, I hope, one he would have rejected.16
The theory takes for granted that men have sense-perception of perceptible
objects. If some more comprehensive metaphysical theory were to claim that the
sensible world is not really real, it does not matter in this connection. For the
present purpose the sensible world is supposed to be real and prior to our
perception of it and to our conceptualising of the percepts. If the metaphysical
theory were to claim that the sensible world is an illusion, the logician is entitled
to say, ‘All right, but for practical purposes we need a theory which treats the
illusion as if it were a reality; and nothing, of course, prevents us from adding a
qualification like “as it seems” whenever we say “real” ’.
Suppose, then, that we have got a social group of some people. They come
across an object. They perceive it and they straightaway get a concept of that
individual object. They somehow establish the convention that when one wants
to externalise his own concept and evoke the corresponding one in another
member of the group, he just has to say [dog].
For the theory to work it must be supposed that it is possible to identify one
man’s concepts with another’s. Porphyry actually made that assumption, as we
shall see later.
It is now time to introduce some essentials of the terminology used by
Porphyry and ‘Porphyrian’ philosophers in the context of the theory.17

16
Porph. Isag. (in particular pp. 6–7); Porph. in Cat. (in particular pp. 57–8, 75, 81, 90–1); Porph. in
Cat. ad Gedalium ap. Simplicium in Cat. 53 & 79 are the main sources for the following. Many details
have been filled in with the help of Dexippus in Cat. (in particular pp. 7–16) and Simplicius in Cat.
(in particular 10,20–13,18; I hold that Simplicius’ interpretation of Porphyry is correct). By using
Dexippus and Simplicius I run the risk of attributing to Porphyry views that were Iamblichus’ and
not his; but, on the whole, Iamblichus will not have deviated from Porphyry (cf. Simpl. in Cat. 2,9ff ).
On imposition, see Porph. in Cat. 57–8; Dexippus in Cat. 11–16; Ammonius in Cat. 11; Simplicius
in Cat. 15; Boethius in Cat. PL 64:159. Sources for details will be given in subsequent notes.
17
The following is a selection of passages containing semantic terms. More may be found via the
indexes of CAG. As for terms for mental and oral discourse, see CLCAG 7, ch. 4.3.6. dêloun Porph.
in Cat. 57,7–8, 20–2; 58,12–14; 60,16–19 [≈ Simpl. in Cat. 22,15ff ]; Dexippus in Cat. 11,14–15;
11,28; 19,20–2. ekpherô Porph. in Cat. 107,35; 112,17–20; 115,31; Simplicius in Cat. 155,6; 160,1;
ps.-Alex.-1 in SE 23,13. On the Stoic use of the word: Ammon. in An. Pr. 68,4–6 ≈ Philoponus in
An. Pr. 243,1–4; cf. Dionysius Halicarnass., de Compositione Verborum ch. 8, secs 45–6, p. 32
(Usener-Radermacher). ekphônein Porph. in Cat. 115,25; Simplicius in Cat. 64,26; 132,9–12;
ps.-Philoponus in An. Post. 421,3–4; ps.-Alex.-1 in SE 23,8; 28,27–8; 36,15. (ex)angellein Porph. in
160 Aristotle Transformed

That of which a concept (noêma) is a concept is a thing (pragma); as long as

the conceptual and linguistic apparatus of men is not much more sophisticated
than in the society at the stage we just considered, the only relevant ‘things’ are a
primitive kind of sensible units which I shall call objects (roughly corresponding
to ‘spatio-temporal particulars’), though the Greeks just say ‘things’. Concepts are
elements of the mental (en dianoiâi) or interior or dispositional (endiathetos, en
diathesei) language or discourse (logos). [dog] is an element of the vocal (en
phônêi) or ‘forthbringing’ (pronouncing) (prophorikos, en prophorâi) language or
discourse (logos). Though this is not yet known to our primitive men, such
elements of the vocal language may be considered from various points of view.
In so far as they are sounds ‘brought forth’ (pronounced) (propheromena), ‘brought
out’ (ekpheromena) or ‘sounded out’ (ekphônoumena), they are rightly called
words (phônai, also translatable as ‘(articulate) sounds’). Their role is to
externalise elements of the interior language, and in that capacity they are
expressions (lexeis, literally ‘sayings’ or ‘callings’), whereas the concepts are
expressed (legomena). As the theory under construction will only incidentally
deal with [dog] and the like as mere words, we may perhaps forgive the old
philosophers for not always observing the distinction between ‘word’ and
‘expression’. Expressions may be compounds of several simple elements (words);
a compound expression is often called a compound word (sound) (sunthetos
phônê). The term ‘expression’ has to do with the externalising of concepts seen
from the perspective of the speaker. Seen from the perspective of the person
who hears the word [dog], it may be said to be a sign (sêmeion) of the presence
of a certain concept in the speaker, and so it is said to signify or signal (sêmainein)
the concept, which is the signified or significatum (sêmainomenon). To indicate
more clearly the communicative function of words, Porphyry and others liken

Cat. 58,23; Nemesius de Natura Hominis ch. 1, p. 43 (Matthaei); Ammonius in Int. 5,15; 18,31;
Olympiodorus in Cat. 19,20–4 (in what purports to be a quotation of Alexander): Olympiodorus in
Phaedonem 4, sec. 10, p. 85 (Westerink) = p. 27 (Norwin); Elias in Cat. 130,11–13; Stephanus in Int.
5,20–36; Psellus (?) Schol. in Int. ch. 1 MS Vat. Barb. gr. 164:17rM; ps.-Alex.-1 in SE 11,27; Michael
Ephesius & Anon. Coislinianus in Int. (CAG 4.5, xlvi); Leo Magentinus in Cat., prooemium, MS
Vat. gr. 244: 32v. hermêneuein Ammonius in Int. 5; Boethius in Int.2 pp. 10–11 (Meiser); Philoponus
in Cat. 9,29–31. legein Dexippus in Cat. 7–10; Comm. 2 in SE 166a1–6; ps.-Alex.-1 in SE 11,26.
mênuein Porph. in Cat. 56,6; 57,27; 63,16; 125,18. (kat)onomazein Porph. in Cat. 57,21, 28; 60,15ff;
91,6, 22; Dexipp. in Cat. 6,14ff; 9,16–20. paristanein Porph. in Cat. 55,8–12 [≈ Dexipp. in Cat. 6,18–
20]; 57,23; 58,3. propherein Porph. in Cat. 59,12; 71,5,29–30; 104,27; 112,18; Dexippus in Cat. 8,28–
9,1; 9,14; Philoponus in An. Pr. 273,4; Comm. 2 in SE 166a1–6. sêmainein Porph. de Abstinentia 3,3;
Porph. in Cat. 57,28; Dexipp. in Cat. 7–10; Ammonius in An. Pr. 1,8–9; Philoponus in Cat. 8–9;
Simplicius in Cat. 9–13; Boethius in Int.2 pp. 26–36 (Meiser); Comm. 2 in SE 166a26–8. (kata)
tassein epi/kata Porph. in Cat. 65,33; Dexippus in Cat. 20,3–4; Simplicius in Cat. 27. tithenai Porph.
in Cat. 57,28–31.
 Porphyry’s legacy to logic: a reconstruction 161

them to messengers (angeloi, in Latin: nuntii) and say that they ‘bring forth
messages about’ or enunciate18 (exangellousi) concepts.
Thus far the terminology is reasonably clear, though the reader may expect
trouble to arise from the liberties taken in the use of ‘word’ and from the fact that
the terminology does not allow for a distinction between word form token, word
form type, and lexeme.19 But this is nothing compared to the trouble that is
created by an extended use of ‘signify’, ‘enunciate’ and ‘express’.
In our simple language, at least, in which we have to consider only the word
[dog], the concept of which it is a sign, and the object of which the concept is and
which is the cause of the concept, it seems reasonable to say that just as [dog] is
a sign of the presence of the concept in the speaker, so the concept is a sign of the
presence of the object in sensible reality. But then it would seem that signification
is a transitive relation. If one of our primitive men is a cave-dweller and says
[dog], and the word signals the presence in his mind of a concept which signals
the presence outside the cave of an object, is it not reasonable to say that [dog] is
a sign of the presence of the object, or for short, that it signifies the object?
Porphyry and his followers thought so and they would say ‘the word signifies
the thing’ without hesitation except when they feared lest people forget that the
concepts also play a role. Then they would say ‘the word signifies primarily the
concept and secondarily the thing’ or ‘the word signifies the thing via the concept’.
They would also say ‘the speaker signifies (by means of a word) . . .’ without
sharply distinguishing this usage from the one in which the subject-term is ‘the
word’. With ‘express’ they seem in practice to have respected the stipulation that
the subject must be a person, the proper objects being concepts, though
secondary usage allowed for things or even words as the objects. ‘Enunciate’ was
treated exactly like ‘signify’, as, probably, was ‘inform of ’ (mênuein) which seems
to mean exactly the same as ‘enunciate’. Further, they would say that the speaker
or words interpret (hermêneuousi) certain kinds of interior discourse (categorical
propositions), make present (paristanousi) the things (to the collocutor), and that
men name ((kat)onomazousi) things, by calling (kalountes) them something.
When people first call a thing by a name, they lay down (tithentai) a name for it
or put a name on it (epitithenai: in the following I shall translate this by impone)
or assign a name to it (katatattein epi). ‘Names’ occur as elements both of the

18
The Latin enuntiare seems to copy exangellein, but in extant literature it appears to represent
apophainesthai; cf. Gabriel Nuchelmans, Theories of the Proposition: ancient and medieval
conceptions of the bearers of truth and falsity, Amsterdam and London 1973,107.
19
On the distinction, cf. John Lyons, Semantics, vols 1–2, Cambridge 1977, 1.13ff.
162 Aristotle Transformed

vocal and of the mental discourse,20 but usually a ‘name’ is understood to be a

word and to be a ‘name of ’ (onoma + a noun in the genitive) a thing. Finally, ‘to
indicate’ (dêloun) is used to cover almost any semantic relation, much like
‘signify’, though most often we find speakers or words ‘indicating’ things.
The invention of the art of writing further complicates the picture. Men write
(graphousi) names or expressions (onomata, lexeis) which inform of, enunciate
or signify the corresponding elements of vocal language, and thus ultimately of
the things.
Hardly any ancient or medieval philosopher was quite consistent in his use of
‘semantic’ terms, and the locution ‘concepts signify/enunciate things’ was not
commonly used because ‘man and his words signify concepts which signify
things’ was normally represented by the telescoped locution ‘man and his words
signify things via the concepts’.21
With this terminology in mind we may return to our primitive society in
which the word [dog] is used as uniquely signifying (the concept of) some
particular object. Their concept of that object has an important peculiarity: it is
structured. They may not know so and they have no words to describe the fact,
but nevertheless it is a fact which is going to influence their behaviour.
Men of a more sophisticated age may say that something in the structure of
the objects or in our perception and conceptualisation of them makes us have
structured concepts of them. Each concept may be described as an ordered set of
components, each component being a value of some feature for which we
instinctively investigate the object. The concept of the first object called dog is
likely to be analysable in something like the following way: +corporeal, +animate,
+sentient, –rational, +quadrupedal, +able to bark, +floppy-eared.22
Armed with such a concept our primitive men come across more objects and
form concepts of them. It so happens that each of the new concepts has exactly
the same components as the original one with only one exception: some or all of
the new concepts lack the component +floppy-eared. Without reflecting on it the
primitive men decide in each case that the proper word to use as a sign of the
concept is [dog].

20
cf. Porph. in Int. ap. Boethium in Int.2 I, ch. 1, p. 30 (Meiser).
21
Variants of the formula hai phônai sêmainousi ta pragmata dia mesôn noêmatôn are ubiquitous. See,
e.g., Ammonius in An. Pr. 1,8–9; Philoponus in Cat. 9,14–15; Simplicius in Cat. 42,5–6; Comm. 2 in
SE 166a1–6; ps.-Alex.-1 11,25–7.
22
My ‘components’ are idiotêtes in Greek; see Porph. Isag. 7 (e.g.). Concepts of individuals are rarely
mentioned by the commentators (Aristotle’s text did not invite them to do so); but see Dexippus in
Cat. 10,20; Philoponus in Cat. 12,4–9; Boethius in Int.2 28,18ff. (Meiser); Simplicius in Cat. 69,15–16.
It is possible that Porphyry would not use the word noêma but ennoia when speaking of concepts
of individuals.
 Porphyry’s legacy to logic: a reconstruction 163

To some extent, this is impractical, because it makes it difficult to refer

unambiguously to one object. At some point in their mental development people
will solve the problem by introducing real proper names like [Snuffy], [Rover]
and so on.23 Porphyry does not comment on this, but he does mention, though
rather briefly, the important discovery which is a presupposition for the
introduction of a separate class of proper names.24
The interesting discovery is this: people realise that they use the word [dog]
to signal all such concepts as have the components + corporeal, +animate,
+sentient, –rational, +quadrupedal, +able to bark, or, to put it differently, they
realise that they use [dog] as a name of all such objects as are corporeal, sentient,
irrational, quadrupedal and able to bark. (Observe once more that it does not
matter whether the objects seen from the perspective of some metaphysical
theory really exist and really are corporeal, etc. What matters is that they are
sensible objects, i.e. the kind of objects whose existence must be taken for granted
in the world of experience, and that they are conceptualised as bodies that are
sentient, irrational etc.)
The new discovery results in the acquisition of a concept of such an object.
And the word people will use when they want to signal a mental discourse
involving the new concept will be [dog].
This means that we are left with two uses of the word [dog]. One in which it
is employed as a sign of a sensible and ostensively identifiable object, a ‘this
something’ (tode ti) whose existence prior to perception and conceptualising
must be assumed; another in which it is employed to signify a non-sensible and
not ostensively identifiable ‘such a thing’ (toionde ti) whose ‘existence’, for all we
know, is entirely a product of our concept of it. If it so happens that there really
exists something matching our concept of ‘such a thing’, this is a contingent and
irrelevant fact seen from the perspective of the present theory (logic), unless,
indeed, it is argued that we could never recognise similarities25 between sensible
objects if there were no pre-existing intelligible objects possessing, in some
particularly genuine way, all the features whose values we were able to identify
across the boundaries of individual objects, thus being able to see the objects as
similar and to name them identically. Porphyry certainly did not consider the
question unimportant;26 but as it was an intricate and much disputed one, he

23
‘Socrates’ is a word of the first imposition according to Dexippus in Cat. 15,27.
24
Porph. in Cat. 90–1.
25
On similarity as an important factor in abstraction: Boethius in Isag.2 I, ch. 11, p. 166 (Brandt) = PL
64:85B (Boethius may depend on Alexander of Aphrodisias – see p. 164 (Brandt) = 84B (Migne) –
but Porphyry would hardly disagree); Simpl. in Cat. 103,7.
26
See CLCAG 7, ch. 4.4.2.
164 Aristotle Transformed

thought it need not bother the mind of the student of logic, at any rate not the
incipient student, for whichever the solution to the problem, logic would not be
invalidated: if the recognition of similarities turns out to be explainable in terms
of the independent existence of intelligible objects, so much the better; if not,
this will leave logic unimpaired – the only thing the logician needs to do is to
postulate the ability to discover similarities.
But let us have another look at the two uses of [dog]. We have seen that only in
the case of the very first bearer of the name – and that dog may have died and
become irrelevant a long time ago – was the choice of the name [dog] really
arbitrary (for all we know, though, considerations of no concern for logic might
point to a not completely arbitrary choice).27 All subsequent naming was systematic
and the people of our primitive society have even realised this. It was based on an
analysis of the objects in which features like ‘floppy-eared’ were consistently left
out of consideration, though people certainly had concepts including all the
components of a dog-concept plus a component like +floppy-eared. At this stage
of development, at least, it is no longer possible to say that men signify the object
a more sophisticated age would call ‘Snuffy’ via a solely individual-identifying
concept when they speak of it by means of the word [dog]. The concept involved
somehow does and does not include the individualising component.
If we say it does not, then it has exactly the same components as the concept /
common dog/, and the difference between /this particular dog/ and /common
dog/ is rather syntactical than one of inner structure: /this particular dog/ is
necessarily associated with a concept like /Snuffy/ – which is otherwise identical
with it but includes an individualising component; /common dog/ is definitely
not associated with any components later in the ordered series than +able to bark.
If you cannot quite make up your mind on the question, and I think Porphyry
could not, /this particular dog/ has a strange intermediary status between /
Snuffy/ and /common dog/. In a sense it is identical with the former, in a sense
with the latter and in a sense, maybe, it is a compound of the two.
The terminology used by Porphyry and his followers is indicative of their
hesitation. It reflects both a will to identification and a will to separation of
concepts on different levels.
Let us repeat: Not all significata of concepts are accessible to direct inspection;
only individual sensible objects, such as <Snuffy> are so, <common dog> is not.
The components of /common dog/ are the initial segment of the series of
components constituting /Snuffy/, but they are considered qua not having had a

27
See ibid., ch. 4.4.8.
 Porphyry’s legacy to logic: a reconstruction 165

position assigned to them in any such series, and so /common dog/ may be called
the concept of the ‘un-ranked’ (akatataktos). If we consider the same components
qua having had a position assigned to them in a series that comprises ‘lower’
components too, we have a concept of a ‘ranked’ dog (katatetagmenos) which is
in a sense identical with the unranked dog and in a sense identical with Snuffy.28
Unranked universals are created by lifting components out of their context;
therefore they may be called aposulêmata dianoias,29 i.e. the booty that the mind
has carried off (from the sensible things); they are secondary to and come after
the particulars (they are epi tois pollois, husterogenê, post rem)30 and subsist only
by virtue of a secondary conceptualisation.31 Ranked universals have a kind of
intermediary status between the truly universal and the particular; they are the
universal/common ‘in’ or ‘as considered in’ the particulars ((theôroumenon) en
tois pollois, in re), and the dog in Snuffy may be described as the dog that is
‘ranked with’ or ‘ranked in’ (sunkatatetagmenos, enkatatetagmenos) Snuffy.
Unfortunately, the ‘ranked/unranked’ terminology was also used in another
sense. In Neoplatonist terminology ranking was the deployment of an idea (or
any supra-sensory entity) on even lower levels; each time we pass to a new level,
some undifferentiated feature of the idea acquires a differentiated value to the
exclusion of others: thus, in a way, the ranked is poorer than the unranked which
contains the differentiae of all its subordinates in an undifferentiated way and is
prior to them (pro tôn pollôn, ante rem).
I do not think Porphyry used ‘unranked’ in the sense of ante rem in his
Aristotelian commentaries, but he did use the term, and to later commentators it
may not have been obvious he had meant post rem. However, they are generally
aware of the fact that Aristotelian universals are not ante rem; Ammonius and
later scholastics sometimes say Aristotelian universals are in re, but in practice

28
katatassein, suntassein and their derivatives are frequent in the commentaries, but usually in the
Neoplatonist sense. See, e.g., Dexippus in Cat. 26,9; 45,22; 51,13; Simpl. in Cat. (see the index of
CAG 8 s.v. katatattein); Philop. in An. Post. 133,24; Asclepius in Metaph. 189,27; 447,25; Syrianus in
Metaph. 7,12; 20,9; 28,14; 30,28; 32,9–10; 36,28; 95,11; 98,34; 155,29; Elias in Cat. 154,17–19; ps.-
Alex. in Metaph. 524,9–12; Damascius in Phaedonem I.502 & II .51 (Westerink); Damascius in
Philebum sec. 80, p. 39 (Westerink) (further references in Westerink’s index, p. 137); Iamblichus in
Timaeum fr. 60 (Dillon); Proclus in Parmenidem col. 939 (Cousin) (and in many other places).
29
This use of aposulan, aposulêma is apparently not attested before Ammonius. Cf. my ‘Hoc aliquid
– quale quid and the Signification of Appellatives’, Philosophia 5–6, 1975–6, 382; add to the
references: Leo Magentinus in Cat., prooem., MS Vat. gr. 244: 32v. [I have now discovered a similar
use of ‘aposul-’ in Iamblichus’ commentary on Nicomachus’ Arithmetic, so Ammonius is not the
earliest source.]
30
This terminology is used by Ammonius in Isag. 41–2 and by several later authors.
31
This formulation was used by Porphyry. See the index of CAG 4.1 s.vv. epinoein, epinoia. In
Simplicius in Cat. 104,21–2 katatakhthenta eidê are distinguished from ta kath’ hauta epinooumena;
the passage is probably of Porphyrian origin.
166 Aristotle Transformed

they often neglect the distinction between in re and post rem universals. Porphyry
may also have neglected it sometimes, but at any rate both he and others were
willing to distinguish between concepts on different levels and with different
syntactical possibilities, so we may assume that they would admit that any
particular man has a limited number of dog-concepts: one of the post rem
dog and a finite number of concepts of the common dog in the particulars.
But it is easily seen that in virtue of his having the concept of ‘such a thing
as a dog’, any man is able, in principle, to enlarge indefinitely his set of concepts
of particular dogs; only his death or shortage of dogs could stop the process.
And what the individual man cannot do because he is mortal may be possible
for his race. Suppose the sensible world is eternal and neither men nor dogs
become extinct. Then the set of objects of which people will have particular
dog-concepts is infinite. And suppose the abstraction of the concept of the
absolutely common dog is not only conceived of as a unique historical event,
but also as occuring in each individual man. It will then be true that seen from
the perspective of history the basis of abstraction for common concepts are
infinite sets.
No such reasoning is found in the writings of Porphyry, but a couple of
statements of his seem to require it; for in the Isagoge32 he says that the number
of individuals are of the infinite order, and in the minor commentary on the
Categories33 he says that common concepts are derived from all the particular
things that bear the same name as the universal.
A related difficulty is this: Porphyry sometimes, in Aristotelian contexts,
speaks of the common comprising (periekhon) all the particulars falling under it,
and of the particulars as sharing or participating (metekhonta) in it.34 If my
account of Porphyry’s way of thinking has been correct to any reasonable degree,
he cannot have meant by ‘participation’ anything that presupposes ante rem
universals, and he cannot have meant by ‘comprising’ that the significatum of a
concept of something common in abstracto – and hence of the word signifying
the concept – can be extensionally delimited. What he must have meant is this:
as long, at any rate, as our language and our concepts are used for their primary
purpose, both interior and exterior discourse ultimately referring to the sensible

32
Porphyry Isag. 6; cf. Porph. in Cat. 58,7–8, where it is said that both things and words are virtually
(skhedon) infinitely many. It is uncertain if the restriction skhedon bears on both things and words
or only on the latter. If Porphyry interpreted Arist. SE ch. 1 165a10–12 literally, he must have
believed that in the Aristotelian system things are assumed to be infinitely many, but not words.
33
Porph. in Cat. 90,32–4.
34
cf. the index of CAG 4.1 s.vv. metekhein, periekhein.
 Porphyry’s legacy to logic: a reconstruction 167

world, universal concepts have no function at all except by virtue of their having
each an associated set of concepts of particulars. For an unranked concept to
comprise ranked concepts and for the unranked universal to comprise the
particulars is simply to have such an associated set.
This interpretation is supported by some carefully phrased passages of the
Isagoge and the minor commentary on the Categories. In one place he explains
that ‘man’ ‘makes present’ (paristêsi) two things at least: an essential quality and a
community between the subordinate substances.35
Later commentators embroider the Porphyrian explanation and make generic
and specific names ‘signify a set and a quality’ and ‘indicate a gathering and
collection of the particulars’ (so Ammonius36), or ‘indicate a set and community
comprising several members, and a quality’ (Philoponus37), or ‘indicate a
qualified set’ (Commentator III 38). I do not think Porphyry would have approved
of those changes of his wording.
He does use such words as ‘set’ and ‘collection’ himself. But notice how careful
he is. At the very beginning of the Isagoge he examines the meaning of genos.39
The first use of the word, he says, was to designate the origin of each man’s birth,
and next it came to designate the set (plêthos) or collection (athroisma) of people
sharing a common origin of birth. It is not in any of these senses, he continues,
that we use the word when we speak of the genos to which the species are
subordinate, but probably the logical use was established in imitation (kath’
homoiotêta) of prior usage: for in a sense the genus of logic is the origin of its
subordinate species, and it comprises, as it were, the whole subordinate set (kai
gar arkhê tis esti to toiouto genos tôn huph’ heauto kai dokei kai to plêthos
periekhein pan to huph’ heauto). As I understand the text, the purpose of writing
arkhê tis and dokei periekhein instead of the plain arkhê and periekhein is to
point out that the genera of logic are neither ante rem nor simply identical with
a collection of extensional objects they ‘comprise’. Similarly, for the particular
objects to participate in the common thing and for the concepts of them to
participate in the common concept, means for each of them to be a member of a
set associated with the common thing or concept whose ranked form they
display. They participate or ‘share’ the unranked by each of them displaying a
ranked ‘manifestation’ associated with it.

35
Porph. in Cat. 96,19–20.
36
Ammon. in Cat. 49,6–11.
37
Philoponus in Cat. 72.
38
Comm. 3 in SE 169a33 (sch. 12).
39
Porph. Isag. 2.7–12; cf. Porph. in Cat. 77,27–30; 90,4–9.
168 Aristotle Transformed

It is now time to return to the concept- and language-forming men of the

primitive model-society. We left them with concepts and names of nothing but
dogs, but, of course, the same procedure that enabled them to think and speak of
dogs will enable them to acquire other concepts on the same levels, like that
of Socrates (components: +corporeal, +animate, +sentient, +mortal, + x (x is
individualising)) and of man (same components less x) and to impone the
relevant names. Using once more the same procedure, they acquire concepts of
a higher level, such as that of an animate sentient body, and associate names with
them, such as ‘animal’.
The examples used so far might suggest that the conceptual field of the men
in Porphyry’s model is limited to substances. But this is not so. Porphyry is
explicit on the point that they also acquire names (and, of course, concepts) of
numbers, magnitudes, colours, and so on, in the same way.40 In short, they create
all the kinds of words necessary in order to be able to signify the things of the
sensible world. These are the ten kinds of words dealt with by Aristotle in the
Categories, words of the first need (tês prôtes khreias) and of the first imposition
(prôtês theseôs).41
But men have more needs than the first. They do not only wish to signal
single concepts of things. It is all right to be able to point out Snuffy and say
[Snuffy] or [dog] or [animal] or [barking]. But mental discourse is more
complicated than that. We do not just have concepts presenting themselves
to consciousness in disconnected succession, each one alone. We have
compounds (suntheseis, sumplokai) of concepts, such as /Snuffy + deep + hole +
barking/. Such a compound is a mental proposition – atomic or molecular –
and to be able to signal it in an intelligible way people must invent a new
kind of words that will not each signal a concept but the structure of such
compounds.
The new kind of words are connectives and include what the grammarians
call conjunctions, articles, prepositions, plus certain adverbs and, probably, the
copula. They are not in themselves significative – no particular concept being
associable with each one of them – but taken together with significative words
they help express complex concepts. Hence they are said to be consignificative
(sussêmantikai). Porphyry in his lost commentaries on the Categories and
de Interpretatione no doubt used a time-honoured simile, comparing such

40
Porph. in Cat. 57.
41
prôtê thesis Simplicius in Cat. 15,7–8, e.g.; prôtê/deutera khreia occurs several times in Dexippus in
Cat. (see the index of CAG 4.2), but not in other authors, as far as I know.
 Porphyry’s legacy to logic: a reconstruction 169

connectives to glue and to the bolts and tar that hold the planks of a ship
together.42
With connectives at their disposal men can frame such sentences as [Snuffy is
barking in a deep hole], [Snuffy is barking near a deep hole] and [If Snuffy is
barking, the hole is deep].
The notion of compounds of concepts has many obscure sides to it. I do not
know whether Porphyry could give a satisfactory account of the general relations
of such a compound to its constituents, to its and their significata, and to
expressions. Only as regards essential predication do I think we can catch a
glimpse of how his machinery was supposed to work.
Essential predication43 or predication hôs kath’ hupokeimenou ‘as of a subject’,
occurs when the [P] of [S is P], without being either the name of an individual
<S> or the expression of an individualising component, is the expression of one
of the components constituting the concept /S/ or the expression of a concept
constituted by a string of components which is identical with some initial part of
the string of components constituting /S/. Examples are [Socrates is rational],
[man is animate] and [Socrates is a man], [man is an animal].
To predicate, Porphyry says, is (1) to ‘mention the things in respect of some
significatum’, or more generally (2) ‘to say a significative expression of things’.44
If we say [Socrates is an animal], we mention <Socrates> but we invite people
to use their concept /Socrates/ only to single out the thing that we talk about and
then to consider him via the concept /animal/, i.e. to pay attention to those
components of /Socrates/ that are also components of /animal/. [Socrates is an
animal] evokes two concepts, that of <Socrates> and that of <animal>, and
indicates a composite concept which is that of the particular ranked animal of
which we also have the concept/Socrates/.
Description (2) simply says that we use the word [animal] and that we claim
that it is one correct way of signifying the subject – which, in fact, it is, because
one of the immediate significata of [animal] is the concept of the ranked animal
which is Socrates.

42
On connectives: Dexippus in Cat. 32–3; ps.-Augustinus Categoriae Decem 2; Ammonius in Int.
12–13; Boethius in Int.2 I, ch. 1, p. 6 (Meiser); Boethius Syll. Cat. I, PL 64: 796C–D; Simplicius in Cat.
64–5; they all certainly reflect Porphyry – who in Cat. 57,32 implies that articles have been
introduced before the second imposition. The glue-and-bolt simile was known as Peripatetic to
Apollonius Dyscolus (Fragmenta = GG 2.3, p. 31; cf. Priscianus Inst. 11.2.6) and in all probability to
his source, the first-century BC grammarian Trypho as well. On ‘is’ as a connective, see Porphyry in
Int. ap. Boethium in Int.2 I, ch. 3, p. 77 (Meiser) (cf. Ammonius in Int. 55–7 and also Simplicius in
Cat. 42,23–4).
43
The term ‘essential predication’ may not be Porphyrian, but that is of no consequence here.
44
Porph. in Cat. 58,16–18, cf. 56,8–13 and the quotation in Simpl. in Cat. 17,3–7.
170 Aristotle Transformed

Someone might object that on this account an essential predication is a mere

tautology like ‘Socrates is Socrates’ or ‘an animal is an animal’. Simplicius45
informs us that some people actually did so and objected to Aristotle’s use of
heteron . . . heteron ‘one (thing) . . . another’ in Cat. 3 1b10–15: ‘When one [thing]
(heteron) is predicated of another (kath’ heterou) as of a subject, all [things] that
can be said of the predicate can also be said of the subject; for instance, man is
predicated of the particular man and animal of man; consequently animal may
also be predicated of the particular man.’ Aristotle ought not, the objection ran,
to have said ‘one of another’, for animal is predicated of man qua animal and
colour of white qua colour. Porphyry’s rejoinder, Simplicius says, consisted in
pointing out that the notion of animal is twofold (dittê hê epinoia tou zôiou), viz.
of the ranked and of the unranked. When animal is predicated of man it is the
unranked animal that is predicated of the ranked, and thus it is correct to say
that it is one (thing) that is predicated of another.
The passage is important for an understanding of Porphyry’s thought. He
cannot possibly have meant that the expression [unranked animal] is predicated
of the expression [ranked animal]. For one thing, this would conflict with his
explanation of what ‘to predicate’ means; and for another, the actual words of the
sentence are not [unranked animal is predicated of ranked animal]. Nor does it
make sense to impute to him the theory that the significata of [man] and [animal]
are [ranked animal] and [unranked animal] respectively.
He cannot, either, have meant that the ante rem animal is predicated of the in
re animal.46 If he had intended to say that, the word epinoia would have been
very infelicitous and he would have abandoned the position that the universals
we operate with in Aristotelian logic are not self-subsisting entities, but just
mental constructs, the products of epinoia.
I do not think it is possible to make sense of Porphyry’s statement unless we
assign some role to concepts, in spite of the fact that he does not mention them
explicitly.47 What he can have meant is this: If we put the pure concept of animal
on top, as it were, of the concept of man, the result will be a concept of man in
which the components it shares with that animal are emphasised. That is, the
attention is directed to the ranked animal in man. In a sense this is putting
/animal/ on top of /animal/, but only in a sense, for the predicate /animal/ is

45
Simpl. in Cat. 53,4ff. Cf. Plotinus 6.3.5.
46
As supposed by P. Hadot, ‘La métaphysique de Porphyre’, Fondation Hardt, Entretiens 12, Porphyre,
1966, 152.
47
Ammonius does. On two occasions (in Int. 21,21ff and 51,20–2) he says that to be a subject or
predicate is not a property of things, but primarily of concepts and secondarily of words.
 Porphyry’s legacy to logic: a reconstruction 171

unranked, while the /animal/ which the predicate draws attention to in the
subject is ranked.
The mental proposition signified by [man is an animal] is thus describable as
a concept of man with emphasis on its animal components, as a concept of that
particular ranked animal which is man, or as a compound of the concepts of
man and animal. To make the compound we must have the concept of the
unranked animal, but by using it in connection with /man/ we rank it.
It might be asked whether the concept /man/ is really /this ranked man/ or
/this ranked man, that ranked man . . ./ or /man unranked/, but the answer is of
no consequence, as /unranked animal/ can help us see /ranked animal/ in /man/
whether the significatum of /man/ is the mental construct <common man> or
any or all of the individual men. The properties of /man/ which interest us in the
present connection are not such as belong exclusively to /unranked man/. So in
the context, the /man/ may be treated as delimiting a ‘this something’ (tode ti)
that is co-extensional with <this ranked animal>.
Thus the significatum of the subject term is, in a sense, the same as the
significatum of the whole proposition. In this sense the subject thing is ranked,
and we draw attention to the fact by ‘mentioning it in respect of ’ the concept
/unranked animal/ – in short, we predicate animal of it.
It is unfortunate that we have so little left of Porphyry’s disquisitions into the
matter. But in spite of difficulties in interpreting the remains, it seems clear that
the main point he wanted to make was that although even a very elementary
form of discourse (mental and oral) cannot be explained except by assuming
that people have and employ concepts of universals, this does not imply that
their discourse is about universals as such.
We have no account of how Porphyry would verify or falsify ‘man is an animal’
or ‘Socrates is a man’. But scattered information about statements of his and later
echoes of his works give some foundation for a reconstruction.
It can be taken for granted that somehow truth and falsity were dependent on
an adaequatio rerum et intellectuum and resided in a composition of concepts,
i.e. in mental propositions. It is a priori likely that a sine qua non – and maybe
sufficient – cause of the falsity of a mental proposition was the failure of such a
thing as it signifies to exist.
But how can men have wrong compounds of concepts?
It might be assumed that a man can have incorrectly structured concepts and
so be able to form in his mind compounds which would not be possible if his
concepts were correctly structured. Probably Porphyry acknowledged such
pseudo-concepts. Yet he would not even call them concepts (noêmata) because
172 Aristotle Transformed

the man who holds them cannot properly be said to have conceived or understood
(nenoêkenai) the objects. This argument occurs in a passage in Boethius which
indubitably reproduces a Greek commentary; it can hardly be anyone’s but
Porphyry’s.48
As Porphyry certainly believed in wrong compositions of genuine concepts,
it must be assumed that he thought that we can have correctly structured
concepts without, at first, being able to analyse them correctly, and that it is
possible to form in the mind pseudo-compounds of concepts which are not
really combinable.
Verification or falsification of a mental proposition of the essential-predication
type then presumably takes place as follows: we perform a kind of checking of
the subject concept to see if it can really be analysed in such a way as to reveal
the predicate concept ranked in it. If doubt lingers because the structure of the
subject concept remains opaque, the decisive check will consist in a confrontation
with the sensible objects ultimately responsible for the formation of the concept
in question. If they do, in fact, give rise to a concept which on close scrutiny and
under elimination of various disturbing factors is seen to be of the desired
structure, the proposition was true, otherwise false.
In practice, of course, one will usually have to be content with an epagogical
(imperfectly inductive) procedure, no man being able to make a complete survey
of the infinitude of objects underlying a universal concept.
A special case occurs if the objects of interest are all non-contemporaneous
with the speaker and his collocutors.49 Then other procedures, such as
confrontation with the writings of earlier authors, must replace the direct
confrontation with the objects.
If, finally, the objects held responsible for the formation of the concept turn
out to be not there, as would be the case with centaurs or goat-stags, the concept
is that of a fictitious entity and the proposition involving it is either neither true
nor false, nothing enabling us to check how the concept is to be correctly
analysed,50 or the proposition is false if every positive statement about the non-
being is false.51
Many questions, particularly about sentences in which the predicate is the
name of an accident, remain to be solved. I shall not try to solve them. Only one
thing seems certain: the theory of such statements must, as far as possible, be

48
Boethius in Int.2 ch.1, p. 41 (Meiser).
49
Dexippus was aware of this problem (in Cat. 7–8,10).
50
This was probably Dexippus’ opinion. See his in Cat. 7–8.
51
So Ammonius in Int. 52.
 Porphyry’s legacy to logic: a reconstruction 173

reconstructed so as to employ the same machinery as the theory of essential

predication.
We have followed Porphyry’s model men to a point where they are able to
speak of all kinds of things that occur in the world of experience prior to their
own conceptualising of it. They are also, presumably, able to judge the truth of
propositions, though probably they cannot yet say [This is true] or [This is false].
Presumably they have to cast their judgment about ‘Socrates is a man’ in the form
[Socrates is a man].
The ‘invention of universals’ was a major event in man’s intellectual history.
But still the possession of concepts of universals and of the words that signify
them has only the function of letting us form handily short propositions and
sentences ultimately concerned with the objects that we found in the sensible
world when our own intellectual activity started.
The decisive turn which makes scientific man out of merely rational man
occurs when he begins to reflect on products of his own intellectual activity:
universals, concepts, names. He then becomes capable of forming concepts of
non-ranked universals qua non-ranked and making statements about them. He
can bar the gate to the descent to individuals and say [man is different from
Coriscus] meaning ‘the pure concept of man, i.e., /man unranked/, has less
components than the concept of Coriscus, and so the mental construct <man>
and the sensible object <Coriscus> are not identical’. He can also make statements
about names of universals or of individuals, saying, for instance, [man is
monosyllabic]; – or about the ‘syntactical’ properties of names and concepts,
saying, e.g., [man is a species], meaning ‘the word [man] is such that there is a
concept of unranked man with an associated set of several concepts of ranked
man whose significata are individuals; /man unranked/ fits into any concept of a
ranked man and any of the several possible sentences of the form [x is a man]
expresses a true mental proposition if x is the name of one of those individuals.’
In none of the three statements is it presumed that the predicate concept is one
that could be arrived at by abstraction of the upper string of components of the
concepts given us by inspection of each particular man. And to that extent such
cases resemble accidental predication, as when we say [Socrates is white], more
than they resemble the superficially similar proposition expressed by means of
the words [man is an animal].52
The examples [man is monosyllabic] and [man is a species] are remarkable
for the fact that they include words that can only be used when we talk about the

52
See Porph. in Cat. 81.
174 Aristotle Transformed

products of our mental activity: they can never refer to the primary objects of
sensation. Porphyry called the institution of such expressions the second
imposition of expressions/names (hê deutera thesis tôn lexeôn/onomatôn).53
With the invention of metalanguage man has reached a sort of linguistic and
intellectual maturity. He is now able to speak about the sensible world and he is
even capable of speaking of his cognition of it. Of course, new concepts and new
words may arise, but they will all be describable within the framework we have
delineated.
The logician’s task, then, consists in pointing out the conditions for vocal
sentences expressing true reasoning, and for that purpose he will have a
permanent need to refer to the genesis of expressions.
It might be objected that on this account we should not be able to speak about
supra-sensory realities; and we should not be able to judge statements about
them if such statements were nevertheless expressible. I think we can imagine
what Porphyry would answer.54 He would probably say that supra-sensory
realities are neither conceptualisable in the same way as the objects of the
sensible world, nor expressible (they are arrêta). Some sort of acquaintance with
them is possible and the best we can do is to talk as if they were cognisable in the
same way as objects of the sensible world, using our regular object- and meta-
language, yet knowing that all such talk has the character of a simile. If such talk
is to be at all intelligible it must obey some of the basic rules of the language in
which it is couched, but the rules of logic are not infallible guides in such
questions. Suprasensory reality is outside the scope of Aristotelian logic.
But much is within its scope, and we shall now return to the consideration of
fallacies and see how some of them are explainable in the Porphyrian framework.
We shall also see how that very rough outline of the genesis of concepts and
language can be refined.

4. Ambiguity and its bearers

In Porphyrian theory ambiguity must be a property of sentences, not of mental

propositions; i.e., if a sentence is ambiguous, it must be capable of signifying
at least two non-ambiguous mental propositions. A simple example of an

53
In Porph. in Cat. onoma and rhêma are the only examples of words of the second imposition
(57,32–3); Dexippus in Cat. 13–16 and 26 adds many more; I think Dexippus had understood
Porph. correctly.
54
cf. Dexippus in Cat. 41,18ff; Simpl. in Cat. 73–4.
 Porphyry’s legacy to logic: a reconstruction 175

ambiguous sentence would be one containing an equivocal expression – and in

Porphyrian theory that must mean an expression that signifies two or more
concepts.
The point is nicely made by Dexippus, a staunch supporter of what I style
‘Porphyrian semantics’. He is very insistent on the importance of recognising that
the primary significata of expressions are concepts, things only being signified in
a secondary manner (kata sumbebêkos).55 In support of the thesis he claims, inter
alia, that sentences derive their truth or falsity from mental propositions and that
the explanation of amphibology, equivocation, and all other fallacies dependent
on expression is that one expression corresponds to several concepts so that the
mind is directed now to one, now to another of them.56
This is in perfect harmony with a scholium of Porphyry’s on de Interpretatione
1 related by Boethius, who informs us57 that Herminus had raised a difficulty
concerning the ‘passions of the soul’ or ‘concepts’58 which according to Aristotle
are the significata of expressions. For how could they be the same to all when in
the case of equivocation one and the same utterance has several significata (unus
idemque vocis modus plura significat)? Porphyry solved it by saying that: (a) The
speaker has in his mental discourse a certain concept that signifies one thing
only. (b) When he talks he intends that one thing only. (c) The recipient of the
message expects a message about one thing and interprets the message he gets in
one sense; but it may be another than the one the speaker intended, i.e. the two
relate the utterance to different concepts corresponding to different things (the
ultimate significates). (d) When the difference in interpretation is discovered,
the speaker must formulate his message in a clearer way; and then (e) the
speaker and the recipient of the message will join in having the same concept.
Consequently the fact that equivocation occurs is no objection to Aristotle’s
claim that people have identical concepts and of the same things.
The main conclusion to be derived from this passage is that the concepts had
by different individuals are cross-identifiable. Admittedly, Boethius does not say
that Porphyry said that he himself believed so. He says that Porphyry thought
Aristotle believed so and explained how such a belief could be defended. But
here, as elsewhere, it must be assumed that Porphyry did not want to disagree
with the Stagirite.

55
Dexippus in Cat. 7,1–2.
56
ibid. 10,3–10.
57
Boethius in Int.2 I, ch. 1, pp. 39–40 (Meiser).
58
Boethius says intellectus. There is no doubt possible that it renders noêmata: cf. Boethius in Int.2 I,
ch. 1, p. 28 (Meiser), e.g., and Ammonius in Int. 24,29–30.
176 Aristotle Transformed

The passage further shows that he thought it possible to signify the same
concept by means of an ambiguous and an unambiguous expression.
The man who, in good faith, has expressed his concept /x/ by means of
expression [z] which happens to signify also concept /y/, can express /x/ by
means of some expression [x] which does not signify any other concept. Boethius
does not tell us how Porphyry thought expression [x] would look, but we are
hardly off the mark if we suggest that he would consider the following an obvious
example: /x/ = /Ajax, son of Telamon/; /y/ = /Ajax, son of Oileus/; [z] = [Ajax];
[x] = [Ajax, son of Telamon]. A passage in his preserved commentary on the
Categories gives us the right to think so.59 He is likely to have held other methods
of disambiguation to be equally useful: some scholars have not hesitated to claim
a Porphyrian origin for the four types listed in Boethius’ de Divisione.60 The claim
has weak foundations, but even so it is extremely unlikely that Porphyry would
have disagreed with Boethius on the point. In fact, neither he nor any of his
successors among the Greek commentators seem to have thought there was any
problem at all: they considered it a matter of course that the ambiguity of a term
may be neutralised in various ways and that one way is to use the same term, but
provided with some restricting addition. If the old scholastics saw any problem
at all, it was rather the reverse: how is it possible for a term to signify ambiguously
if to signify is to signify some thing? If the only way of signifying one of the
supposed significata of [dog] is to say [dog that can bark] and the only way of
signifying another of them is to say [dog that is a star], these are two unambiguous
names and no role as a name is left for the supposedly ‘common name’ [dog].61
To the modern observer the possibility of disambiguation by means of
restricting additions may also seem so obvious that there is no need to make a
point of it. Yet, many western thirteenth-century philosophers saw a problem
here, and they did so exactly because they had accepted one important part of
Porphyry’s logic, namely the theory of imposition of names. If, it was argued, a
name has been imponed to express several significata, nothing can prevent it
from doing so; and consequently ‘Ajax, son of Telamon, duelled with Hector’ is
no less ambiguous than ‘Ajax duelled with Hector’. The solutions they advanced
are of no relevance here, but it is interesting to notice to whom the problem
seemed one and to whom it became a non-problem. The Parisian masters of the

59
Porph. in Cat. 64,9–17.
60
See CLCAG 7, ch. 2.3.4 note 19.
61
The problem was formulated by Nicostratus. We find discussions of it in Dexippus in Cat. 19 and
Simplicius in Cat. 26–8. Porphyry must have discussed it too, for it seems that it was he who
transmitted the Nicostratean problems to posterity (cf. Simpl. in Cat. 29–30).
 Porphyry’s legacy to logic: a reconstruction 177

late thirteenth-century were generally much more favourable to ‘hypostasising’

than the Porphyry I reconstruct, and much more interested in relations like
signification that could be assumed to have been fixed once and for all than in
such elusive relations as supposition (the context-bound relation of a term to
something else – concept or extramental thing). For these Parisian masters there
was a problem. For people whose basic approach to logic was more like Porphyry’s
and more like Buridan’s the problem evaporated.62
A third conclusion to be drawn from Boethius’ report of his great predecessor’s
scholium is that Porphyry thought that the terms of vocal language can have
several senses whereas the terms of mental discourse (the concepts) cannot.
It is only consistent that he should have held that sentences can be ambiguous
whereas mental propositions cannot. Once more Boethius provides the evidence.
Before Porphyry, he says,63 nobody had succeeded in producing a correct
explanation of de Interpretatione 5, 17a15–17, ‘for a declarative logos to be one is
either to indicate one [thing] or to be one by conjunction – and to be many is
either to indicate many [things] and not one or to be unconjoined’ (esti de heis
logos apophantikos ê ho hen dêlôn ê ho sundesmôi heis, polloi de hoi polla kai mê
hen ê hoi asundetoi).
A comparison of Boethius’ report of Porphyry’s exposition with Ammonius’
scholium on the passage64 shows that in the main the latter can be used as a
substitute for the Phoenician’s lost text, and so I shall borrow a few details from
that source; but they are not essential for the purpose. As far as the essentials go,
Boethius’ account is clear enough.
Porphyry, then, showed that the passage need not baffle commentators, for
every bit of it becomes intelligible when we remember that a logos has not
necessarily the same properties when we look at the expression and when we
look at the expressed; Aristotle speaks of the logos seen from both points of view,
and what he wants us to understand is that the expression is no infallible guide
when we ask about unity and plurality. If we say [a rational mortal animal is
walking], we get the impression of something very complex, but there is just one
logos, exactly as when we say [a man is walking]. But if we say [Ajax duelled with
Hector], there is more than one logos in spite of surface simplicity, for unity and
plurality is considered on the level of significata, not on the level of expression,
and [Ajax] has two significata. The following may serve as a rule of thumb: the

62
See my ‘The dead man is alive’, Synthese 40, 1979, 43–70.
63
Boethius in Int.2 II , ch. 5, pp. 106ff (Meiser).
64
Ammonius in Int. 72–5; cf. Dexippus in Cat. 9,14–15.
178 Aristotle Transformed

number of logoi equals the number of the significata of the subject expression(s)
multiplied by the number of the significata of the predicate expression(s).65
Porphyry added many niceties that need not detain us here. The conclusion
emerges clearly enough: ambiguity is a property of sentences and the number of
mental propositions signified by an ambiguous sentence is a function of the
number of the significata of the terms of the sentence. I do not think there was
anything new in this conception. As usual Porphyry’s greatness consisted more
in remembering in the right places what everybody was supposed to know than
in introducing novelties.
His influence lasted till the end of scholasticism. People like Leo Magentinus
did not have very clear conceptions of semantics, but still we find scholia such as
the following:66 ‘or we may understand [sc. 17a15] as follows: you ought not to
characterise a logos as one or many on the basis of the pronounced logos
(prophorikos logos) but of the signified, for “(a/the) dog is walking” looks in the
pronounced logos like one, but nevertheless it is several logoi because “dog”
signifies many [things]; in fact, as many [things] as “dog” signifies, so many
propositions have you said.’ (ê houtôs noêteon, hoti ouk opheileis kharaktêrizein
ton hena logon kai tous pollous apo tou prophorikou logou, all’ apo tou
sêmainomenou; to gar ‘kuôn peripatei’ei kai heis tôi prophorikôi logôi dokei, alla
polloi logoi eisi, dioti to ‘kuôn’ polla sêmainei; hosa gar sêmainei to ‘kuôn’, tosautas
kai protaseis eirêkas.)

5. Truth and its bearers

If ambiguity is a property of some expressions, truth can hardly be a property of

expressions, at least if it is held that (a) any categorical proposition must be either
true or false and cannot be both; (b) ‘p is true’ ↔ ‘∼p is false’ and ‘p is false’ ↔ ‘∼p
is true’. For supposing that the ‘propositions’ which are the bearers of truth and
falsity may be sentences (expressions), it is possible to construct examples that
force us to admit that (1) p is both true and false; (2) p is true and ∼p is also true;
(3) p is false and ∼p is also false.
Suppose the world once contained a pair of heroes, both bearing the name of
Ajax, one of whom had Telamon for his father, while the other was the son of
Oileus. Suppose that it is a fact that the son of Telamon once duelled with Hector,
whereas the son of Oileus never did so. Suppose further that we, the members of
65
cf. Ammonius in Int. 73,8–11; 126,11–14.
66
Leo Magentinus in in Int. 17a15, MS Vat. gr. 244: 104v, sch. lê.
 Porphyry’s legacy to logic: a reconstruction 179

a linguistic community, have just two concepts with which we associate the name
‘Ajax’, namely one concept of the son of Telamon and one of the son of Oileus. It
is irrelevant whether the feature by which we can distinguish the two concepts
is ‘being the son of X and not Y’: what matters is only that we have, in fact,
distinct concepts. Let it be the criterion of truth of a declarative sentence that the
concepts named by its subject and predicate terms are joined in the way it states
and that there is a corresponding composition in rebus. Then consider the
following pair of ‘contradictory’ sentences:

(1) Ajax duelled with Hector.

(2) Ajax did not duel with Hector.

We have concepts of two distinct persons called Ajax, and we know the
relevant facts concerning duelling with Hector about each of them. To determine
the truth-value of the sentences, we may try to replace the word ‘Ajax’ with
expressions that signify our distinct concepts in an unambiguous way. For
convenience I shall use ‘Ajax, son of Telamon’ as an expression that unambiguously
signifies one of the concepts, and ‘Ajax, son of Oileus’ as an expression that
unambiguously signifies the other of the concepts with which we associate the
word ‘Ajax’. The following rewritten pairs of sentences might be considered:

(I) (1) Ajax, son of Telamon, duelled with Hector.

(2) Ajax, son of Telamon, did not duel with Hector.
(II ) (1) Ajax, son of Oileus, duelled with Hector.
(2) Ajax, son of Oileus, did not duel with Hector.
(III ) (1) Ajax, son of Telamon, duelled with Hector.
(2) Ajax, son of Oileus, did not duel with Hector.
(IV ) (1) Ajax, son of Oileus, duelled with Hector.
(2) Ajax, son of Telamon, did not duel with Hector.
(V) (1) Ajax, son of Telamon, or Ajax, son of Oileus, duelled with Hector.
(2) Ajax, son of Telamon, or Ajax, son of Oileus, did not duel
with Hector.
(VI ) (1) Ajax, son of Telamon, and Ajax, son of Oileus, duelled with Hector.
(2) Ajax, son of Telamon, and Ajax, son of Oileus, did not
duel with Hector.
(VII ) (1) Ajax, not definitely the son of Telamon and not definitely the
son of Oileus, duelled with Hector.
(2) Ajax, not definitely the son of Telamon and not definitely
the son of Oileus, did not duel with Hector.
180 Aristotle Transformed

(VIII ) (1) Ajax, both definitely the son of Telamon and definitely the son
of Oileus, duelled with Hector.
(2) Ajax, both definitely the son of Telamon and definitely the
son of Oileus, did not duel with Hector.

(I–IV ) are certainly permissible rewritings of the two sentences on the

suppositions previously stated concerning our concepts. (V)–(VIII ) are more
dubious, but they might be tried to see if they could help to find a way of
interpreting (1) and (2) so that they can be ascribed definite truth-values without
elimination of the double signification of ‘Ajax’. Now, the truth-values we will
have to assign to the rewritten sentences are these:

(1) (2)
(I) T F
(II ) F T
(III ) T T
(IV ) F F
(V) T T
(VI ) F F
(VII ) ∼T&∼F ∼T&∼F
(VIII ) ∼T&∼F ∼T&∼F

If we consider (I)–(IV ) only, it appears that the non-permissible combination

TT occurs in (III ) and the similarly non-permissible combination FF in (IV ). In
Greek terms, (1) and (2) sunalêtheuousi and sumpseudontai (‘are true together,
are false together’). Further, (I)–(II ) show that (1) is both true and false – and
(2) likewise.
The trick of substituting a disjunctive or a conjunctive phrase for ‘Ajax’ leads
to the illicit combinations TT and FF in (V) and (VI ).
The final, desperate attempts of (VII )–(VIII ) probably lead to ∼T&∼F for
both sentences, for we have no concept of an Ajax who is not definitely one or
other of the two we know or definitely both, and so the sentences as phrased in
(VII )–(VIII ) are meaningless: it is not possible to find a pair of mental
propositions and compare them to the facts of the world to see whether there is
correspondence or not. We might expect a philosopher in the Porphyrian
tradition to say that they are neither true nor false.
So, what must our philosopher do? He wants to save the principles of
contradiction and excluded middle. He holds that sentences have mental
propositions for their primary significate and that verification and falsification
 Porphyry’s legacy to logic: a reconstruction 181

require comparison of the mental propositions with the extramental realities.

The simple – and the only – way out of the difficulties is to say that the real
bearers of truth and falsity are the mental propositions. If sentences can
nevertheless be called ‘true’ or ‘false’, their truth values must be a function (the
sum) of the truth values of the mental propositions they signify.
An inspection of the sources bears out our expectations. With a single
exception we have not got any statements about the question with direct
ascription to Porphyry, but the later commentators who are known to depend
on him agree so well that it is inconceivable that he should have thought
otherwise.
The one certainly Porphyrian statement unfortunately occurs in the concept-
less minor commentary on the Categories. Yet, it is indicative of what we might
expect from his less compendious works. In agreement with Aristotle he
mentions67 logos as an example of what falls under the category of quantity. For
precision he adds that the logos in question is not the one in the mind but the one
whose constituents are long and short syllables. A little later68 he tells his pupil
that a quantum has no opposite and shows this to be the case with lines, surfaces,
etc. – ‘Not even logos has an opposite, qua logos’. This elicits the question: ‘How
does it then come that one logos is called true and another false?’ To which
Porphyry answers: ‘But it was not in so far as it is significative that it was taken to
belong to the category of quantity: it was in so far as it is in pronunciation and is
measurable in terms of long and short syllables’. Strictly speaking this passage
does not prove that Porphyry thought truth and falsity do not at all reside in
vocal sentences. What it proves for certain is that at least he would not say that
sentences understood just as phonic stuff could be bearers of truth and falsity.
But if we remember that Porphyry in this concise commentary speaks as little as
possible about mental propositions, I think we may interpret the passage to the
effect that he means: ‘The true bearers of truth and falsity are the mental
propositions. If sentences can be said to be true or false this, at least, is only in
virtue of their signifying true or false mental propositions.’
That was how Dexippus saw the situation. He says69 ‘There could not possibly
be a true or false logos if the logos did not receive its truth or falsity from the
concepts, for the logos in the mind is prior, and it is probably after it that “saying”
(legein) and the vocal logos are named.’

67
Porph. in Cat. 101,24ff.
68
ibid. 106,11ff.
69
Dexippus in Cat. 10,3–6.
182 Aristotle Transformed

Boethius did not disagree.70 ‘In view of the fact that simple concepts do not
yet share in truth or falsity it is obvious that the vocal expression (prolatio) of
such a simple concept must also be beyond either. But when a composition or
division involving ‘to be’ [compositio secundum esse facta vel etiam divisiolearning,
and so epiprosthen tôn ôikeiôn tithentai ta allotria. occurs in the concepts, –
which is where truth and falsity are primarily engendered – then, since the words
receive their meaning (significatio) from the concepts, they [i.e. the words] must
also be true or false according to the quality [i.e. truth-value] of the concepts.’
(The translation paraphrases slightly.)
Ammonius concurs;71 ‘And it will be established [sc. in de Interpretatione] that
nouns and verbs are simple words, which signify neither truth nor falsity [. . .]
but the true and the false belong to (theôreitai peri) the declarative logos, which
results from the intertwining (sumplokê) of them; and that truth and falsity
belong to the concepts before the words, as the concepts are the causes of the
words. For concepts, too, are either simple or composite. The simple ones, which
are signified by the simple words, cannot be bearers of any truth or falsity; but
the composite ones, which are about composite things and are indicated by
intertwined words, can be bearers of truth and falsity. But to the things as such
no one could possibly ascribe such properties, not even to the composite ones.’
(The translation paraphrases slightly.)
We have no ancient discussion of the ‘disambiguated’ sentences (V–VII );
(V–VI ) were the object of discussion in Western medieval scholasticism, and we
can only suppose that the Greeks would have declared that they do not solve the
problem of equivocation for the same reasons as their Western colleagues did.72
(VII ) is hinted at by the same Western schoolmen. We have no discussion of it
in Greek, but Dexippus, at least, seems to have held that sentences that contain
words to which no significatum corresponds or whose significatum is a concept
of a fictitious entity are neither true nor false.73 (VIII ) is dismissed as meaningless
by Boethius.74
Examples like (I–IV ) are discussed in all commentaries on de Interpretatione.
Ammonius’ comments on 17a15–17 and 17a35–775 give us fair compensation for
the loss of Porphyry’s ipsissima verba.

70
Boethius in Int.2 p. 49 (Meiser).
71
Ammonius in Int. 18,2–12; cf. 21,4ff.
72
cf. my ‘Can equivocation be eliminated?’, Studia Mediewistyczne 18.2, 1977, 103–24.
73
Dexippus in Cat. 7,30–8,1 in particular, and pp. 7–10 in general.
74
Boethius in Int.2 II , ch. 8, p. 182,17–20 (Meiser) (Boethius uses another example).
75
Ammonius in Int. 72–5; 84–6.
 Porphyry’s legacy to logic: a reconstruction 183

Having established the principle that counting of propositions must happen

on the level of significates, Ammonius exemplifies as follows;76 ‘If we say “Ajax
duelled with Hector” without determining which of the two Ajaxes we are
making a statement about, no reasonable man would say that this is one
proposition as we see it to be simultaneously true and false.’
Some pages later77 he explains that ‘Ajax duelled with Hector’ and ‘Ajax did not
duel with Hector’ may both be true and both be false although they appear to be
constituted of the same terms; however, the identity between the proposition
which may truly be affirmed and the one that may at the same time be truly
negated is only an identity on the level of expression, not on that of the significata.
To make it necessary for the affirmative and the negative propositions – and this
time Ammonius ought to mean ‘on the level of expression’ – to have opposite
truth-values, one must disambiguate the expression so that it is clear about
which Ajax one is talking.
He returns to the question in a scholium78 on 18a26–7, explaining that a pair
like ‘this [or: there] is a white cloak’ and ‘this [or: there] is not a white cloak’ are
exactly parallel to ‘Ajax duelled with Hector’ – ‘Ajax did not duel with Hector’, if a
convention states that ‘cloak’ is a name that can be used as a substitute for ‘horse’
as well as for ‘man’. In such cases, he says, the use of an equivocal expression
makes it possible for the two sentences (protaseis) to be both true; both false; or
one of them true, the other false.
The terminology in which protasis, logos etc. are used both of entities on the
level of expression and of their significates creates many difficulties of
interpretation, and certainly helped confuse the discussion already in antiquity.
Intending, no doubt, to be faithful to Porphyry’s way of thinking, Boethius
on one occasion79 explains that ‘Alexander abducted Helen’ will have different
truth-values when intended as a statement about the son of Priam and when
about the king of Macedonia. His wording strongly suggests that he considers
‘Alexander abducted Helen’ to be one proposition that changes from true to
false when compared to one of the two portions of reality signified by it
rather than to the other, so that in this account the mental language has
dropped out, vocal language being seen as having realities for its immediate
significates.

76
ibid. 73,11–14.
77
ibid. 84,30–85,3.
78
ibid. 128,3–14.
79
Boethius in Int.1 1, ch. 6, pp. 81–2 (Meiser).
184 Aristotle Transformed

Certain passages in Ammonius can be read in the same way. Thus in the scholium
on 18a26–7 he says80 that the reason why a proposition (protasis) with an equivocal
subject and its negation need not have different truth-values is that nothing prevents
one of the several things signified by the equivocal word from participating and
another of those things from not participating in that which is predicated.
Yet, it is clear that the thought that underlies all of his discussion of ambiguity,
and which is a Porphyrian thought, is that truth and falsity are strictly speaking
properties of the propositions of mental language,81 even if vocal sentences may
be called true or false in a derivative sense. The principles of contradiction and
excluded middle apply to the mental propositions only – but there they certainly
apply.82 Porphyry is known to have defended them,83 and the context in which
his defence is mentioned even suggests that he said something like, ‘In spite of
the counter-argument that may be derived from ambiguous sentences, the
principles must be upheld.’
Porphyry’s teaching was never quite forgotten as long as scholasticism lasted,
and the discussion of ‘Ajax duelled with Hector’ was a standard item of
commentaries on de Interpretatione and the Sophistici Elenchi till the end of the
Middle Ages.84 No one ever ventured a frontal attack on the great scholar’s
positions and consequently we find fragments of Porphyrian doctrine even in
the works of Michael of Ephesus and Leo Magentinus. In fact, their works are
scarcely intelligible if one forgets about the Porphyrian background. But it is also
clear that few of his successors had his ability to remember trivialities in the
right places. No one doubted the importance of mental language, but most
commentators became like Michael of Ephesus who duly mentions it in his
scholia on the initial chapters of de Interpretatione85 and in one scholium on the
Elenchi86 – but who forgets about it in his comments on chapter 17 of the
Elenchi,87 where it is obvious that many difficulties disappear if one remembers

80
Ammonius in Int. 128,4–14.
81
cf. Boethius in Int.2 I, ch. 1, p. 28 (Meiser) (a Porphyrian passage: cf. p. 26,17 (Meiser)).
82
Except, perhaps, if they are affirmative statements about fictitious entities, though Ammonius in Int.
62 says that any negative proposition about ficta is true and seems to hold that any affirmative
proposition about them is false.
83
Boethius in Int.2 II , ch. 6, p. 134,20ff (Meiser).
84
See, e.g., Psellus Paraphrasis Int., MS Vat. Barb. gr. 164: 5v; Michael Ephesius (?) Sch. in Int., MS
Paris. gr. 1917: 24rM; Leo Magentinus in in Int. 17a34, MS Vat. gr. 244: 106r sch. mb’ and MS Paris.
gr. 1917: 25r; ps.-Magentinus in Int., MS Paris. gr. 1928: 141r; Blemmydes Logica, PG 142: 900C
(from Ammonius?); Comm. 3 in SE 175b7, cod. 150: 200r. Related examples in ps.-Alex.-1 in SE
129,8–9; 141,6–7; 179,6ff.; Comm. 3 in SE 176a3–4 (sch. 19).
85
MS Paris. gr. 1917:17r ff. Some extracts are printed in CAG 4.5, pp. xlv–xlvii.
86
ps.-Alex.-1 11,23–8.
87
ibid. 128,14ff.
 Porphyry’s legacy to logic: a reconstruction 185

the distinction between non-ambiguous true or false mental propositions and

surface sentences, which may contain ambiguities and hence be excepted from
the principles of contradiction and excluded middle.
Quite generally, it may be said that the post-Porphyrian commentators tend
all too often to regard complex concepts (mental propositions) and complex
expressions (sentences) as equally suited to be the bearers of truth and falsity,88
i.e. they tend to overlook the counter-argument derivable from the occurrence
of ambiguous sentences. They also, reasonably enough against such a background,
tend to disregard mental propositions altogether, leaving us with sentences and
facts or things of the world as the only relevant elements in the semantic process.
These tendencies are quite obvious in all extant collections of scholia on the
Elenchi.
To some extent, Porphyry himself carries the responsibility for the fate of his
logic. He was a man of compromises, at least as far as expression goes. Hardly a
single piece of new terminology can be shown to be of his making, and hardly a
single piece of theory can be demonstrated to originate with him. What he did
was to accept what everybody accepted and give some pieces of doctrine a twist
so as to fit into the rest. He was more consistent than most people in his dealings
with the ‘triangle of signification’, but he probably never thought of eliminating
the dangerous ambiguity of semantic terms like ‘discourse’ (mental or vocal) and
‘significatum’ (mental or real). He tended to say: ‘Primarily truth belongs to a
logos viewed this way, and derivatively to a logos viewed that way’, rather than:
‘Sentences are neither true nor false; propositions are either true or false’, just as
he said: ‘The notion of man is twofold’, rather than: ‘There are several concepts
sharing a common origin and a common name, but which are distinct concepts’.
What distinguishes Porphyry from his followers is that he treats the differences
indicated by such locutions as ‘qua this . . . qua that’ as important differences, so
much that I think it is justifiable to translate his ‘logos qua this’ as ‘mental
proposition’ and his ‘logos qua that’ as ‘sentence’ – whereas Ammonius and later
commentators often seem to think that ‘qua this’ or ‘qua that’ does not make any
important difference.

88
cf., e.g., ps.-Magentinus in Int., MS Paris. gr. 1917: 18v-19r; Stephanus in Int. 6,17–21; Michael
Ephesius in Int., MS Paris. gr. 1917: 18vM; Leo Magentinus in Int., MS Vat. gr. 244: 94r sch. ê.
186
 8

How did Syrianus regard Aristotle?*

H.D. Saffrey

The great inquiry which Paul Moraux initiated into the history of the Aristotelian
tradition will cover only the period from the first century bc to the second century
ad.1 The results of this great work will obviously be decisive for appreciating
the use which first Plotinus2 and then his Neoplatonic successors made of
Aristotle’s treatises in order to develop their own philosophy. Porphyry had been
able to write that the Enneads were a ‘digest’ of Aristotle’s Metaphysics.3 This
observation expresses the profound regard Porphyry had for Aristotle, and
we shall see that this regard remained as a matter of tradition in Athenian
Neoplatonism. By taking us to the very beginning of the fifth century ad, I would
like to offer to Paul Moraux, as a token of my esteem, some items for the record by
showing how Syrianus, the master of the Neoplatonic school in Athens, regarded
Aristotle.
Syrianus is the author of an extant commentary of Platonic inspiration on
books 3, 4, 13 and 14 of the Metaphysics4 and of commentaries on the Categories,
the de Interpretatione, the Prior Analytics, the de Caelo and the de Anima, which
are lost and are known to us only from later authors. The first document that
I would like to submit for examination is the prologue to the commentary

* Translated by Victor Caston.

1
Paul Moraux, Der Aristotelismus bei den Griechen von Andronikos bis Alexander von Aphrodisias.
Bd. 1, Die Renaissance des Aristotelismus im I Jh. v. Chr. (Peripatoi 5), Berlin 1973. Bd. 2, Der
Aristotelismus im I und II Jh. n. Chr. (Peripatoi 6), Berlin 1984. The third volume, dealing with the
doctrine of Alexander of Aphrodisias, will be published under the direction of J. Wiesner.
2
For a specific example of the influence of Alexander of Aphrodisias and Aristotle on Plotinus, cf. Paul
Henry, ‘Une comparaison chez Aristote, Alexandre et Plotin’ in Les Sources de Plotin, Entretiens sur
l’Antiquité classique, vol. 5, Vandoeuvres-Génève 1960, 427–49. As P. Henry notes in this article, it is
Émile Brehier to whom we owe the finest remarks on the influence of Aristotle and Aristotelianism
on Plotinus. It is obvious that the work of Paul Moraux will help to revive this important question. In
the meantime, see T.A. Szlezák, Platon und Aristoteles in der Nuslehre Plotins, Basel 1979, 135–43.
3
cf. Porphyry, Vita Plotini 14,5–7.
4
On this commentary generally, cf. Karl Praechter, ‘Syrianos’, 1, in RE IV A, cols 1769,56–1775,2.

187
188 Aristotle Transformed

on book 13 of the Metaphysics. As is well known, the long prologues to the

Neoplatonists’ principal books have in general a programmatic value.5 This
prologue proves to be no exception:

Syrianus, in Metaph. 13 (CAG 6.1), pp. 80,4–81,14:

I do not count myself among those who systematically try to pick a quarrel with
Aristotle, nor indeed among those who make him their master with regard
to a small number of subjects or on trivial points. On the contrary, I belong
among those who admire in an ordinary manner his logical methods and who
appreciate his moral treatises and natural philosophy a great deal. And – so as
not to appear too lengthy in enumerating here all the very fine and excellent
doctrines which this great philosopher produced – who6 would not rightly
admire, if he is endowed with good sense, whatever in this very perfect treatise
is said of relevance to the subject of forms united with matter and to the
definitions, or whatever is pointed out by means of demonstrations appropriate
to the subject of divine and immobile causes which transcend the world entirely,
although they are objects which surpass all composite thought and all rigorous,
discursive argumentation? Indeed, who would not name the author of this work
the benefactor of humanity?
For these reasons, we owe recognition no doubt as great as do those who have
sensed the sagacity of his spirit.7 Yet since – for what reasons I do not know – in
other parts of his treatise on theology no doubt, but particularly in the last two
books, 13 and 14, Aristotle was moved to charge violently in an assault on the
first principles of Pythagoras and Plato, without saying anything sound or
adequate against them and even, if one must speak the plain truth, without
succeeding for the most part in touching them (because in these attacks he puts
forward his own presuppositions), it appeared reasonable, in order to prevent
the newest among my auditors, impressed by the author’s deserved reputation,
from being borne away by the flood of contempt for the divine realities and
the Ancients’ divinely inspired philosophy, to submit the text to a critical and
impartial examination using the best of our abilities and to show that the

5
On the prefaces of Proclus, for example, see H.D. Saffrey, ‘Théologie et anthropologie d’après
quelques préfaces de Proclus’, in Images of Man in Ancient and Medieval Thought (= Studia Gerardo
Verbeke . . . dicata), Leuven 1976, 199–212.
6
I have adopted here Praechter’s correction (in Gött. gel. Anz. 1903, 517; = Kleine Schriften,
Hildesheim 1973, 250); I read tis ou [pas] (80,9).
7
tês ankhinoias (80,16). In his in Tim., 1,24–2,1, Proclus says: ‘That Plato, right down to the finest
detail in which he wrought the present instruction, was alone able to preserve from beginning to
end the traits which characterise Pythagoreanism in natural science – this is what the most
sagacious spirits must examine (tous ankhinousterous)’ (from the translation of Festugière). For
Proclus, the sagacity of Aristotle did not reach quite that far.
 How did Syrianus regard Aristotle? 189

doctrines concerning Plato’s and Pythagoras’ first principles remain unrefuted

and unshaken, and that Aristotle’s undertakings against these doctrines most
often miss their mark8 and examine subjects which in no way concern these
divine men, whereas in a small number of cases they strive to say something no
doubt which holds against them and to criticise them, but without being able to
effect a refutation, whether great or small. And it had to be thus, since ‘the truth
can never be refuted’, as this divine man says [Plato, Gorgias 473B10–11], and
because the fathers of these treatises concerning first principles, by making them
resemble the realities, founded them ‘in durability and infallibility to the degree
attainable by treatises’ [Timaeus 29B6–8].
But enough of preambles. And since we have just stated the intention and
plan of this work, it is time to engage in the combats9 themselves. However that
may be, whether we should be called a combatant because we answer charges
brought against the finest and best of philosophies, or whether we should rather
be called a judge who decides between the most simple and most intellective
intuitions of the epoptic philosophy of Pythagoras’ pupils and the aporias raised
in a dialectical manner against them by one who in this field is the strongest of
all that have ever been heard speak, Aristotle, we will not refuse to bear these
titles; the only thing which matters is to conduct these battles with justice,
intelligence and love of the truth.

This fine text begins with a declaration of intent. Syrianus intends to distinguish
himself from his predecessors: he does not reject the authority of Aristotle purely
and simply, and he does not have the hypocrisy to appeal to him only where his
intervention is without importance and solely decorative. In the long dispute
over the agreement between Aristotle and Plato, he maintains an equal distance
from the two extremes, and strives to separate the Platonic and Aristotelian
positions.10 This is why he emphasises his admiration for Aristotle the logician,
the moralist or natural philosopher, but withholds judgment when he exposes

8
para thuras apantôsas (80,28). On this phrase, cf. L.G. Westerink’s note on Damascius, in Phaed. 1,
§380,6 (207): ‘ “come to the wrong door”, that is, answer the wrong question; or, more vaguely,
probably owing to a confusion with apantan “answer”: give the wrong answer, a number of times in
Olympiodorus, in Gorg. (see index).’
9
It seems that tôn agônôn has its normal sense here, since in what follows Syrianus will claim in turn
to be a combatant and a judge. Yet in rhetoric, the opposition of agônes to prooimion is known in the
attenuated sense of the development of the treatise in contrast to the prologue, and this sense is
known to Syrianus himself, cf. in Herm. 2 (Rabe), 111 and 170, and Olympiodorus, in Alc. 213,3–4.
10
For the Neoplatonists’ appreciation of Aristotle, see R.T. Wallis, Neoplatonism, London 1972, 24–5
and 143–4, and Fr. Romano, Studi e richerche sul Neoplatonismo, Naples 1983, ch. 3,‘Lo “sfruttamento”
neoplatonico di Aristotele’, 34–47. For the assimilation of Aristotelian logic, cf. A.C. Lloyd,
‘Neoplatonic and Aristotelian logic’, Phronesis I, 1955/56, 58–72 and 146–60, and his ‘The
assimilation of Aristotle’s logic’ in The Cambridge History of Later Greek and Early Medieval
Philosohy, ed. A.H. Armstrong, Cambridge 1967, 319–22.
190 Aristotle Transformed

Aristotle’s attacks concerning the Platonic first principles. Proclus, even more
than his teacher, will note the divergences between Aristotle and Plato, since he
will pass the same global judgment on the Physics that Syrianus passes on the
Metaphysics, and will even go so far as to say: ‘As for the great Aristotle, it is my
opinion that he arranged as much as possible his treatise on nature in a spirit of
rivalry with the teachings of Plato’,11 and his commentary on the Timaeus will at
every opportunity establish the superiority of Plato over Aristotle.
That having been said, however, it is necessary to notice the wholly exceptional
title which Syrianus gives to Aristotle when he proclaims him the ‘benefactor of
humanity (euergetên tou tôn anthrôpôn biou)’. In itself, the mere epithet of
‘benefactor’ is ordinary – all those to whom you owe some kindness are your
‘benefactors’: the cities used to declare their great men the ‘benefactors of the
city’, and a friend or parent who wished to honour a dead man to whom he owed
much would often have engraved on the tomb that he was his benefactor.12 To
remain within Neoplatonic usage, one might compare what Proclus said of his
teacher, Syrianus: ‘If we had to repay the debt of gratitude, incurred by his
kindnesses to us (tôn euergesiôn), not even the whole of time would suffice.’ Thus,
Proclus considered Syrianus as his benefactor. We shall come back to this.
Ammonius, for his part, said the same thing of Proclus when he called him ‘our
teacher and benefactor (euergetês)’.13 But the appellation ‘benefactor of humanity’
is much more significant. To my knowledge, it was more or less reserved for the
emperor himself, who is often acclaimed with the formula ‘Saviour and benefactor
of the entire world’, for the reason that his providence in fact exerted an influence
over the whole of the empire.14 Yet it is open to doubt whether Syrianus truly
intended to compare Aristotle to the emperor. Looking elsewhere, it should be
remembered that the Stoics had taught the Greeks to deify the first inventors of
the sciences and crafts as the great benefactors of humanity, for the reason that
their inventions were made for the benefit of all men.15 Aristotle, for Syrianus,
might be considered the inventor of theology, since his treatise, the Metaphysics,

11
Proclus, in Tim. 1,6,21–4 and its whole development until 7,16.
12
cf. B. Kötting, s.v. ‘Euergetes’ in RAC 6, cols 848–60, especially cols 858–9, ‘Verwendung in
nachkonstantinischer Zeit’.
13
Proclus, Theol. Plat. (Saffrey-Westerink) 1.1,7,10; Ammonius, in Int., CAG 4,5 (Busse) 181,30–1.
14
cf. A.D. Nock, ‘Soter and Euergetes’ in The Joy of Study, Papers . . . F.C. Grant, New York 1951, 127–48,
reprinted in Essays on Religion and the Ancient World, Oxford 1972, 720–35. On the emperor as
euergetês, cf. K.M. Setton, The Christian Attitude towards the Emperor in the Fourth Century, New
York 1941, 51.
15
On this subject, see the arguments of Cicero in de Nat. Deor. 1.15.38, with Pease’s note on the words
‘a quibus . . . utilitas’, and 2.24.62, with Pease’s note on the words ‘beneficiis excellentis viros’.
 How did Syrianus regard Aristotle? 191

had ‘Theological Treatise (theologikê pragmateia)’ as its title, at least in Syrianus’

day, which he even mentions in the text cited above when he speaks of certain
‘parts of his treatise on theology’.16 Did Syrianus think that Aristotle, in creating
the first treatise of theology proper, had merited being called, on a level with the
emperors, the ‘benefactor of humanity’?
Whatever the reason, this benefactor of humanity will have to be contested by
Syrianus in the name of the first principles of Pythagorean and Platonic
philosophy, by which one must understand the philosophy of the One. This is
why, in engaging in the commentary on book 13 of the Metaphysics, Syrianus
tells us that he is in fact engaged in a combat in which he will be one of the
participants and, in turn, the judge who will determine the victor. It is again as a
judge between Plato and Aristotle that Syrianus is presented by Ammonius in his
commentary on the de Interpretatione, when he says:

The great Syrianus has judged this question in an excellent and very penetrating
manner, by showing very clearly what is the nature of these realities (the
contraries) and their mutual distinction, by emphasising the agreement of our
common and innate notions with the words of the treatise, by examining each
of Aristotle’s arguments in themselves, and by teaching through an analysis of
the wording of the treatise what in this text has been transmitted correctly,
on the one hand, and, on the other, in a way no doubt probable but inexact.17

We see that here also Syrianus uses the same method, dividing the true from the
false, the exact from the approximate, the argument which holds from the one
which runs aground. We know, moreover, that at least once in the same
commentary Syrianus dares to contradict Aristotle and show the falsity of his
doctrine.18
In this sense, the great text of the commentary on book 13 of the Metaphysics
ends with a second declaration of intent. Far from wanting to indulge in a base
polemic with a philosopher whom he admires and respects, it is in the name of
‘justice, thought and the love of the truth’ that Syrianus will judge the debate

16
en allois men meresi tês theologikês pragmateias (80,17); see also Elias, Prolegomena Philosophiae,
CAG 18,1 (Busse), 20,19–20: ‘for it is written in the theological treatise, the Metaphysics (en gar têi
Meta ta phusika pragmateiai theologikêi epigegrammenêi . . .)’, a name based on the Aristotelian
division, Metaph. 6.1, 1026a18–19: ‘so that there would be three contemplative philosophies:
mathematical, natural, and theological (hôste treis an eien philosophiai theôrêtikai, mathêmatikê,
phusikê, theologikê)’, with Ross’s remark (ad loc.): ‘The designation of metaphysics as theologikê is
confined to this passage and the corresponding passage in 11,7, 1064b3’.
17
Ammonius, in Int., CAG 4,5 (Busse), 253,12–17.
18
cf. L. Tarán, Anonymous Commentary on Aristotle’s De Interpretatione (Beiträge zur klassischen
Philologie, Heft 95), Meisenheim am Glan 1978, 120,13–14: prôtos de Surianos ho philosophos
anteipen kai apodeiknusin pseudos on.
192 Aristotle Transformed

which he initiates between the philosophy of Aristotle and that of Pythagoras

and Plato.
These texts reveal to us Syrianus’ philosophical qualities and bear testimony
to a tradition that engendered great respect for Aristotle in the Neoplatonic
school of Athens, even when one is committed to refuting him. It should no
longer be surprising that the best commentaries on Aristotle, those of Simplicius,
are the last productions of this Athenian Neoplatonic school, even after the
closing of the school by Justinian in 529.
Yet another document confirms this favourable impression. Syrianus had been
the teacher of Proclus, and we said that the latter celebrated Syrianus as his
benefactor because of the perfect education which he had received from him.
Now, the manner in which Syrianus proceeded in educating Proclus has been
described by Marinus in his writing entitled Proclus, or On Happiness. For the first
anniversary of the death of his teacher (485), Marinus composed a sort of funeral
eulogy, whose theme is the happiness the philosopher acquires by practising the
six degrees of virtue. To illustrate his subject, Marinus recounts numerous facts
from Proclus’ life, and in particular the education which he received from
Syrianus. The division of this treatise into chapters was introduced in 1700 by J.A.
Fabricius and was followed afterwards by all the editors including J.F. Boissonade,
the author of the last edition in 1814.19 A summary of these texts follows.
(Proclus’ Aristotelian education had in fact begun before his arrival in Athens,
although, even in Alexandria, he attended the course of a professor named
Olympiodorus, of whom we know practically nothing.20)

1. Marinus, Proclus §9 (in Alexandria).

He begins to attend the lectures of professors on the philosophy of place.

For Aristotelian studies, he would frequent the school of the philosopher
Olympiodorus, who had a great reputation. For mathematics . . . Whenever he
attended the lectures of Olympiodorus – who had a great facility in speaking
and who, because of the rapidity of his elocution and the sublimity of his
argumentation, was accessible only to a small number of his auditors – Proclus,
after leaving the lecture, would repeat its entire content from memory before his
fellow students in its original formulations even though there was a great deal
of it . . . Moreover, Proclus even understood the logical treatises of Aristotle
easily, and this after a simple reading, although they are difficult for those who
study them to understand.

19
J.F. Boissonade, Marini Vita Procli, Leipzig 1814.
20
cf. PLRE, s.v. ‘Oympiodorus’ 2.
 How did Syrianus regard Aristotle? 193

2. Marinus, Proclus §12 (once in Athens, Syrianus puts the young student in the
hands of the old Plutarch of Athens).

Proclus thus read Aristotle’s treatise On the Soul and Plato’s Phaedo with Plutarch.

3. Marinus, Proclus §13 (after the death of Plutarch, Syrianus takes him into his
own hands).

In less than two complete years, Syrianus read all the treatises of Aristotle with
him, the logical, ethical and political and those on the theological science (the
Metaphysics) which surpassed them all. Then, once he had been well introduced
by these words as if by certain preparatory sacrifices and small mysteries, he was
led step by step to the mystagogy of Plato . . .

4. Marinus, Proclus §14.

Proclus acquired the political virtues by reading Aristotle’s Politics and Plato’s
Laws and Republic.

From these four texts, we can extract precious information on the education
Proclus received as regards Aristotelian philosophy. We note that from an early
age, Proclus learned Aristotelian logic, of which he had a thorough mastery;
he knew by heart the treatises which he understood in a single reading (text 1).
On the other hand, in the Neoplatonic school at Athens, the complete study of
Aristotle had always been considered as preparatory to the study of Plato:21 the
reading of the de Anima as preparation for the reading of the Phaedo (text 2), the
study of logic, ethics, politics, physics, and metaphysics as preparation for
Platonic ‘mystagogy’ (text 3); the reading of the Politics as preparation for the
reading of the Laws and the Republic (text 4). One sees this: if the Neoplatonists
held that assuredly Plato’s philosophy surpassed Aristotle’s in every respect22
they did not for all that neglect extensive study of the Aristotelian treatises. It
suffices to read their writings to collect the evidence of their genuine familiarity
with Aristotelian philosophy.23
This regard for Aristotle in the Neoplatonic school of Athens and with
Syrianus in particular must be interpreted within the context of the long debate
on the problem of the agreement between Aristotle and Plato. In the period in
which Aristotle was known perhaps solely through his exoteric works, the thesis

21
This is also Themistius’ point of view, cf. Orat. (Downey-Norman) 20, vol. 2, 6,10–19; Aristotelianism
serves as the proteleia to the epoptia which is Platonism.
22
cf. L.G. Westerink, Anonymous Prolegomena to Platonic Philosophy, §9, 24–34.
23
See, for example, the index of the different books of Proclus’ Theol. Plat.
194 Aristotle Transformed

of agreement could be easily invented by Antiochus of Ascalon and diffused by

Cicero. But familiarity with the great Aristotelian philosophical treatises must
have modified radically the idea that people had of the relation between Aristotle
and Plato. The thesis of agreement, rejected first by Atticus and taken up again
immediately by Alcinous in the form of the doctrine that the Ideas are the
thoughts of God, could be maintained by putting Plato’s intelligible realm within
Aristotle’s transcendent Intellect. With Plotinus, the correlation of the intelligible
and the intellect came to entail a superior principle, beyond the Intellect and
Being, the One. In this way, Neoplatonism’s fundamental hierarchy of the three
hypostases was established.
From this point, what Aristotle will come to be reproached with continually
is precisely that he stopped at the Intellect in the doctrine of the first god, without
ascending further to the One.24 In the confines of the inferior hypostases a
certain agreement between Aristotle and Plato will always be sought, but the
superiority of Plato will always be loudly affirmed because of his doctrine of
the first hypostasis, the One, the first god. This is the position which we have
discovered in Syrianus, when in his commentary on books 13 and 14 of the
Metaphysics he sets Pythagoras’ and Plato’s doctrine of first principles in
opposition to Aristotle’s. Syrianus taught this position to Proclus, and through
him to all the philosophy of the Middle Ages and above all the Renaissance.25

24
See, for example, Proclus Theol. Plat. 2.4.
25
cf. Joseph Moreau, ‘De la concordance d’Aristote avec Platon’ in Platon et Aristote à la Renaissance
(De Pétrarque à Descartes XXXII ), Paris 1976, 45–58.
 9

Infinite power impressed: the transformation of

Aristotle’s physics and theology
Richard Sorabji

The focus of this chapter will be on an argument about infinite power, which was
started by Aristotle, but which was subsequently connected with the idea of a
Creator God.
In the thirteenth century, Bacon, Bonaventure and Thomas Aquinas all
thought, in one sense or another, that Aristotle’s God was a Creator and Sustainer
of the universe.1 Some thought that he was a Creator in the sense of being
causally responsible for the beginningless existence of the universe, some in the
sense of giving it a beginning. Any such belief in his causal role is hard to credit,
for the main action of Aristotle’s God in the world seems to be merely that of a
mover. He moves the heavens, but he does not seem to give them existence.
Admittedly, he has a certain indirect responsibility for the existence of compound
bodies down below the heavens. For he unconsciously inspires the sun’s motion,
which by the obliquity of its angle in turn produces seasonal disturbances, and
so makes earth, air, fire and water turn into each other, relocate themselves and
mingle to form new compounds. But it is not obvious that he has even indirect
responsibility for the existence of the main masses of earth, air, fire and water, or
of the heavens themselves.
The contrary view of Aristotle’s God as Creator or Sustainer was based partly
on spurious works, several of them associated with the ninth century ad. A
particularly important one, to which I shall shortly return, is the Theology of
Aristotle. Others which mention a Creator or Sustainer God and which were
wrongly attributed to Aristotle are the early de Mundo from around the beginning

1
For references, fuller detail and translations of the main passages discussed below, see Richard
Sorabji, Matter, Space and Motion, London and Ithaca NY, 1988, ch. 15. The materials were first
prepared for the Charles Schmitt Memorial Colloquium at the Warburg Institute, and presented
there, as here, in abbreviated form.

195
196 Aristotle Transformed

of our era, the Liber de Causis and the Secretum Secretorum. But I want to draw
attention to a quite different kind of influence, that of the Neoplatonist
commentator Ammonius (c. 435/45 – c. 517/26 ad ). Ammonius maintained that
Aristotle did indeed make God the efficient cause of the world, in the sense of
making him causally responsible for its beginningless existence. His aim was to
harmonise Aristotle with Plato, for Plato’s God in the Timaeus was a Creator, and
Ammonius took him to be a Creator in the same sense, of vouchsafing a
beginningless existence. Plato makes his Creator explain that because the heavens
are a composite entity, they can in principle be dismantled but that they will in
fact be held together by his good will.2 One later interpretation among others
was that the heavens are a compound of Aristotelian matter and form. On this
view, the prime matter might be expected to survive a process of dismantling,
but the heavens, being a compound of prime matter with form, would not.
Ammonius was not the first to represent Aristotle’s God as a Creator or
Sustainer. Arius Didymus was Stoicising in the first century bc , when he made
Aristotle’s prime mover ‘hold together’ (sunektikon) the heavenly bodies.3 In the
fourth century ad, St Ambrose was reading or misreading an early dialogue, On
Philosophy, from Aristotle’s days in Plato’s school, when he characterised
Aristotle’s efficient cause as the operatorium who brought things about merely by
willing. This does not represent Aristotle’s mature view and may represent only
the view of an interlocutor in the now fragmentary work.4 A desire to harmonise
Platonism with Aristotle accounts for the Neoplatonist Dexippus, in the same
century as Ambrose, understanding Aristotle’s talk of dependency (êrtêtai) in
Metaphysics 12 as meaning that substances owe their unity and at least in some
cases their life and form to the supreme Neoplatonist deity, the One.5 A still more
extreme harmonisation is attempted by the fifth-century Neoplatonist Hierocles,
when he ascribes to both Plato and Aristotle belief in divine Creation not out of
any substratum. He further credits for the general thesis of harmony his teacher
Plutarch of Athens, and an earlier Ammonius – Ammonius Saccas, who taught
Plotinus in the third century ad.6 But the interpretation of Aristotle’s God as
Creator or Sustainer did not yet stick. We shall see that Plutarch’s other two

2
Plato Timaeus 41A–B.
3
Arius Didymus, fr. 4, in Dox. Gr. 450,16.
4
Ambrose Hexaemeron 1.1.1–2, p. 3, 10–13 (Schenkl). The imprint in mud of a papyrus recently
found in Afghanistan may contain a further fragment of On Philosophy, assigning to God a different
causal role, that of making things participate in Forms: P. Hadot, ‘Les textes littéraires grecs de la
trésorie d’aï Khanoum’, Bulletin de correspondence hellénique 111, 1987, 225–66.
5
Dexippus in Cat. 40,28–41,3; 41,7–18, with reference to Aristotle Metaph. 12.7, 1072b14.
6
Hierocles ap. Photium Bibliotheca 171b33ff; 172a22ff (Bekker).
 Infinite power impressed 197

pupils, Syrianus and Proclus, resisted it. So it was left to Ammonius to drive it
home, and we are told that he wrote a whole book on the subject.7
Ammonius’ influence on later Greek Neoplatonist commentators seems to
have been decisive. His view of Aristotle’s God was published by Asclepius and
endorsed also by Philoponus, Simplicius, Olympiodorus, Elias and (if he is the
author of the extant commentary on Aristotle’s de Anima, Book 3) Stephanus.
Even more important, Ammonius is cited, along with the spurious Theology of
Aristotle, as an authority for the view of Aristotle’s God as Creator by the Islamic
philosopher Farabi (c. 873–950 ad ), in his work The Harmony of Plato and
Aristotle,8 although Farabi diverges from Ammonius in denying that Aristotle’s
cosmos is eternal. I shall draw attention below to a neglected passage in which
Avicenna (c. 980–1038) accepts that Aristotle’s God is a Sustainer, and
Maimonides (1135–1204) repeats that Aristotle’s God is the efficient cause of the
world’s beginningless existence.9 Farabi was read by Avicenna, Farabi and
Avicenna were read by Maimonides, and Maimonides was read, in Latin
translation, by our thirteenth-century authors. So there is a perfectly possible
route by which Ammonius may have acted, alongside other influences, to shape
the thirteenth-century misreadings of Aristotle’s theology.
Tantalisingly, Farabi tells us that Ammonius’ arguments are too well known to
need repeating. But Simplicius (writing after 529 ad ) gives a series of arguments
drawn from Ammonius, the last of which is the argument about infinite power.10
The original argument on infinite power appears in Physics 8.10, where
Aristotle maintains that what moves the heavens cannot be a spatially extended
entity, nor a power lodged in a spatially extended entity. For what produces
unending motion must be infinite, whereas any extended entity, and any power
lodged in an extended entity, is finite. If the mover is neither a spatially extended
entity, nor a power (or soul) lodged in one, this opens the way to the view made
explicit in Metaphysics 12.7, where the argument is repeated (1073a5–11), that
the mover is God: God can supply the required infinite power.
So far the argument concerns only the motion of the heavens. But it was
changed to an argument about their existence by the work of four successive
Neoplatonists who stand in a teacher-pupil relationship: Syrianus, Proclus,

7
Simplicius in Phys. 1363,8–12; cf. in Cael. 271, 18–21 and Farabi’s Harmony, discussed below.
8
Farabi, Harmony of Plato and Aristotle, with German translation, Dieterici (ed.), Alfarabi’s
philosophische Abhandlungen, Leiden 1892, 24,13–25,1 and 28,22–29,5.
9
Maimonides Guide for the Perplexed, part 2, ch. 21 (translations by M. Friedländer, London 1904,
and S. Pines, Chicago 1963). Maimonides’ Aristotle believes, but does not think he can prove, the
world to be beginningless, part 2, ch. 15.
10
Simplicius in Phys. 1361,11–1363,12, esp. 1363,4–12.
198 Aristotle Transformed

Ammonius and Philoponus. Syrianus (died c. 437 ad ) protests that Aristotle

should have admitted that the source of the heavens’ infinite motion is also the
source of their unending being. Aristotle would then have been in agreement
with Plato, as Syrianus interprets him. Plato even implies at Phaedrus 245E, in
Syrianus’ view, that in giving motion to the heavens, God thereby gives them
being.11 The connecting link between motion and being is not explained, but one
might be found in the view later propounded by Aristotle, that circular motion
is the defining characteristic of the celestial body. That might well be taken to
imply that the preservation of its motion guarantees its preservation as the
celestial body. A connexion between motion and being such as that suggested by
Syrianus was already hinted at by Aristotle’s immediate successor Theophrastus12
and is later found on one interpretation in Simplicius,13 in the passage of
Avicenna to be discussed below, and in Abravanel.14
But of more relevance is the argument suggested by Syrianus’ pupil Proclus
(c. 411–485 ad ). For instead of moving directly from motion to being, Proclus
suggests doing so indirectly, by means of reapplying the infinite power argument.
The same infinite power argument, says Proclus, which excludes an infinite
power of moving from a finite body will also exclude from it an infinite power of
existing. So this power of existing must equally be derived by the cosmos from
elsewhere, most plausibly from the same divine source:15

In just the same way, everything shows that it [the cosmos] will obtain its infinite
power of existing from there [from the Intellect], because of the argument which
says that an infinite power never exists within a finite body . . . .
Something else, then, will give it the power of existing, and will give it not all
at once, since it would not be capable of receiving it all at once. It will give it,
then, in the amounts it can take, in a stream that flows and ever flows onto it. No
wonder the cosmos is for ever coming into being and never has being.

11
Syrianus in Metaph. 117,25–118,11. I am indebted for my account of Syrianus and Proclus especially
to Carlos Steel, who was kind enough to show me an advance copy of his article, ‘Proclus et Aristote
sur la casualité efficiente de I’lntellect divin’, in J. Pépin and H.D. Saffrey (eds), Proclus – lecteur et
interprète des Anciens, Actes du Colloque Proclus, CNRS , Paris 1987.
12
Theophrastus asks, Metaph. 6a5–13; 10a10–15, whether the heaven would stop existing as a heaven,
if it lost its rotation.
13
Simplicius in Cael. 301,4–7. So Robert W. Sharples, ‘The unmoved mover and the motion of the
heavens in Alexander of Aphrodisias’, Apeiron 17, 1983, 62–6, at 63, but Simplicius might instead
intend Proclus’ re-application of the infinite power argument, since he endorses that elsewhere (in
Phys. 1363,4–8).
14
Isaac Abravanel, The Deeds of God 9.9.
15
Proclus in Tim. (Diehl) 1,267,16–268,6 (quoted in part here) and in a work entitled Examination of
Aristotle’s Objections to Plato’s Timaeus, ap. Philoponum contra Proclum 238,3–240,9; 297,21–300,2;
626,1–627,20.
 Infinite power impressed 199

Syrianus and Proclus were both deprecating Aristotle’s failure to agree with
Plato on the issue, although they thought he agreed on many others. But Proclus’
pupil Ammonius takes the harmonisation of Plato and Aristotle further. The
argument sketched by Proclus is what Aristotle actually intended: he meant us to
re-apply the infinite power argument and conclude that God produced the being
as well as the motion of the heavens. This is the last of the arguments drawn by
Simplicius from Ammonius’ book:16

And if, according to Aristotle, the power of any finite body is itself finite, clearly
whether it be a power of moving or a power that produces being, then, just as it
must get its eternal motion from the unmoved cause, so it must receive its eternal
being as a body from the non-bodily cause. My teacher Ammonius wrote a whole
book offering many proofs that Aristotle thought God was also an efficient cause
of the whole cosmos, a book from which I have here taken over some items
sufficient for present purposes. More complete instruction on the subject can be
got from there.

Ammonius’ pupil Philoponus (c. 490–570s ad ) agrees that Aristotle makes

God the efficient cause of the world’s beginningless existence.17 But he himself is
a Christian, and accordingly he tries to rewrite the infinite power argument, to
support the Christian belief that God created the universe in the more
conventional sense of giving it a beginning. As a result, he makes the argument
much more complex, and his version has been very variously interpreted.18 But I
believe it can be seen to have four stages.
First, Philoponus argues on behalf of Christianity that, since the world cannot
contain infinite power, it must have a beginning and end.19 But he knows that

16
Simplicius in Phys. 1363,4–12. The best account of Ammonius’ theology that I know is that of
Koenraad Verrycken, God en Wereld in de Wijsbegeerte van Ioannes Philoponus, Ph.D. diss., Louvain
1985. The relevant section is summarised in Chapter 10 below.
17
Philoponus in Phys. 189,10–26; 240,18–19; in GC 136,33–137,3 (cf. 286,7); in DA 37,18–31; in An.
Post. 242,26–243,25.
18
The interpretation which seems to me right is that of Lindsay Judson, ‘God or nature? Philoponus
on generability or perishability’, in Richard Sorabji (ed.), Philoponus and the Rejection of Aristotelian
Science, London and Ithaca NY 1987. I shall be diverging in various ways from the interpretations
of H.A. Wolfson, ‘The Kalam arguments for creation in Saadia, Averroes, Maimonides and St.
Thomas’, Saadia Anniversary Volume, American Academy for Jewish Research, Texts and Studies 2,
New York 1943 at 202 and 240; id., The Philosophy of the Kalam, Cambridge Mass. 1970, at 377–8
and 381; H.A. Davidson, ‘John Philoponus as a source of medieval Islamic and Jewish proofs of
creation’, Journal of the American Oriental Society 89,1969, at 362; id., ‘The principle that a finite
body can contain only finite power’, in S. Stein and R. Loewe (eds), Studies in Jewish Religious and
Intellectual History Presented to Alexander Altmann, Alabama 1979, at 80 and n. 37; Michael Wolff,
Fallgesetz und Massebegriff, Berlin 1971, at 94–9; Carlos Steel, ‘ “Omnis corporis potentia est finita” ’,
in Jan P. Beckmann, Ludger Honnefelder, Gangolf Schrimpf, Georg Wieland (eds), Philosophie im
Mittelalter, Hamburg 1987, 213–24.
19
Philoponus contra Proclum 1,18–2,11; contra Aristotelem ap. Simplicium in Cael. 142,22–5; in an
Arabic summary of a later lost work, translated into English by S. Pines, ‘An Arabic summary of a
200 Aristotle Transformed

Proclus has an answer to this, and he records, as the second stage in the argument,
Proclus’ reply that the world can perfectly well be unending, provided that the
infinite power needed for this result is not housed within the world.20 He
therefore has to reply to Proclus, and he does so in the third stage of his argument
by reversing an idea about nature employed much earlier by Alexander of
Aphrodisias.21 Proclus claims to be following Plato, but he has failed to ask
whether Plato’s cosmos is imperishable ‘in accordance with the logos of its own
nature’, or rather through God’s overriding its nature. In fact, its nature as a
composite entity means that it can in principle be dismantled, so if it is
imperishable, this must be due to God’s overriding its nature. Proclus has
therefore misrepresented Plato, whose world must be perishable in accordance
with the logos of its own nature. It ought therefore to follow, by a principle which
Proclus himself accepts,22 that the world is also generable in accordance with the
logos of its own nature.23
This third stage of the argument is resisted by Simplicius, who says that the world
is not really perishable by its own nature, because it is by nature fitted (epitêdeios) to
receive the eternity that God gives it.24 Proclus in one passage concedes Philoponus’
point that his cosmos will have to be perishable by its own nature (ek tês oikeias
phuseôs),25 but in another insists that the heavens get their imperishability from
their own constitution (sustasis), which belongs to their essence, even though they
receive their constitution from God.26 He adds that the cosmos ‘by its very being
and nature desires the Intellect’ from which it gets its being.27

lost work of John Philoponus’, Israel Oriental Studies 2, 1972, 320–52, at 323–4 (reprinted in his
Collected Works, vol. 2, Jerusalem and Leiden 1986), and into French by G. Troupeau, ‘Un épitomé
arabe du “de contingentia mundi” de Jean Philopon’ in Memorial A.J. Festugière = Cahiers
d’Orientalisme 10, Geneva 1984, 77–88, at 84; and in a work recorded by Simplicius in Phys.
1327,14–16; 1329,17–19. In his earlier in Cat. 50,28–51,12, Philoponus was content merely to
expound Aristotle’s version of the infinite power argument.
20
Philoponus contra Proclum 238,3–240,9; 297,21–300,2; 626,1–627,20 quotes Proclus’ Examination
of Aristotle’s Objections to Plato’s Timaeus. The same argument is ascribed to Plato by Philoponus
contra Proclum 235,4–19 and propounded without ascription, Philoponus contra Proclum 2,11–14;
Arabic summary translated Pines 324, Troupeau 84.
21
Alexander Quaest. 1.18,30,25–32,19; and fragment of the lost in Cael. preserved at Simplicius in
Cael. 358,27–360,3, both translated by R.W. Sharples, in ‘Alexander of Aphrodisias: problems about
possibility 2’, Bulletin of the Institute of Classical Studies 30, 1983, at 99–102, and discussed by him
there and in ‘The unmoved mover and the motion of the heavens in Alexander of Aphrodisias’,
Apeiron 17,1983,62–6.
22
Philoponus contra Proclum 119,14–120,14; cf. 549,7–550,24.
23
ibid. 225,14–226,19; 240,23–6; 241,26–7; 242,11–22; 304,4–9.
24
Simplicius in Cael. 143,17–29; in Phys. 1331,30–3; cf. 1358,29–30. Same point made against
Alexander in Cael. 361,10–15, cf. 360,23–4.
25
Proclus Examination of Aristotle’s Objections to Plato’s Timaeus, ap. Philoponum contra Proclum
627,17–18; cf. 239,1. I am grateful to Andrew Smith for the Proclus references.
26
Proclus in Tim. 3.212,21–2 (Diehl).
27
Proclus in Tim. 1.267,10, just before the passage translated above.
 Infinite power impressed 201

In the fourth stage of his argument, Philoponus tries to go further. He does

not deny Proclus’ claim that God might override the world’s natural perishability,
although he does not agree that God will in fact do so. But he seems to think
it impossible even for God to override the world’s natural generability and give
it a beginningless existence after all. Hence he ascribes to Plato, and himself
upholds, the view that the world is generable not just by nature, but without
qualification, and that it was actually generated and had a beginning in time.28
He does not, however, explain the asymmetry which allows God to override
natural perishability and impose endless existence, but not to override natural
generability and impose beginningless existence.
The four stages of Philoponus’ argument stand out particularly clearly in the
Arabic summary of a lost work of his which has been translated by Shlomo
Pines. I will mark the stages (i) to (iv):

Then he said: (i) if the world is a finite body, as has been demonstrated by
Aristotle in the first treatise of his book on the Heaven and (if) the forces of
every finite body are finite, as has been likewise demonstrated by Aristotle at the
end of the eighth treatise of the Book of Physics, (then) because of what we have,
and of what Aristotle has demonstrated, the world must have been created in
time (and) have come into existence after not having existed.
Supposing, however, that someone says: (ii) as Aristotle has explained and
demonstrated, the body of the world is finite and it is impossible that a finite
body should have an infinite force; however the force which has ensured the
preservation of the essence of the world in the eternity a parte ante [and] which
will ensure its permanent and perpetual preservation in the future, is the force of
the Creator, may He be blessed, who causes the heaven to move in perpetual
motion; we should answer:
(iii) The disagreement between us and him does not (concern) the (portion)
of the Creator, may He be blessed and exalted, which is imparted to the world.
The disagreement solely (concerns) your saying that the world is by nature
eternal a parte ante. For if, as you say, it is eternal a parte ante, it does not need a
force ensuring the preservation of its essence to be imparted to it. For it is a
characteristic of a thing which is eternal a parte ante by nature that the force
which ensures the preservation of its essence should be a force natural to it and
not drawn by it from something else. [If however this force is drawn by it from
something else] and received by it from a thing other than its essence (the thing
in question) is not eternal a parte ante by nature, for it is not then . . . It is (106b)
eternal a parte ante only because of a force (belonging) to some other (thing).

28
Philoponus contra Proclum 237,7–15.
202 Aristotle Transformed

(iv) Its eternity a parte ante would have been abolished if there had not been
this thing which gives it permanent (being) and ensures the preservation of its
essence preventing the latter from perishing. Since the world is not by nature
eternal a parte ante, as is asserted by the Eternalists, it must be by nature created
in time, as is asserted by us. And it is impossible for what is by nature created in
time to be eternal a parte ante. In this there is a refutation of the doctrine
concerning the eternity a parte ante of the world adopted by the Eternalists.
The main points of the first treatise of the discourse of John the Grammarian
are finished. Glory be to God always, for ever, eternally.29

Philoponus’ argument was available in Arabic not only in the summary just
presented, but also thanks to the translation of his contra Proclum and contra
Aristotelem. The detective work of Islamic scholars has shown that various stages
of the argument were known to the Jewish philosopher Saadia (died 942 ad ), to
Farabi’s Christian pupil Yah.ya Ibn cAdî (892–1073) and in turn to his Christian
pupil Ibn Suwâr (born 942). But what has not to my knowledge been discussed,
nor translated, is a portion of commentary by Avicenna (c. 980–1038) on Aristotle’s
version of the infinite power argument, as it appears in his Metaphysics. I am
grateful to Fritz Zimmermann for supplying me with an abstract. I believe that
Avicenna here does four things. First, he endorses, without ascription, the view of
Ammonius that Aristotle’s God is the Sustainer, as well as the Mover, of the
heavens (sec. 6). Secondly, he distinguishes the type of argument which we found
in Syrianus which moves directly from God’s preservation of the heavens’ motion
to his preservation of their existence (2b). Thirdly, he reapplies the infinite power
argument, to make Aristotle’s God the Sustainer of the heavens, very much in the
way we saw Proclus doing, and Philoponus recording in the second stage of his
argument (2d). Finally, he adopts a position like that which Philoponus forces on
Proclus in the third stage of his argument: the existence of the heavens is necessary
(they are imperishable) because of God, not because of their own nature (3,4).
Averroes later describes Avicenna as following Philoponus’ point here – although
he misidentifies it as Alexander’s – and it is he (Averroes) who puts Avicenna’s
point, as does Abravanel after him, in terms of God versus nature.30

29
‘Philoponus, an Arabic summary of a lost work’, translated by S. Pines, Israel Oriental Studies 2,
1972, at 323–4, repr. in his Collected Works 2, 1986; French translation by Troupeau 84–5.
30
Averroes Middle Commentary on Aristotle’s de Caelo, translated from Hebrew version into Latin by
Paul Israelita, available as ‘paraphrasis’ in the Juntine edition of 1562–1574, vol. 5, 293 v, a, H-I,
partly translated into English from the Hebrew by H.A. Wolfson, Crescas’ Critique of Aristotle,
Cambridge Mass. 1929, 681–2. Avicenna is here described as the companion (comes in Latin) of
‘Alexander’, and the terminology of nature is associated with Avicenna here and later by Abravanel
The Deeds of God 2.3, 126, translated from the Hebrew by H.A. Wolfson, op. cit., 597 and 682.
 Infinite power impressed 203

Avicenna’s argument is hard to follow, because the commentary is probably a

reconstruction, whether by himself or by someone else, of part of a larger
production in twenty volumes, which he lost in the sack of Isfahan three years
before his death. It may be significant that in announcing his hope of
reconstructing the work, he mentions that he is studying not only Alexander and
Themistius, but also Philoponus.31
I need to show in more detail how the line of thought I have described
emerges from the passage. Avicenna is attacking (sec. 6) commentators on
Aristotle, especially (sec. 4) the Nestorian Christian Abu Bishr Matta of Baghdad
(died c. 940), who had translated Aristotle, Alexander and Themistius into
Arabic. Abu Bishr had said (2a) that the existence of the celestial sphere was
necessary in itself, so that (4) God’s role was confined to producing motion and
to making motion perpetual. Avicenna objects (3,4) that nothing other than God
is intrinsically necessary; all other necessity is owed to God, including that of the
celestial sphere. This is the view that Averroes later connects with the third stage
of Philoponus’ argument, which he misidentifies as Alexander’s. Avicenna goes
on (6) to accept the interpretation of Aristotle provided by Ammonius and his
followers including Philoponus, according to which Aristotle’s God is the
Sustainer of the universe. On the other hand he thinks (1) that neither the
commentators he is attacking nor Aristotle himself supply an actual argument
for this conclusion. The only argument they supply for God concerns the motion,
not the existence, of the universe. In particular, they do not use the idea which we
found in Theophrastus, Syrianus and possibly Simplicius (2b). Avicenna then
refers to an argument which would do the trick (2d), and would show God to be
the Sustainer of the universe. It is the infinite power argument used by Proclus
and recorded inter alios by Philoponus. According to the misguided interpretation
of certain commentators, the celestial sphere will be receptive (a concept found
also in the Greek texts) of the motion imposed by God, and even perpetually
receptive of it, despite the absurdity that no provision has been made for its
existence to be perpetual. For its own power is finite, so that it has no internal
source of perpetual existence. Its perpetual existence cannot then be assured,
unless these commentators are wrong, and the divine source of its motion, which
acts as a final cause filling it with desire, also acts as an efficient cause, that is, I
take it, as an efficient cause of its existence. (‘It will have to be receptive and

31
Avicenna, Letter to al-Kiyâ, M. Badawi (ed.), Aristû ‘ind al-‘Arab, Cairo 1947 (repr. Kuwait 1978),
120–2, second part translated into French by S. Pines, ‘La “philosophie orientale” d’Avicenne et sa
polémique contre les Bagdadiens’, Archives d’histoire doctrinale et littéraire du moyen âge 19, 1952,
5–37, at 6–9. See esp. 9.
204 Aristotle Transformed

perpetually in motion despite the fact that its power is finite – unless there is
another cause and emanation on account of which the desired end becomes an
effective principle’.) The following are the relevant sections of Avicenna, with key
portions italicised:

Comments on Metaph. 12.6, 1071b5–31

(1) He [Avicenna] criticises Aristotle and his commentators as follows. It is

improper to argue to God from the fact of motion. For in that way they cannot
show that the true One is the principle of all existence. If the first principle is a
principle of the motion of the sphere, it need not therefore be a principle of the
substance of the sphere.

(2) He [Avicenna] says: Let us examine their belief that the motion of the sphere
is necessary and without beginning or end. They do not show:

(a) that the existence of the sphere is necessary in itself;

(b) that for it to exist is to be in motion, so that if it ceased to move it would
cease to exist.
(c) Rather, they argue as follows: the sphere exists; it is in motion; therefore its
motion must be beginningless. But that is a non sequitur. For suppose we knew
it existed but not whether it had ever moved. From the fact of its existence we
could not tell whether it moved, perpetually or otherwise. Neither could we
proceed to tell that or how it, its matter and form, had been brought into being
[by a creator].
(d) In a word, how is it supposed to be desirous, receptive, eternal? It will have
to be receptive and perpetually in motion despite the fact that its power is finite –
unless there is another cause and emanation on account of which the desired end
becomes an effective principle.

Comments on Metaph. 12.7, 1072b3

(3) If it sets things in motion by dint of being intrinsically desirable, and because
the thing moved is fit so to be affected as to get moving, that motion is brought
about jointly by conditions in both mover and moved. If so, act and effect follow
necessarily, if only in faculties close to reason (Greek: logos). Act and effect will
thus be necessary, and will constitute a sublime necessity which has a noble
existence, since the order of the whole universe flows from it. We do not mean to
say that that necessity is one of compulsion or ineluctability. It is one where
things could not conceivably be different. This passage (1072b3) does not mean
that the heavenly motion is intrinsically necessary, in the sense that it could not
be otherwise. It is necessary in the sense explained earlier. Nothing exists of
 Infinite power impressed 205

necessity when considered in isolation from the First. Indeed, if the tie to the
First could be broken, everything would reduce to nothing. In respect of itself,
everything is vain and perishes, except the face of the First Truth.

(4) Some, failing to distinguish between ‘conditionally necessary’ and ‘strictly

necessary,’ have mistaken the necessity here at issue for an intrinsic necessity,
arguing as follows, (lacuna). Now I ask Abu Bishr: on the assumption that it is of
necessity that things are as they are, what part is there for the first cause to play in
relation to them? He answers: to maintain them in perpetual motion. But that is
absurd. The part played by the First is that it is on its account that there is necessity,
with no necessity accruing to anything else on account of itself. A sign of the folly
of that (Baghdadian) position is that it makes necessity intrinsic to things,
perpetuity extrinsic: that intrinsic necessity of theirs is not supposed to entail
perpetuity; perpetuity is to be furnished from outside. But if perpetuity is
bestowed on motion, does that not also make it necessary? If it were necessary
by itself, would not that make it independent of its mover?

(5) The truth is that motion, its existence, the necessity of its existence while it
exists, and the perpetuity of its existence, all depend on the causes of motion.
God is too lofty to be made a cause only of motion. It is He that furnishes the
existence of every substance. It is through Him that everything else must be and
acquires the necessity consequent upon the relation there must be between Him
and it.

Comments on Metaph. 12.7, 1072b13–14

(6) ‘Thus upon such a principle does the heaven depend’. The principle in question
is primary, single, simple, intrinsically intelligible, absolutely good, the object of
desire for the whole universe. It is from it that necessity emanates to things.
Besides it, no importance can be attached to non-existence [as a possible source
of existence]. It either excludes non-existence altogether or extends a power to
exist to things in the measure in which they are fit to be affected. ‘Depends’:
Aristotle means to say that the universe itself is sustained by that principle. It is
wrong to confine its blessings to the lowly gift of motion, as do those commentators.32

That the infinite power arguments recur in Averroes (c. 1126–1198) has been
well brought out in some very interesting work by modern Arabists. What I
think can now be clarified is the relationship of Averroes’ texts to the preceding
ones of Avicenna, Philoponus and others. Perhaps the most important point is
that the first three stages of Philoponus’ argument recur in Averroes, but modern

32
Avicenna, Notes on Metaphysics 12 from his Kitâb al-Ins.af, Badawi (ed.), Aristû ‘ind al-‘Arab, 23–6
(whole text 22–33), abstract by Fritz Zimmermann.
206 Aristotle Transformed

scholarship has not been aware of this, because Averroes was not aware himself.
He correctly identifies the first stage as coming from Philoponus, but by a signal
error ascribes the second and third stage to Alexander, and his ascription is
taken at face value by Thomas Aquinas and by more recent commentators.33
It is clear that the second-stage argument for God as Sustainer, which
Philoponus takes from Proclus, should not be ascribed to Alexander. For
Alexander argues on the other side that, if the world escapes destruction, this
cannot be due to an external cause, such as God.34 And a passage of Simplicius
informs us that he goes only as far as making his Aristotelian God an efficient, as
well as a final, cause of the world’s motion, not a cause in any way of its existence.35
It is equally wrong to ascribe to Alexander the idea from the third stage of
Philoponus’ argument, that the world is destructible by nature, but preserved by
God. That is another thing that Alexander argues to be impossible.36
Averroes addresses the subject in five of his commentaries on Aristotle and in
his de Substantia Orbis. The three most informative commentaries are the Long
Commentary on Physics 8, the Middle Commentary on de Caelo and the Long
Commentary on Metaphysics 12.37 I will mark separately the first three stages of
Philoponus’ argument, as it appears in some of Averroes’ fullest formulations, along
with a relevant part of Averroes’ reply, and his assimilation of Avicenna to Philoponus
(misiden-tified as Alexander). But I shall slightly alter Genequand’s translation of
the Long Commentary on Metaphysics 12, 1628, to distinguish Philoponus’ argument
from Averroes’ reply by making the quotation from Philoponus end earlier.

33
H.A. Wolfson, The Philosophy of the Kalam, Cambridge, Mass. 1976, 381; H.A. Davidson, ‘John
Philoponus as a source of medieval Islamic and Jewish proofs of creation’, Journal of the American
Oriental Society 89, 1969, 361, n. 41; S. Feldman, ‘The end of the universe in medieval Jewish
philosophy’, Association for Jewish Studies Review 1986, 76.
34
ap. Simplicium in Cael. 301,3; 359,14.
35
Simplicius in Phys. 1361,31–3; 1362,11–15. On this particular point I diverge from C. Genequand’s
valuable discussion Ibn Rushd’s Metaphysics, Leiden 1984, 36.
36
Alexander Quaest. 1.18,30,25–32,19; fragment of lost in Cael. ap. Simplicium in Cael. 358,27–360,3.
37
(i) Averroes Long Commentary on Physics 8, comm. 79 (available in the thirteenth-century Latin
translation by Michael Scot, in the Juntine edition of 1562–1574, vol. 4, reprinted Frankfurt 1962, 425
verso, col. a, sec H – 427 recto, a, B); (ii) Middle Commentary on de Caelo translated from Hebrew
version into Latin by Paul Israelita (available as ‘paraphrasis’, Juntine edition, vol. 5, 293, v. a, G – 293 v.
b, K; 294 r. b, D – 295 r. a, B), also translated in part directly from the Hebrew version by H.A. Wolfson,
Crescas’ Critique of Aristotle 596–7; 681–2; (iii) Long Commentary on Metaphysics 12, comm. to text
41, 1626–38, translated from Arabic by C. Genequand, op. cit., 162–8. Below I use the the translations
from Hebrew and Arabic by Wolfson and Genequand and otherwise translate from the Latin. Other
works are: (iv) Long Commentary on de Caelo 2, comm. 71, Latin translation by Michael Scot, in
Juntine Edition, vol. 5; (v) de Substantia Orbis 3; and (vi) a work inaccessible to me, the Hebrew MS
version of the Middle Commentary on Metaphysics 12 (Casanatense Heb. MS no. 3083, 140(141)b –
141(142)a). This reference is given, along with the others, by H.A. Davidson ‘The principle that a finite
body can contain only finite power’, in S. Stein and R. Loewe (eds), Studies in Jewish Religious and
Intellectual History Presented to A. Altmann, Alabama 1979, 75–92, at notes 53, 58, 61, 64.
 Infinite power impressed 207

Averroes’ Long Commentary on Metaphysics 12, 1628, translated from Arabic by

Genequand.

Stage (i): John the Grammarian [Philoponus] raised strong objections against
the Peripatetics concerning this problem. He says: ‘if every body has a finite
power and the heaven is a body, then it will have a finite power; but everything
finite is necessarily corruptible, so that the heaven is corruptible.’

Stage (ii): If it is said that it acquires incorruptibility from the eternal separate
power,

Averroes’ reply: there will be something destructible but eternal. But this has
been shown to be impossible at the end of the first book of the de Caelo et Mundo.

Averroes’ Long Commentary on Physics 8, 79, Juntine vol. 4, 426v, a, I–L.

Now in the proposition here assumed, according to which the power of any body
is finite, it can be doubted whether it applies to the celestial body or not. If it
does, then the power of the celestial body will be finite, but what has finite power
is destructible. . . . This first question at least is very difficult and full of snags.

Stage (ii): And Alexander answers in some of his treatises, and says that the
celestial body receives (adeptus fuisse) eternity from its mover, who is not
enmattered.

Averroes’ reply: But this would make it something that can be destroyed, yet
never will be destroyed. This is Plato’s opinion too, that is, that there is something
eternal which can be destroyed. But Aristotle proved at the end of the first book
of the de Caelo et Mundo [1.12] that there could not be anything eternal that
contained a potentiality for destruction.

Stage (i): Now John the Grammarian [Philoponus] maintained his own opinion
[reading opinionem with the Latin of the 1489 edition] against the Peripatetics,
in that he thinks that the world is destructible and generable. And this is the
strongest of all the doubts that can befall us on this, especially because Aristotle
says expressly in the second book of the de Caelo et Mundo [2.12, 293, a, 10–11]
that the power of the heaven is finite.

Averroes’ Middle Commentary on de Caelo 1, translated from Hebrew version

by Wolfson (Latin translation from Hebrew in Juntine vol. 5, 293v, a, G and
294v, b, M).

Stage (iii): The inference must therefore be that while the sphere by its own
nature has the possibility of being corrupted, it must be free of corruption on
account of the infinite immaterial force, outside the sphere, which causes its
208 Aristotle Transformed

motion. That this is so is maintained by Alexander in a treatise of his, and he is

followed by Avicenna. . . .

Averroes’ reply: On this account, i.e. by virtue of its being simple, the celestial
body has no substratum and no contrary. Hence Aristotle maintains that it is
ungenerated and incorruptible.

Averroes’ reply reveals his relation to a number of further thinkers. It includes

a point which modern scholars have understood as coming from Philoponus,
and indeed as forming the nub of Philoponus’ argument, but in fact I believe that
he does not use it here.38 The point is that the pseudo-Alexandrian view violates
Aristotle’s conclusion in Cael. 1.12, that an eternal thing cannot have a potentiality
for destruction.39 Averroes concludes that the view must be rejected. He prefers
to return to the idea which we found in Abu Bishr, that the heavens’ existence is
necessary of itself. One reason why the celestial body endures by its own essence
is that Averroes thinks it is not a composite of matter and form.40 Another reason
is that it lacks a contrary, and so has no potentiality for destruction. Motion
differs, according to Averroes, because it does have a contrary, rest, and so it gets
its perpetual motion, unlike its necessary existence, from elsewhere.41 In thus
separating motion from existence, Averroes distances himself from the type of
argument found in Syrianus, according to which preserving the celestial body’s
motion is the same thing as preserving its existence. For this departure he was
later rebuked by Abravanel.42 Had he but known it, Averroes should have claimed
Alexander as an ally, not an opponent. For the true Alexander would have agreed
that the heaven’s existence is necessary of itself.
Averroes’ treatment of the subject is important because of its extensive
influence on subsequent Christian and Jewish discussions. Of the three
commentaries of Averroes from which I have quoted, two were translated into

38
I am persuaded here by Lindsay Judson, op. cit. I do not think that Philoponus is represented as
using the point at Simplicius in Phys. 1333,28–30; 1334,37–9. This passage rather concedes for the
sake of argument the position of Plato and Proclus that the potentiality for destruction will remain
for ever unactualised. The contrary interpretation is that of Wolfson (1943 at 202 and 240; 1976 at
377–8 and 381) and Davidson (1969 at 362 and 1979 at 80 and n. 37) cited above. But Davidson
expresses some misgivings as to how it fits, in ‘The principle that a finite body can contain only finite
power’ (1979) at n. 37. Averroes’ type of point is to be found elsewhere in Philoponus, but in a
different context, at contra Proclum 131,26–132,28.
39
Averroes Long in Phys. 8, 426v, b, K; Middle in Cael. 293v, a, I; 294r, a, F; Long in Metaph. 12, 1628.
40
Averroes Middle in Cael. 294r, b, D – 295r, a, B. Its non-compositeness does not seem to be used for
the same purpose at Long in Phys. 8, 426v, b, M – 427r, a, B, nor (pace Genequand, 46) at Long in
Metaph. 12, 1634.
41
Averroes Long in Metaph. 12, 1631; Middle in Cael. 294v, b, M.
42
Isaac Abravanel, The Deeds of God 9.9, cited by S. Feldman, ‘The end of the universe in medieval
Jewish philosophy’, Association for Jewish Studies Review 1986, at 76.
 Infinite power impressed 209

Latin from Arabic in the thirteenth century by Michael Scot, namely the long
commentaries on the Physics and Metaphysics. Consequently, Averroes’ discussion
was known to such thirteenth-century scholars as Robert Grosseteste and Roger
Bacon in Oxford and Bonaventure and Thomas Aquinas in Paris, to name them
in rough chronological order.43 Grosseteste, surprisingly, takes a view like that of
Averroes.44 Bacon sides with what I have called Philoponus’ opening move,45 while
Bonaventure reverses Philoponus’ view. According to him, God does indeed give
the world a beginning, but in doing so overrides, rather than follows, its nature.46
The most thoughtful discussion is that of Thomas Aquinas.47 He refers to
Averroes’ long commentaries on Physics and Metaphysics 12 and to his de
Substantia Orbis, and he reports the controversy in Averroes’ own terms as being
between Averroes and Alexander. The belief that Alexander made God a Sustainer
eternally overriding the world’s natural tendency to finite duration takes on a
new significance in Thomas. For it will have confirmed him in his view that this
was also the opinion of Alexander’s master, Aristotle.
Thomas does not side with Philoponus’ opening move, because unlike
Philoponus, he accepts that, though the world had a beginning, it will have no
end.48 He therefore favours the view, wrongly ascribed to Alexander, that the
celestial body receives its eternity not from any infinite power of its own, but
from the infinite power of the divine mover. He does so, even though the so-
called ‘Alexander’ goes beyond him, by treating the celestial body as not only
endless, but also beginningless in time. He defends ‘Alexander’ from Averroes’
reply by insisting that the celestial body does not have a potentiality for non-
being, but is indestructible. So Alexander is immune to the objection that he
violates the ban in Aristotle Cael. 1.12 on an eternal thing being destructible.49

43
Richard Dales tells me he finds arguments against infinite power also in the relatively early Vatican
MS Latin 185.
44
Robert Grosseteste de Motu Supercaelestium, in L. Baur (ed.), Die philosophischen Werke Grossetestes,
96, 29–99, 4.
45
Roger Bacon Opus Maius, part 7, Moral Philosophy, part 4, ch. 1 (in Opus Maius, J.H. Bridges (ed.),
vol. 2, Oxford 1897).
46
Bonaventure, Commentary on Lombard’s Sentences 2, d.1, par. 1, a.1, q. 2, conclusion (translated by
Paul Byrne, in Cyril Vollert, Lottie Kendzierski, Paul Byrne, St. Thomas Aquinas, Siger of Brabant,
St. Bonaventure, On the Eternity of the World, Milwaukee, Wisconsin 1964, 105–13, at 109–10); and
in Hexaemeron Collatio 6, n. 5 and Collatio 7, n. 2.
47
Thomas Aquinas, Commentary on Aristotle’s Physics 8, lectio 21, 1147, 1152, 1154. There is an English
translation of the Commentary by Richard J. Blackwell, Richard J. Spath and W. Edmund Thirlkel,
London 1963.
48
Thomas Aquinas, Commentary on Aristotle’s Physics 8, lectio 21, 1147 and Summa Theologiae 1,104,
a.4, ad 1 and 2.
49
See the excellent account in Carlos Steel, ‘ “Omnis corporis potentia est finita” ’, in Jan P. Beckmann,
Ludger Honnefelder, Gangolf Schrimpf, Georg Wieland (eds), Philosophie im Mittelalter, Hamburg
1987, 213–24.
210 Aristotle Transformed

This is not to deny that the celestial body receives its perpetual existence from
God, since necessary things can perfectly well have a cause of their necessity. It
has been found puzzling that the celestial body can be necessary both by its own
nature and because of God.50 But this is very close to the position that we found
in Proclus, who argued that it gets its nature from God, and in Simplicius, who
maintained that it is by nature fitted (epitêdeios) to receive the indestructibility
God gives it:

Even with his intention granted, however, Averroes cannot draw a conclusion
against Alexander. For the latter did not posit that the celestial body acquires
eternity from another on the understanding that of itself (de se) it has potentiality
for being and non-being, but on the understanding that its being does not come
from itself (ex se) . . . Now the only being it has from him is perpetual being. So
it has its perpetuity from another. And Aristotle’s words are also consonant with
this, when he says in Metaphysics 5 and above at the beginning of this eighth
book that some necessary things have a cause of their necessity.51

Elsewhere Thomas tries out other arguments.52 He also rules out the type of
move found in Syrianus, according to which the heavens would have a potentiality
for acquiring existence from their motion.53
Averroes’ influence did not stop with Thomas, but extended to the Jewish
thinkers Gersonides (1288–1344), Crescas (1340–1410) and Abravanel (1437–
1509), the first two of whom attacked the foundations of the infinite power
arguments, with results of considerable interest for dynamics.54 It is with some of
the implications of the infinite power arguments for dynamics that I want to
finish.

50
Carlos Steel, op. cit.
51
Thomas Aquinas Commentary on Aristotle’s Physics 8, lectio 21, 1154.
52
Thomas denies that the heavens have a finite potentiality for being, on the grounds that their only
potentiality is for motion, not for being (Commentary on Lombard’s Sentences 2, d.1, a.1, q.5, ad
contra 8). But that solution clashes with Physics Commentary 1153 and Summa contra Gentiles 1.20,
obj.3, which allow a potentiality for being. Consequently, the latter passage offers a different
solution. The heaven has the power of changeless being, but that need not be an infinite power,
because changeless being has no quantitative extension and is not touched by time, whether the
heaven endures for an instant, or for infinite time.
53
Thomas Aquinas, Commentary on Lombard’s Sentences 2, d.1, a.1, q. 5, ad contra 8.
54
On Crescas, see The Light of the Lord, translated in part by H.A. Wolfson, Crescas’ Critique of
Aristotle, Cambridge Mass. 1929, proposition 12, part 2, 271, with Wolfson’s notes; and H.A.
Davidson, ‘The principle that a finite body can contain only finite power’, in S. Stein and R. Loewe
(eds), Studies in Jewish Religious and Intellectual History Presented to A. Altmann, Alabama 1979,
75–92, at 85–9. For Gersonides Ways of the Lord 5.3.6; 6.1.3; 6.1.14, see the account of C. Touati, La
pensée philosophique et théologique de Gersonide, Paris 1973, 308–15; and for Abravanel Deeds of
God 9.9, see S. Feldman, ‘The end of the universe in Jewish medieval philosophy’, Association for
Jewish Studies Review 1986, 53–77, at 74–6.
 Infinite power impressed 211

I believe there are several implications. First, I have argued elsewhere that
Philoponus must be regarded as the originator of impetus theory, if an impetus
is viewed not merely as an internal force, but as an internal force impressed from
without (vis impressa).55 But this claim must now be qualified, for we have seen
that it is an impressed force (dunamis) which Proclus, before Philoponus,
imagines God as implanting in the heavens.56 I doubt, however, if the qualification
is very significant. For one thing, Proclus does not extend the idea of impressed
force to the most interesting case of projectiles. For another, he seems not to have
influenced Philoponus in this regard, even though Philoponus quoted some
of the relevant passages cited above from Proclus. For Philoponus started by
applying the idea of impressed force to projectiles. He did so in his commentary
on Aristotle’s Physics,57 of which the first edition can be dated to 517. It was only
some thirty or forty years later in the de Opificio Mundi of 546–9, or even 557–
60, that he first thought of making God implant an impetus in the heavens.58 And
even then he was not following Proclus, because Proclus’ argument for God’s
impressing a force depended on a belief which Philoponus rejected, the belief in
the infinite duration of the heavens.
The infinite power arguments carry a second implication for dynamics. Why
should an infinite force be needed to keep the celestial spheres rotating for ever?
In his attack on motion in a vacuum, Aristotle includes two pertinent difficulties.
If there were motion in a vacuum would not the speed be infinite in the absence
of resistance,59 and where would the motion ever stop, if no stopping place in a
vacuum differs from any other?60 Avicenna endorses the second objection in his
own person, and argues in detail that projectile motion in a vacuum would not
stop.61 Is there not, then, we may ask, a worry about the rotation of the celestial
spheres? Although Aristotle surrounds them not with a vacuum, but with
nothing at all, the question retains its force: why should they stop at one position
or time rather than another? And this in turn creates a clash with the demand for

55
Richard Sorabji, ‘John Philoponus’, in Sorabji (ed.), Philoponus and the Rejection of Aristotelian
Science, London and Ithaca NY 1987.
56
There is no reference to impressed force in the much earlier pseudo-Aristotelian de Mundo 398b20,
where God is said to transmit the power (didôsi dunamin) which moves the inner celestial spheres
from the outermost one. For the simile of endosis used here is not that of impressing a force, but of
striking a key note.
57
Philoponus in Phys. 641,13–642,20; cf. 384,29–385,11.
58
Philoponus Opif. 28,20–29,9.
59
Aristotle Phys. 4.8,215b22–216a4.
60
ibid. 215a19–20.
61
Avicenna, Shifâ 1, p. 154f, translated into French and discussed by S. Pines, ‘Études sur Awh.ad al-
Zamân Abu’ l-Barakât al-Baghdâdî’, Revue des Études Juives n.s. 3, 1938, repr. in his Collected Works
1, Leiden 1979, 52–6.
212 Aristotle Transformed

infinite power. For no infinite power is needed to spin a finite body, and once
it is spinning without differentiation in its surroundings, why, if Aristotle’s
argument is acceptable, should it stop? Conversely, if we take seriously the
demand for infinite power, we presuppose that a finite power would be exhausted.
But then we can answer Aristotle’s question why something should stop after
three miles, rather than four miles, of travel in a vacuum.
An analogous difficulty was noticed by Philoponus about Aristotle’s other
suggestion, that speed in a vacuum would be infinite, not because of the lack of
differentiation but because of the lack of resistance. Rotation, he pointed out,
and in particular the rotation of the heavens, encounters no resistance, but is
none the less finite in speed.62 The point is repeated by Ibn Bajja or Avempace
(died 1138).63
Averroes and Thomas Aquinas both try to rescue Aristotle, Averroes by
arguing that it is the heavenly spheres themselves, and not any surrounding
medium, which provide the resistance to motion.64 Thomas suggests that
Aristotle’s connexion of lack of resistance with infinite speed is a merely
dialectical move against opponents who postulate vacuum as the sole cause of
motion,65 but I have to say that I see no sign that Aristotle’s opponents are as
extreme as that.
One way or another, it looks as if something in Aristotle’s dynamical theory
must give. Or does he have a way out of the difficulties? As regards the lack of
differentiation, someone might plead that we will find differences, if we look
beneath the heavens, rather than beyond them. At some stage in the celestial
rotation, the most prominent heavenly body, the sun, will be directly over the
highest mountain. So it cannot be said that there are no differences between
different possible stopping points for the celestial rotation. But it may still be
wondered if these differences would provide a good reason for the celestial
rotation to stop at one point rather than another. As regards the lack of resistance
to celestial motion, an attempt might be made to find sources of friction
by pointing out that some of the celestial spheres draw round with them

62
Philoponus in Phys. 690,34–691,5.
63
Avempace is reported by Averroes, Long Commentary on Physics 4, text 71, Latin in Juntine edition
1562, repr. Frankfurt 1912, 160r, b, E-F, translated in E.A. Moody, ‘Galileo and Avempace, the
dynamics of the leaning tower experiment’, Journal of the History of Ideas 12, 1951, at 226–7.
Thomas Aquinas gives the argument in his Commentary on the Physics 4, lectio 12,534, translated by
R.J. Blackwell, R.J. Spath, W.E. Thirlkel, London 1963.
64
Averroes, Long Commentary on Physics 4, text 71; Thomas Aquinas gives the argument in his
Commentary on the Physics 4, lectio 12, 535.
65
Thomas Aquinas, Commentary on Physics 4, lectio 12, 536. On the other hand, Thomas does not
discount as dialectical Aristotle’s ‘why stop here?’ question, lectio 11, 526.
 Infinite power impressed 213

(sumperiagein) the lower belts of fire and air, creating heat by inducing friction
there (parektribesthai).66 But whether they suffer friction in return is not said.
There is another reason for wondering why infinite power should be needed
for everlasting celestial rotation.67 In an earlier treatment of celestial motion in
On the Heavens, we find no integral and undisputed reference to the role of God.
Instead, Aristotle combines two ideas, that the heavens are made of an
indestructible fifth element which can undergo no change but motion, and that
circular motion is natural to that element. If it can undergo no other change,
what reason could there be for its natural motion to cease? It cannot tire, or grow
bored, or perish, for these would be changes. Nor does circular motion have a
terminal destination. It looks, then, as if in his earlier work Aristotle allowed for
a motion that was eternal not because of any infinite power, but because of
immunity to further change in something to which circular motion was natural.
There is a final corollary. For the question whether finite power would be
exhausted has implications for the definition of impetus. It has been suggested
that impetus is by definition something that gets exhausted,68 and indeed
Philoponus does say that it would be. This is his answer to Aristotle’s question
why motion in a vacuum would stop here rather than there: the force impressed
in a projectile would be exhausted (exasthenêsêi), as Aristotle himself admits in
regard to the different force which he postulates in the pockets of air behind a
projectile.69 However, it seems dangerous to elevate exhaustibility into a defining
characteristic of all impetus, since Avicenna maintains in the passage just
discussed that only resistance exhausts the force in a projectile, and that in a
vacuum the force would not be exhausted.70 Similarly Buridan insists71 that the
impetus by which God sets the heavens moving at the time of Creation will
never run out.
I shall conclude by drawing attention to the role played by the Neoplatonist
commentators on Aristotle, not only in dynamics, but also in the earlier issue of

66
Aristotle Meteor. 1.3, 341a17–22; 1.7, 344a11–13; Cael. 2.7, 289a20–35.
67
I owe the point to John Cleary.
68
Michael Wolff, ‘Philoponus and the rise of preclassical dynamics’, in Richard Sorabji (ed.),
Philoponus and the Rejection of Aristotelian Science, at 84–5. For his reply to the doubt about
Avicenna, see p. 85.
69
Philoponus in Phys. 644,16–22.
70
Avicenna Shifâ 1, p. 154f.
71
Buridan in Phys. 8.12, fol. 120, translated into French and discussed by P. Duhem, Études sur
Léonard de Vinci 3, Paris 1913, 40–2; also discussed by G. Sarton, Introduction to the History of
Science 3, part 1, Baltimore 1947, 543; M. de Gandillac, Le mouvement doctrinale du XIe au XIVe
siècles, Paris 1951, 459; C. Touati, La pensée philosophique et théologique de Gersonide, Paris 1973,
312–13; Marshal Claggett, The Science of Mechanics in the Middle Ages, Madison, Wisconsin 1959,
510–15.
214 Aristotle Transformed

Creation. What started in Ammonius as an attempt to harmonise Aristotle’s God

with Plato’s Creator ended up by helping (only helping, of course) to harmonise
Aristotle with Christianity. And what began in the Christian Philoponus as an
effort to argue for the beginning of the universe, finished in Thomas Aquinas
as an argument ascribed to Alexander against its ending. The Aristotelian
philosophy that was handed on to the Middle Ages was not just that of Aristotle,
but a far richer compilation transformed by the preoccupations of the intervening
commentators.72

72
I am very grateful for help to David Barlow, John Cleary, John Dillon, Jill Kraye, Ian Mueller, Tom
Settle, Robert Sharples, Andrew Smith and above all Fritz Zimmermann. This chapter is a revised
version of a paper originally written for the Charles Schmitt Memorial Symposium held at the
Warburg Institute in 1987. The original version was published in Sarah Hutton and John Henry
(eds), New Perspectives in Renaissance Thought: Essays in the History of Science, Education and
Philosophy: Studies in Intellectual History in Memory of Charles B. Schmitt, London 1989, and
presented in a more detailed version in my Matter, Space and Motion, London and Ithaca NY 1988.
 10

The metaphysics of Ammonius

son of Hermeias
Koenraad Verrycken

It is no overstatement to say that the latest period of Alexandrian philosophy

(starting from Ammonius) is virgin territory for the greatest part. Up to now
students of ancient philosophy seem to have had rather strong feelings of
disinterest, discouragement and even aversion to all those cheerless volumes of
Greek. As a result, an area of considerable importance has remained largely
unknown for too long. This situation has allowed several rather serious
misconceptions on the subject of the Ammonian school to persist. In this
chapter I shall try to prove that the prevailing view of Ammonius’ philosophy
is incorrect, and that a better understanding of its true nature may
contribute towards a more adequate definition of the ‘problem of Alexandrian
Neoplatonism’.
In what follows, I shall refer to the three levels (or hypostases) of orthodox
Neoplatonic theology: (1) the One, the Good, the highest or first principle, the
First, the first cause; (2) the Demiurge of Plato’s Timaeus, the Platonic Creator,
the Intellect, Mind, Nous, the second principle; (3) the world soul, the third
principle. The scheme treats each level as the efficient cause that produces the
level below, while each lower level is said to proceed by a procession from
the higher. The level which the world soul produces is the still lower level of
the sensible world. The supreme producer, however, which is directly or indirectly
responsible for all levels, and especially for the level of matter, is the One.
The Intellect is said to exist in the realm of intelligible and intellective entities.
Being and non-discursive thought are associated with it. The One, by contrast,
is above being and thought and other characteristics, so that it can be described
either not at all, or only negatively, while the world soul engages only in a
lower kind of thinking – discursive thought. The Platonic Demiurge produces
lower levels (directly or indirectly) by containing the Platonic Ideas as

215
216 Aristotle Transformed

creative logoi within itself. Aristotle’s God had usually been understood
before Ammonius differently from the Platonic Demiurge, not as an efficient
cause of existence, but only as a final cause or goal which inspires the heavens
to move.
At the beginning of this century Karl Praechter initiated the view that
Alexandrian Neoplatonism was clearly distinct from Athenian Neoplatonism.
According to Praechter there were four different types of Neoplatonic philosophy:
(1) Plotinus and Porphyry, the founders of the system. (2) The speculative
branch, which developed metaphysical thinking to the utmost. It is to be divided
into (a) the Syrian school (Iamblichus, etc.) and (b) the Athenian school (Plutarch
of Athens, Syrianus, Proclus, Damascius, Simplicius, etc.). (3) The religioustheurgic
branch, i.e. the school of Pergamon (Maximus of Ephesus, Julian the Apostate,
etc.). (4) The ‘learned’ branch: (a) the Alexandrians (Hypatia, Hierocles, Hermeias,
Ammonius, Olympiodorus, Philoponus, Elias, David, etc.) and (b) the Western
Neoplatonists (Macrobius, Calcidius, Boethius).
Simplicius’ work, according to Praechter, stands midway between the Athenian
school and the Alexandrians.1 Typical of Alexandrian philosophy, Praechter
believed, was its being influenced by two local traditions, namely the interest in
the (exact) sciences (which made it tend to turn away from metaphysics) and the
study of Christian theology.2 Christian influence, according to Praechter, was
already to be found in the philosophy of Hierocles, a pagan who was active in the
first half of the fifth century.3 Praechter was convinced that Hierocles’ metaphysics
was considerably simplified in comparison with Athenian Neoplatonism, and
that this simplification was characteristic of Alexandrian Neoplatonism, or
rather Alexandrian Platonism, in general. This philosophy was influenced, of
course, by the style of the great speculative systems (Iamblichus-Proclus). At the
same time, however, the ‘theology’ of the Alexandrians was of a more elementary
character, and their exegesis of Plato and Aristotle was comparatively sober.4
Praechter held that Hierocles regarded the Platonic Creator, the Demiurge, as

1
K. Praechter, ‘Richtungen und Schulen im Neuplatonismus’, Genethliakon C. Robert, Berlin 1910,
155–6. Praechter’s classification was meant to improve on that on which E. Zeller, Die Philosophie
der Griechen in ihrer geschichtlichen Entwicklung III 2 had based his treatment of Neoplatonism.
Zeller had classed most of the Alexandrians under ‘The Athenian School’, some of them (Hypatia,
Synesius and others) under ‘Iamblichus and the Syrian School’.
2
K. Praechter, op. cit., 151; id., ‘Christlich-neuplatonische Beziehungen’, Byzantinische Zeitschrift 21,
1912, 2–3.
3
K. Praechter, ‘Christlich-neuplatonische Beziehungen’, 3–27; id., ‘Hierokles 18’, in G. Wissowa and
W. Kroll (eds), Realencyclopädie der klassischen Altertumswissenschaft VIII , 2, 1913, col. 1482.
4
K. Praechter, ‘Hierokles’, col. 1481.
 The metaphysics of Ammonius son of Hermeias 217

the highest principle; there was no One. Accordingly, he did not hesitate to call
the system ‘theistic’,5 meaning that it accepted a single supreme God, though
with the proviso that this theism was inspired, not so much by Christianity, but
rather by pre-Plotinian Platonism.6 On the other hand, Hierocles’ Demiurge is
said to create the world from eternity out of nothing by his mere will. In
Praechter’s view this doctrine clearly has its origin in Christian theology.7
The manner in which Hierocles solves the problem of the relation between
human free will and divine providence was believed to be influenced by
Christianity likewise.8 It is important that Praechter extended the characteristics
of what he called Alexandrian Neoplatonism to Ammonius, Philoponus
and further. Ammonius, he believed, increasingly had to take into account
the presence of Christians among his hearers (e.g. Philoponus): he had to
make concessions, and he was no longer free to engage in metaphysical
speculation in the style of Proclus, because of its connexions with Hellenic
polytheism. The young Christians in the philosophical school were not
looking for the highest truth, which they possessed already in their religion.
What they were looking for was the knowledge of Plato and Aristotle.
According to Praechter, Ammonius concentrated on Aristotle, in particular
on his logical writing, the Organon, which was neutral matter. As a consequence
the Alexandrian school more and more lost its positively Platonic character
and became ‘an institute for general philosophical education’. This was exactly
what enabled it to survive after 529, when the Athenian school was closed.9
Praechter supposed a continued effect of pre-Plotinian theism on the
Alexandrian school after Hierocles. Meanwhile, this theism clearly had
the function of diminishing the distance between Platonism and Christianity.
On the other hand Praechter was convinced that the Alexandrian variant
of Neoplatonism is also to be found in Simplicius’ Commentary on Epictetus’

5
K. Praechter, ‘Christlich-neuplatonische Beziehungen’, 3–4: ‘His (i.e. Hierocles’) supreme entity is
the Demiurge. He is conceived in an absolutely theistic way as the creator, the organiser, the
legislator and the leader of the world. He is simply “god (ho theos)”. Beyond him there neither is the
One nor any other hypostasis, He is the only supra-mundane being, the “God of gods (theos theôn)”,
the “supreme and best God (theos hupatos kai aristos)”. Immediately after Him come the intra-
mundane gods (enkosmioi theoi), the astral gods.’
6
K. Praechter, ‘Christlich-neuplatonische Beziehungen’, 4–5. In his article on Simplicius Praechter
refers to Origen the pagan and Longinus (K. Praechter, ‘Simplikios’, in G. Wissowa and W. Kroll
(eds), Realencyclopädie der klassischen Altertumswissenschaft IIIA , 1, 1927, col. 208).
7
K. Praechter, ‘Christlich-neuplatonische Beziehungen’, 5–9; id., ‘Hierokles’, col. 1482.
8
id., ‘Christlich-neuplatonische Beziehungen’, 12–21.
9
id., ‘Richtungen und Schulen’, 151–4.
218 Aristotle Transformed

Enchiridion,10 which, he thought, was written at the time Simplicius was still a
pupil of Ammonius in Alexandria and not yet influenced by the Athenian
Neoplatonism of Damascius.11 The theism of Ammonius, according to Praechter,
is documented in the Commentary on the Isagoge,12 if not in the in Metaphysica
of his pupil Asclepius.13
Up to now Praechter’s picture of the philosophy of Ammonius has remained
largely unchallenged. Eminent scholars like P. Merlan and L.G. Westerink have
agreed with it. I quote Westerink first: ‘. . . we are very incompletely informed
about his (Ammonius’) metaphysics. Yet one essential point is clear: from the
testimonies . . . we learn that to Ammonius the Aristotelian Intelligence is both
the final and the efficient cause; as this Intelligence is identified with the Creative
Mind of the Timaeus, it follows that the Creator is also the Supreme God.
Accordingly, the Alexandrian School has the characteristic habit of referring to
God as“the Demiurge”, a practice continued later by the Christian Olympiodoreans.
The reason for the identification was no doubt partly to bring Plato into harmony
with Aristotle, but the ultimate motive (as already in Hierocles) was to adapt
both to Christian monotheism. As a result, the complicated metaphysical
superstructure of Athenian Platonism had to be discarded, or at least had lost its
interest . . .’14 Here we find the main reason why Praechter’s view on Ammonius is
regarded as correct to the present day: Ammonius’ interpretation of Aristotle’s
God is believed to be theistic in character. Everyone concluded that this could
not mean anything other than that Ammonius himself considered the divine
Intellect or Demiurge as the highest principle, and that he consequently rejected
the Neoplatonic distinction of a higher God, the One.

10
id., ‘Simplikios’, col. 207: ‘It is the simpler doctrine of principles of the Alexandrian school, known to
us in its more elementary character from Hierocles. . . . It is only little influenced by Plotinus and the
Neoplatonism based on Plotinus, and essentially links up with the older evolution of Platonic
doctrine. The most important point of difference is the identification of “the principle of principles
(arkhê arkhôn)” with the Demiurge who creates and governs the world. This identification implies
the absence of the One, the principle which transcends the whole intelligible world and which is
characteristic of the doctrine of Plotinus and his followers.’
11
ibid., col. 206.
12
id., ‘Christlich-neuplatonische Beziehungen’, 5 n. 5: ‘I shall only indicate here that other Alexandrian
Neoplatonists, just like Hierocles, know only one supra-mundane God. Ammonius for instance (in
Isag. 3,9ff; 9,16f; 9,21; 11,11 etc.) talks about “God (ho theos)”, his properties and activity completely
in the traditional way, without betraying any knowledge of the complicated theological system
elaborated by non-Alexandrian Neoplatonism.’
13
id., ‘Simplikios’, col. 208: ‘. . . his (i.e. Ammonius’) Commentary on the Metaphysics too, recorded by
Asclepius, in comparison with Syrianus shows a more elementary construction of the highest levels
of reality.’
14
L.G. Westerink, The Greek Commentaries on Plato’s Phaedo, vol. 1: Olympiodorus, Amsterdam,
Oxford, New York 1976, 24.
 The metaphysics of Ammonius son of Hermeias 219

Merlan adduces a further argument, which is taken from the so-called

Ammonius or de Mundi Opificio. This dialogue was written by Zacharias
Scholasticus, a Christian who attended the lectures of Ammonius for a while in
Alexandria. It makes no mention at all of any difference, in Ammonius’ philosophy,
between the highest principle of reality and the Demiurge, a fact that brought
Merlan to the following conclusion: ‘If we rely on Zacharias, we shall have to admit
that Ammonius either did not profess the doctrine of a god above the Demiurge,
i.e. Plotinus’ One, or at least gave it scant attention. As to the nature of God there
seems to be no difference at all between Ammonius and Zacharias. . . . It is this un-
Plotinian idea of God as creator and omniscient provider which Boethius shares
with Ammonius. We should not forget that according to Ammonius, Aristotle’s god
was also efficient cause – in other words, Ammonius tried to reconcile the Platonic
concept of the Demiurge with the Aristotelian concept of the Unmoved Mover.’15
The first to oppose Praechter’s theory about Alexandrian Platonism was A.C.
Lloyd. He pointed out that Simplicius’ Commentary on the Enchiridion does not
regard the Demiurge as the highest principle, and that Asclepius’ Commentary
on the Metaphysics does not confirm Praechter’s view either.16 Unfortunately,
Lloyd did not specify his own opinion about the nature of Ammonius’
metaphysics. But his criticism of Praechter was taken up and extended by I.
Hadot. On the basis of a thorough study of Hierocles’ philosophy and Simplicius’
in Enchiridion Epicteti, Mme. Hadot concluded that Praechter’s position17 was
untenable. As to Hierocles first, she showed convincingly that his views on
matter, the Demiurge, soul and providence represented a type of Neoplatonism
which, via Plutarch of Athens, went back to Iamblichus.18 Consequently, there is

15
P. Merlan, ‘Ammonius Hermiae, Zacharias Scholasticus and Boethius’, Greek, Roman and Byzantine
Studies 9, 1968, 200–1; cf. id., ‘Ammonius, son of Hermias’, in C.C. Gillispie (ed.), Dictionary of
Scientific Biography, vol. 1, New York 1970, 137: ‘Its (i.e. the Alexandrian school’s) Platonism was in
many respects pre-Plotinian . . .’
16
A.C. Lloyd, ‘The later Neoplatonists’, in A.H. Armstrong (ed.), The Cambridge History of Later Greek
and Early Medieval Philosophy, Cambridge 1970, 316.
17
As late as 1976 Praechter’s theory was defended by T. Kobusch, Studien zur Philosophie des Hierokles
von Alexandrien. Untersuchungen zum christlichen Neuplatonismus, Munich 1976.
18
I. Hadot, Le problème du néoplatonisme alexandrin. Hiéroclès et Simplicius, Paris 1978, 143. Recently
N. Aujoulat, Le néo-platonisme alexandrin. Hiéroclès d’Alexandrie. Filiations intellectuelles et
spirituelles d’un néo-platonicien du Ve siècle, Leiden 1986, has made a renewed attempt to show that
Hierocles’ Neoplatonism was different from Roman and Athenian Neoplatonism. Hierocles is
believed to represent a typically Alexandrian Neoplatonism, which originated with Ammonius
Saccas and Origen the pagan. In assuming a link between Hierocles and Ammonius Saccas Aujoulat
follows Heinemann, Langerbeck and Theiler. One distinctive feature of Alexandrian Neoplatonism,
according to Aujoulat (55–65), is the fact that the Demiurge was considered the highest principle of
reality. Although Aujoulat is convinced that Hierocles was influenced in this by Christianity, his
general approach to Hierocles does more justice to the Neoplatonic character of his philosophy
than Praechter’s did. Nevertheless, his case for a theistic interpretation of Hierocles is unconvincing.
220 Aristotle Transformed

no reason to suppose that he was influenced by Christianity. The fact that

Hierocles nowhere refers to a principle above the Demiurge does not imply that
his philosophy was theistic. With regard to the de Providentia the argumentum e
silentio is very weak, given the limited length of Photius’ summaries. With regard
to the Commentary on the Carmen Aureum, where no god transcends the level
of Nous, Mme. Hadot emphasised that Hierocles was addressing himself to
beginners in philosophy. Hence a full development of the Neoplatonic theological
system in all its complexity would have been out of place.19 As to Simplicius’ in
Enchiridion Epicteti, the second support of Praechter’s theory: Mme. Hadot was
convinced that it was not possible in any way either to date the commentary
(even relatively) or to determine its place of composition.20 Moreover, Simplicius’
theology in this commentary is not pre-Plotinian and theistic at all. On the
contrary, it is identical with the theology of Simplicius’ other commentaries,
namely the metaphysical system of Damascius.21 But one should not forget that
in this case also considerations of a pedagogic and pragmatic nature lead to a
simplification or rather fragmentation of Neoplatonic metaphysics.22
So far Mme. Hadot did not extend her investigations to the school of Ammonius.
It is clear that her conclusions lead to the following question: if Praechter’s theory
is invalid with regard to Hierocles and Simplicius, what are we to think of its
validity with regard to the other Alexandrians? Isn’t it possible that they too are
Neo-platonists rather than Platonists in Praechter’s sense? And, secondly, if their
philosophy proves to be Neoplatonic, what exactly are the differences – if any –
between Alexandrian and Athenian Neoplatonism? In this chapter I shall focus
these questions on Ammonius. First I shall try to outline Ammonius’ metaphysics
as it appears in Asclepius’ in Metaphysica. Then I shall offer some remarks on
the Ammonius of Zacharias Scholasticus and the Ammonian commentaries on
the Organon. Thirdly, I shall briefly discuss Ammonius’ interpretation of the
Aristotelian concept of God. Finally, I shall deal with the problem of a possible
difference between Ammonian and Athenian Neoplatonism.23

19
I. Hadot, op. cit., 115–16.
20
ibid. 33–43. Mme. Hadot has changed her mind, and now thinks it was written at H.arrân: ‘La vie et
l’oeuvre de Simplicius d’après les sources grecques et arabes’, in I. Hadot (ed.), Simplicius, sa vie, son
oeuvre, sa survie, Berlin 1987, 18.
21
ibid. 47–65. Cf. also I. Hadot, ‘Le système théologique de Simplicius dans son commentaire sur le
Manuel d’Epictète’, in P. Hadot (ed.), Le néoplatonisme, Paris 1971, 265–79.
22
I. Hadot, Le problème du néoplatonisme alexandrin, 189–91.
23
The studies which K. Kremer devoted to the school of Ammonius, ‘Die Anschauung der Ammonius
(Hermeiou)-Schule über den Wirklichkeitscharakter des Intelligiblen. Ueber einen Beitrag der
Spätantike zur platonisch-aristotelischen Metaphysik’, Philosophisches Jahrbuch 69, 1961–2, 46–63;
id., Der Metaphysikbegriff in den Aristoteles-Kommentaren der Ammonius-Schule, Münster 1961, are
for several reasons hardly relevant to the present purpose. Their main defect is that they take little
account of the Neoplatonic character of the Ammonian school.
 The metaphysics of Ammonius son of Hermeias 221

Ammonius’ metaphysics in Asclepius’ in Metaphysica

Asclepius’ in Metaphysica was written ‘from the voice of Ammonius (apo phônês
Ammôniou tou Hermeiou)’, that is, as an edition of his lectures.24 The frequency
and docility of the pupil’s references to his ‘hero’ and ‘master’25 assure us that,
apart from the insertion of excerpts from Alexander,26 we may consider him
a faithful witness to Ammonius’ philosophy,27 and assume that the statements
in the first person plural28 render Ammonius’ viewpoint. Nowhere is there any
trace of divergence of opinion between the pupil and his master. This implies
that, whatever information about Ammonius we get from other sources
(Ammonius’ other commentaries, Zacharias, Philoponus, Simplicius), we should
try to harmonise it with what Asclepius tells us.
The text of Asclepius gives us at least some idea of the Neoplatonic character
of Ammonius’ metaphysics. On three points we have sufficient evidence to
venture a somewhat more detailed reconstruction. These three points are: the
concept of the highest principle, the doctrine of the divine Intellect, and
Ammonius’ view of the nature of the creative process.

(a) The highest principle

I do not think one can any longer doubt that Ammonius’ view of the highest
principle is Neoplatonic. To start with, let us consider what Asclepius says about
the first cause29 of reality. It is the principle (arkhê) of everything,30 the First,31 the
Divine.32 It is the Good, the final cause of the universe.33 At the same time it is the

24
Asclepius in Metaph. 1,2–3; 113,1–2; 137,2–3; 222,2–3.
25
ibid. 92,29–30.
26
cf. M. Hayduck, Praefatio, CAG 6.2, pp. v-vi.
27
L.G. Westerink, Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962, p. xi; L. Tarán,
‘Asclepius of Tralles, Commentary to Nicomachus’ Introduction to Arithmetic’, Transactions of the
American Philosophical Society, n.s. 59, 1969, 8.
28
e.g. Asclepius in Metaph. 102,17; 106,33–4; 140,15–16; 141,22; 145,15; 146,5; 147,28; 148,24; 151,1;
159,16; 163,11; 168,31; 181,32; 189,12; 190,5; 191,7; 194,13. . . .
29
‘The very first and transcendent cause (hê prôtistê kai exêirêmenê aitia)’: Asclepius in Metaph.
38,18–19; ‘the first cause’: (hê prôtê aitia): 56,24; (to prôton aition): 23,3; 35,10–11; 35,15; 39,17; 52,7;
115,37; (120,10–11 Alexander); 121,7–8; 147,4; 301,15.
30
‘(The)(only) principle of the totality ((Hê)(mia) arkhê (tôn) pantôn)’: Asclepius in Metaph. 17,7;
19,34–5; 48,5–6; 60,23–4; 99,5; 105,31–2; 148,10–11; 183,22–3; 186,2; 194,14; ‘properly the principle
of all beings (hê kuriôs arkhê pantôn tôn ontôn)’: 195,26; ‘the very first (of all) principle(s) (hê arkhê
(hê) prôtistê pasôn)’): 147,29–30; 148,33; ‘the only principle properly (hê mia kuriôs arkhê)’: 56,20;
‘the first principle (hê prôtê arkhê)’: 147,7; 148,27–8; 191,8.
31
to prôton: Asclepius in Metaph. 42,14; 52,1–2; 144,28; 146,6; 151,29; 158,18; 198,30; 225,16.
32
to theion: ibid. 439,26–7.
33
ibid. 15,8; (17,29–18,1 Alexander); 47,9; 52,1–2 (‘. . . all things have turned themselves towards the
First’); 106,33–107,4 (‘. . . the first henad, . . . towards which all beings have turned themselves’);
108,25; 123,16–17; (128,21–2 Alexander); 151,22–32; 450,25–7.
222 Aristotle Transformed

productive principle from which everything originates.34 The final cause of

reality is identical with its efficient cause.35 The Good is the One.36 The unity of
the One transcends not only difference (heterotês) but identity (tautotês) as
well.37 Accordingly, even the self-identity of Intellect (or Mind) is transcended by
the One. The unity of the One is neither numerical nor specific: numerical unity
cannot occur without matter, whereas the One, as the origin of matter, is
immaterial;38 specific unity is impossible as well, since the One transcends all
species and all universals.39
Not only is Ammonius’ view of the highest principle Neoplatonic, we can add
that his Neoplatonism must at least to some extent have been in keeping with
that of Syrianus and Proclus. This is obvious from his theory of henad, monad
and dyad. In the metaphysics of Syrianus, monad and dyad as a pair of principles
transcending all being come immediately after the One. They also reappear at
every level of being.40 The supreme monad is the principle of unity, rest and
identity,41 the supreme dyad is the principle of plurality, procession and
difference.42 In Proclus the supreme monad and dyad are called ‘the first limit’
(to prôton peras, autoperas) and ‘the first infinity’ (hê prôtê apeiria, autoapeiria)
respectively.43 Proclus thus uses the terms from the Philebus.44 On the other
hand, he often calls the One itself ‘the henad’.45 In Asclepius we find a combination
of Syrianus’ and Proclus’ terminology: monad and dyad derive from the henad.

34
The productivity of the One is often referred to by the terms paragein-paraktikos (Asclepius in
Metaph. 26,6; 28,6; 35,15–16; 52,7; 56,20–1; 106,23; 146,7–8; 147 passim; 148,27; 191,12–13;
194,13–17; 202,12; 217,3–5; 227,37) and proagein-proaktikos-proagôgos (105,32; 107,12; 123,16;
147,8–9; 151,19; 151,25; 176,12; 201,9–10; 201,26; 206,19; 216,23; 217,2–4; 226,13). The first
principle is said to possess a ‘generative potency (gonimos dunamis)’ (202,11; 202,19; 206,18–19;
208,21–2; 217,3–4; 225,16; 226 passim; 227,37); cf. 20,27–8; 123,15. The One is the ‘source’ of
goodness (21,19) and truth (115,37): everything is produced through its ‘gushing power’ (249,3–4).
The genesis of intelligible and sensible reality is a ‘procession’ (proodos: e.g. 223,35), a ‘proceeding’
(proienai: e.g. 196,6–7; 202,22; 225,15–17; 226,7; 230,4).
35
ibid. 108,23–5; 151,24–7.
36
to hen: Asclepius in Metaph. 33,34–5; 54,25–6; 249,3; ‘that which is essentially One (to hen to kat’
ousian)’: 151,18–19.
37
ibid. 144,28–30. Cf. 148,13–5. The identity of efficient and final causality in the One exists only for
us (151,26–7).
38
Asclepius in Metaph. 144,30–2; 148,24–7; 191,7–11; 217,4–5.
39
ibid. 191,11–13; 217,2–4; 148,28–32.
40
Syrianus in Metaph. 112,14–17. For my account of monad and dyad in Syrianus and Proclus I rely
on A.D.R. Sheppard, ‘Monad and dyad as cosmic principles in Syrianus’, in H.J. Blumenthal and A.C.
Lloyd (eds), Soul and the Structure of Being in Late Neoplatonism. Syrianus, Proclus and Simplicius,
Liverpool 1982, 1–5.
41
e.g. Syrianus in Metaph. 113,15–16.
42
e.g. ibid. 112,35–6.
43
e.g. Proclus El. Theol. prop. 90. See Dodds’s commentary, 246–8.
44
Plato Phil. 16C.
45
References in Proclus Theol. Plat. III , H.D. Saffrey and L.G. Westerink (eds), Paris 1978, p. xv.
 The metaphysics of Ammonius son of Hermeias 223

The One is the henad,46 the supraidentical origin of identity and non-identity.
This ‘definition’ of the highest principle as the henad is the starting point for a
comprehensive derivation of the rest of reality. Without being diminished itself,47
the henad gives rise to the monad and the dyad. The monad is the principle of
union, likeness, formative power, identity, and, in short, of all coherence. The
dyad is the principle of difference, division and unlikeness. The henad is the first
cause, which transcends both the monad and the dyad.48 These two subordinate
principles together are ‘intelligible number’,49 i.e. pure unity and twoness. They
generate definiteness (peras) and indefiniteness (apeiria).50 The duality of unity
and twoness is the basis of arithmetic: odd numbers are definite, even numbers
indefinite.51 The duality of definiteness and indefiniteness is the basic structure
of the whole of reality. Due to its indivisibility and priority, the monad is
identified with the Intellect (nous), which is stable (monimos).52 The dyad
is identical with Soul (psukhê), the moment of procession and discursive
thought.53 Thus the production of the hypostases Intellect and Soul from the
One is understood as the origination of a monad and a dyad from a henad.
The same process is repeated, at a lower level, as the origination of form (monad)
and matter (dyad) respectively from ‘productive Nature (phusis proagôgos)’.54
Form is a limit (peras) which defines and limits matter. Matter is indefinite and
infinitely divisible.55
The transcendence of the One56 excludes not only describing it affirmatively
(affirmative theology), but, strictly speaking, even the possibility of a negative
theology. ‘Concerning the First,’ Asclepius writes, ‘there is room neither for
affirmation nor for negation, since it transcends both being and every human
concept and in a word all intellective activity. For, even if Plato says that I do not

46
Asclepius in Metaph. 35 passim; 38,18; 48,5; 52,6; 99,5–6; 100,3; 106,22; 107,12; 178,1; 201,9;
201,25–6; 202 passim; 208,21; 216,23; 223,36. Cf. Proclus Theol. Plat. III Saffrey-Westerink, p. lxxvi.
47
Asclepius in Metaph. 202,11–12; 202,19–22; 206,18–20. The One is ‘infinite in potency’
(apeirodunamon: 121,8–9), it possesses a ‘superfluity of potency’ (periousia dunameôs: 194,14–15),
a ‘superpotent potency’ (dunamis huperdunamos: 439,26–7).
48
ibid. 35,2–16 (for the ascription to the Pythagoreans cf. n. 68 below); 147,4–6.
49
ibid. 201,23; 202,22–3; 207,18–19.
50
ibid. 38,13–16; 38,21–4; 39,7–9.
51
ibid. 38,21–9.
52
ibid. 36,22: monimos-monas.
53
ibid. 34,28–30; 38,13–14; 198,26–7; 202,8–9. In 201,40–202,9 monad and dyad are identified with
Empedocles’ Love (philia) and Strife (neikos) respectively. Cf. Syrianus in Metaph. 43,6–23 (A.D.R.
Sheppard, ‘Monad and dyad as cosmic principles in Syrianus’, 8–9).
54
Asclepius in Metaph. 38,11–13. In 202,23–5; 207,19–20; 207,23–7 and 208,18–22, on the other
hand, Asclepius calls nature (phusis) the (sensible) monad, not a henad. Hence the henadic function
continues to rest with the One directly.
55
Asclepius in Metaph. 38,22–4; 39,8–9.
56
ibid. 33,34–5.
224 Aristotle Transformed

know what it is, but know what it is not, negation cannot therefore really
comprehend it (i.e. the First). No, negation (in this case) just indicates that there
is no grasping. But the First cannot be grasped . . ., since it is beyond every
concept.’57 Asclepius especially emphasises the transcendence of the One in
relation to the next level down, intelligible reality or being, and for this he relies
on Plato58 and Plotinus.59 It is true that the henad is sometimes called ‘being in
the proper sense (kuriôs on)’,60 ‘being absolutely (haplôs on)’,61 ‘being really (ontôs
on)’,62 ‘substance absolutely (haplôs ousia)’63 or ‘the very first substance (prôtistê
ousia)’.64 However, these formulas are concentrated at the beginning of the
commentary on Metaphysica 4. One gets the impression that they are just
intended to interpret Aristotle’s definition of metaphysics as the theory of being
qua being and to give it a theological sense, i.e. to understand his being qua being
as the Good or the One, and the being of the remaining beings as a participation
in the goodness of the One.65 On the other hand, the One does not seem to lose
its supra-intelligible quality in this context.66 I think that the connexion of the
definition of metaphysics in Metaphysica 4 with the highest principle is to be
considered as a Neoplatonic interpretation of Aristotle rather than as an
Aristotelian contamination of Neoplatonism. For elsewhere in Asclepius’
Commentary it is repeatedly affirmed that the Good transcends being and
actuality.67 The One (autoen) and Being (autoon) are different hypostases. The
One can be predicated of being, but not vice versa. The henad is above being,
because being is a form (eidos), and the henad transcends all form, for it produces
the forms.68 The One not only transcends all being as its origin and final goal, it

57
ibid. 158,18–23.
58
ibid. 42,14–15.
59
ibid. 309,15–18.
60
ibid. 223,35–6; 225,15–17; 225,22; 225,34–226,3; 226,6–8; 230,4.
61
ibid. 226,6; 226,16–17; 227,2; 227,12.
62
ibid. 238,5; it is the subject matter of the ‘very first philosophy (prôtistê philosophia)’.
63
ibid. 226,17–18.
64
ibid. 232,9–10; ‘first substance (prôtê ousia)’: 309,14–15.
65
ibid. 226,3 (‘. . . granting being to all things’); 227,3–4 (‘it grants being to all things’). Cf. G. Verbeke,
‘Aristotle’s Metaphysics viewed by the Ancient Greek commentators’, in D.J. O’Meara (ed.), Studies in
Aristotle, Washington DC 1981, 122, 124. It is also characteristic of this theological approach to
ontology that it gives a theological interpretation to Aristotle’s resolution of ambiguities in different
uses of the predicate being. When Aristotle says that the different uses are all related to one thing
(aph’ henos, or pros hen, cf. e.g. Philoponus in Cat. 17,2–10), the reference to ‘one thing’ is taken to
imply a variety of ways of being dependent on, or oriented towards, the One. Cf. Asclepius in
Metaph. 223,33–6; 227,2–4; 232,18–25.
66
ibid. 226,2–4; cf. 439,26.
67
ibid. 77,16–17; 148,33–4; 202,31–2.
68
ibid. 201,4–12. Asclepius ascribes the theory to ‘the Pythagoreans and Plato’ (201,4), but at 201,
11–12 he accepts it as correct.
 The metaphysics of Ammonius son of Hermeias 225

also supports being, for in its ultimate emanation it takes on the guise of Matter
and in that guise receives the intelligible form of beings. Matter still participates
in the One, but, just like the highest principle itself, it is being and non-being at
the same time.69

(b) The second principle: the demiurgic Intellect

The Good is the source, the ground and the final goal (telos) of all existence. The
second divine principle is the Intellect. Ammonius seems to have accepted the
division of the Intellect into Being, Life and Thought.70 Did he also accept
Proclus’ setting of Plato’s Creator, the Demiurge, at the level of intellective being
(the noeron)?71 Not necessarily. As a matter of fact Asclepius does not define
more exactly or more explicitly the relation of the Demiurge to the Intellect as a
whole. However, the overall impression is that he simply identifies the two. The
Demiurge produces the sensible world on the basis of the demiurgic logoi,72
universal ideas73 which flow from the henad.74 So far, the picture is clearly
Neoplatonic. On the other hand, the use of the term ‘God’ (ho theos) might seem
to suggest the theistic view that the Intellect is after all to be identified with the
One as the supreme God.75 How are we to reconcile these apparently incompatible
doctrines? I think a text like Asclepius in Metaphysica 148,30–4 might contain
the core of the solution: ‘But it is God (ho theos) who sows those universal logoi
in souls and nature; hence He even transcends universality, and with him nothing
is potential, but He is actuality without potentiality in the case of the Intellect,

69
ibid. 233,29–31; cf. Ammonius in Int. 213,8–10.
70
Asclepius in Metaph. 202,27–32: ‘That the henad precedes everything is clear from what follows.
Life is above Intellect. For whatever participates in Intellect also participates in Life, whereas not
everything that participates in Life also participates in Intellect. Again, Being is at all events above
Life. For whatever participates in Life also participates in Being, whereas not everything that
participates in Being also participates in Life. The henad, however, is even above Being, since it is
above existence.’ Cf. e.g. Proclus El. Theol. prop. 101 and Dodds’s commentary 252–3; for further
references W. Beierwaltes, Proklos. Grundzüge seiner Metaphysik, Frankfurt a.M. 1979, 2nd edition,
106 n. 69.
71
cf. e.g. R. Beutler, ‘Proklos 4’, in G. Wissowa and W. Kroll (eds), Realencyclopädie der klassischen
Altertumswissenschaft XXIII , 1, 1957, col. 228.
72
Ammonius-Asclepius not only ascribe the doctrine of the demiurgic logoi to Plato and Aristotle; it
is an essential part of Ammonius’ own system as well. Cf. e.g. Asclepius in Metaph. 48,5–7; 70,2–4;
73,21–3; 75,32–3; 76 passim; 80,30–1; 81,2–4; 88,4–7; 89,21–2; 142,38–143,1; 145,17–18; 165,35–
166,1; 166,29–31; 167,32; 168,34–5; 173,22–3; 173,31–3; 175,16–18; 209,35–7; 216,24–6; 218,3–5.
The term is also to be found in Zacharias, Ammonius ll.371; 388; 511–12; 766–7; 871–2 Colonna.
73
Asclepius in Metaph. 173,22; 175,16; 183,21–3.
74
ibid. 48,5–6; 52,6–7; cf. 54,25–6.
75
cf. Praechter, nn. 5 and 12 above.
226 Aristotle Transformed

whereas the very first principle even transcends actuality.’76 The quotation shows
that the use of the term ho theos is perfectly compatible with a distinction
between the divine Intellect (Nous) and ‘the very first principle (hê prôtistê
arkhê)’. This ambiguity makes it possible to indicate, by the single term ho theos,
either the highest principle alone77 or the Demiurge,78 or, as we have just seen,
both together.79 Now, this possibility of either unfolding or telescoping, as it
were, the concept of God according to the needs of the context, seems to be one
of the basic features of Ammonius’ metaphysics. Hence even an apparently
plainly theistic terminology does not mean that Ammonius identified the
supreme God and the Demiurge. It only means that it is not always necessary to
draw a distinction between the two principles. I shall return to this point later.
But it may already be relevant to refer to a similar practice in Christian theology:
one can consider the Christian God either as a simple principle (namely ad
extra, in His relation to the world) or as a complex principle (namely ad intra, in
so far as the relation between the persons of divine Trinity is concerned).

(c) Divine causality

In the third place, the Commentary of Asclepius gives us some information about
Ammonius’ conception of divine causality. This causality is of a double nature: it is
both efficient and final.80 To take the final aspect first: everything is directed
towards the One as its goal or final cause.81 But the Demiurge too is said to move
the universe as a final cause or good: the circular movement of heavenly bodies is
the sensible imitation of the reflexive thought of the divine Intellect.82 As to
efficient causality: it is conceived as a stratified process. The Good produces, from
the abundance of its power, what are called ‘intelligible forces’,83 ‘intellective
forces’,84 angels and souls.85 The articulation of this intelligible world remains very
vague. Rather, the main emphasis is on its absolute dependence on the highest
principle. Divine creativity is an eternal process; to ‘give being’86 is the essence of

76
Asclepius in Metaph. 148,30–4: . . . energeia aneu dunameôs epi tou nou, hê de prôtistê arkhê kai
huper energeian estin.
77
e.g. ibid. 20,25–6; 28,6; 41,31–3; 108,24; 144,30–1; 195,26–7; 309,12–13.
78
e.g. ibid. 81,2–4.
79
e.g. ibid. 77,8–17; 122,25–6; 147,32; 148,30–4; 158,27–8.
80
cf. n. 35 above.
81
cf. n. 33 above.
82
Asclepius in Metaph. 151, 8–12.
83
ibid. 11,35–6; cf. 147,6–11.
84
ibid. 57,9; 119,33–4 (noerai ousiai).
85
ibid. 57,8; 226,3. See further M. Hayduck, Index verborum, CAG 6, 2 s.v. angelos (p. 455).
86
ibid. 226,2–3; 227,3–4: to einai kharizesthai.
 The metaphysics of Ammonius son of Hermeias 227

the Good. God’s simplicity makes His goodness absolutely necessary.87 At the
same time this necessity is absolute freedom: only God is free, because he does not
need anything outside Himself, whereas everything else needs Him.88 This means
that Ammonius, just like the other Neoplatonists, cannot avoid the aporetic
conclusion that on the one hand the Good is not dependent on anything outside
itself, but on the other hand it necessarily gives existence to other things. But this is
one of the mysterious truths which can only be apprehended in a mystical way.89
Its mythical expression is the duality of ambrosia and nectar. In God, Asclepius
writes, there are two perfections. The first one (absolute freedom) pertains to His
essence, to what He is in Himself: it is symbolised by the solidity of ambrosia. The
second perfection is symbolised by the fluidity of the nectar: it is necessary
emanation.90 At the level of the demiurgic Intellect divine goodness becomes
conscious (though not deliberate) creation. Because the ideas are immanent in it
(dêmiourgikoi logoi),91 the divine Intellect eternally contemplates itself. Hence, at
this level, the creation of the material world is the product of the creative Intellect’s
reflection on itself.92 But Asclepius does not answer the question how divine
causality can break out of this self-sufficient self-consciousness. This means that
the relation between the Intellect’s reflexive thinking and its creativity is just as
problematic as the relation between the freedom and the emanation of the Good.

The Ammonius of Zacharias Scholasticus and

the Ammonian commentaries on the Organon

(a) Zacharias
Zacharias, who later became the Bishop of Mytilene, was a student of Ammonius
in Alexandria c. 485–7.93 The teaching of Ammonius and the contradiction
between his belief in the eternity of the world and Christian doctrine prompted
Zacharias to write, perhaps during his stay in Beirut (487–91), a polemical

87
ibid. 309,10–15; 309,24–6; 368,13–15.
88
ibid. 20,25–8.
89
ibid. 309,15–18.
90
ibid. 195,37–196,10.
91
ibid. 88,4–7.
92
ibid. 151,9–11; cf. 70,26–9.
93
G. Bardy, ‘Zacharie le Rhéteur’, Dictionnaire de Théologie Catholique, vol. 15, Paris 1950, col. 3677;
K. Wegenast, ‘Zacharias Scholasticus’, in G. Wissowa and W. Kroll (eds), Realencyclopädie der
klassischen Altertumswissenschaft IXA , 2, 1967, col. 2212; M. Minniti Colonna, Zacaria Scolastico,
Ammonio, Naples 1973, 23.
228 Aristotle Transformed

dialogue called Ammonius or de Mundi Opificio.94 The main part of the work has
four sub-divisions. The first section (ll.92–350 Colonna) is introduced as the
account of a discussion between Ammonius and Zacharias (referred to as
Khristianos) about the eternity of the world a parte ante and a parte post.95 The
second section (ll.351–937) reports a dispute on the same subject between
Zacharias and the medical specialist (iatrosophist) Ges(s)ius, a pagan student of
Ammonius. In the third section (ll.938–1093) Zacharias discusses, with
Ammonius again, the co-eternity of the world with God. In the last section,
which is very short (ll.1094–1136), the same interlocutors deal with the problem
of the divine Trinity.
At first sight there seems to be no difference between Zacharias’ theistic
concept of God and the one ascribed to Ammonius.96 But one should not forget
that both Zacharias’ and Ammonius’ arguments are focused on the problem of
the eternity of the sensible world. Hence the articulation ad intra of the divine
creative principle is not the point at issue here. The fact that the discussion does
not enter into the Neoplatonic distinction between the One and divine Intellect
does not imply that Ammonius believed in a supreme personal God, any more
than Zacharias’ reference to the Creator excludes his belief in the doctrine of
divine Trinity, expounded ll.1094–1136. The main discussion between Zacharias
and his master is about the relation of the sensible cosmos to its closest creative
principle, the Demiurge. This problem is first of all what one might call a problem
of natural philosophy, whereas the subordination of the divine Intellect to the
highest principle pertains to metaphysics.97 The situation of Ammonius here is
comparable to that of his master Proclus in his de Aeternitate Mundi (which

94
The dating of the Ammonius in 487–91(92) is accepted by G. Bardy, op. cit, col. 3679 and M. Wacht,
Aeneas von Gaza als Apologet. Seine Kosmologie im Verhältnis zum Platonismus, Bonn 1969, 18 n.
17; it is regarded as likely by K. Wegenast, op. cit., col. 2215. Originally, Mrs Minniti Colonna was of
the opinion that Zacharias wrote his dialogue at least 10 or 20 years after 480–90, when Aeneas of
Gaza composed his Theophrastus (‘Zacaria Scolastico, il suo Ammonio e il Teofrasto di Enea di
Gaza’, Università di Napoli. Annali della Facoltà di Lettere e Filosofia 6, 1956, 117 n. 9). In her edition
of the Ammonius (p. 30) Mrs Minniti Colonna thinks that Zacharias probably wrote his dialogue
in Constantinople (where he settled in 491: ibid. p. 25), more specifically after his definitive
renunciation of monophysism (which, maybe, must be set in 512: ibid. p. 26), maybe even after the
death of Aeneas of Gaza (c. 518) (ibid. p. 44). Regardless of whether one prefers the earlier or a later
dating of the Ammonius, this much is certain, that the introduction of the dialogue, the talk between
Zacharias and his friend, is set at the time of Zacharias’ stay in Beirut. Accordingly, the ‘now’ in
Ammonius l.22 Colonna refers to the period 487–91.
95
According to P. Courcelle, ‘Boèce et l’école d’Alexandrie’, Mélanges d’Archéologie et d’Histoire de
l’Ecole française de Rome 52, 1935, 216, this discussion really took place during the summer of 486
(or perhaps 487). Cf. M. Minniti Colonna, Zacaria Scolastico, Ammonio, 39.
96
Zacharias Ammonius ll.108 (‘. . . the Father and Maker of this universe’); 122 (‘. . . the Demiurge . . .’)
etc. God simply is ho theos (ll.117; 123; 126 etc.).
97
cf. below.
 The metaphysics of Ammonius son of Hermeias 229

survives in the de Aeternitate Mundi contra Proclum of John Philoponus).

Nowhere in that work does Proclus speak about the place of the Demiurge in
the world of intelligible reality, because it is sufficient in an argument about the
eternity of the sensible world simply to accept a transcendent creative principle.
And no one would call Proclus’ theology theistic because of his not offering in
his de Aeternitate Mundi a derivation of the demiurgic principle from higher
intelligible reality.

(b) Ammonius’ in de Interpretatione

The foregoing may have shown that Zacharias’ dialogue yields no decisive
indications in support of Merlan’s theistic interpretation of Ammonius’ theology.98
Conversely, it is not possible either to find much in it to prove positively the
Neoplatonic character of Ammonius’ teaching on God and creation. Anyway, the
discussions between Ammonius and Zacharias are chronologically close to
Ammonius’ in de Interpretatione, which, as we shall see, contains clearly
Neoplatonic elements. This commentary indeed must be dated to the early period
of Ammonius’ philosophical activity,99 and Zacharias is referring to the beginning
of Ammonius’ Alexandrian professorship.100 It is not very likely that Ammonius
temporarily gave up Neoplatonism in or about 486 (the time at which Zac.harias
was his student) and afterwards, e.g. in his Commentary on the Metaphysics
(recorded by Asclepius), returned to it.101 As to in de Interpretatione, I would like
to show (1) that the influence of Proclus is found in this commentary, and (2) that
it contains some evidence of the possibility of treating intelligible reality in two
different ways, each of which is valid at its own level.
There can hardly be any doubt that in de Interpretatione, in spite of the
sporadic use of terms like ‘God (ho theos)’102 or ‘the Demiurge (ho dêmiourgos)’,103
has a basically Neoplatonic context. Aristotle’s distinction of spoken sound
(phônê), psychological state (pathêma tês psukhês) and objective reality
(pragma)104 is interpreted as corresponding to the three Neoplatonic hypostases:

98
cf. above.
99
cf. L.G. Westerink, The Greek Commentaries on Plato’s Phaedo, vol. 1: Olympiodorus, 19.
100
Zacharias Ammonius ll.19–24.
101
It seems legitimate to suppose that Asclepius’ in Metaphysica, written ‘from the voice of Ammonius’,
is posterior to Ammonius’ in de Interpretatione, written by Ammonius himself (cf. for this L.G.
Westerink, Anonymous Prolegomena, xi; L. Tarán, ‘Asclepius of Tralles’, 14).
102
Ammonius in Int. 1,11; 217,28–9; 240,10–11.
103
ibid. 250,12–19.
104
Aristotle Int. 16a3–8.
230 Aristotle Transformed

Soul, Intellect, God.105 Intellect is not the highest level of reality: it is transcended
by ‘non-being’, in the sense of that which is above being and form, and which
produces both form and matter.106 The highest principle is just as formless as
matter.107 Moreover, I think it is possible to find, in Ammonius’ in de
Interpretatione, traces of the divine henads which Proclus tries to accommodate
at the level of the One. At one place Ammonius refers to ‘divine ranks (theiai
diakosmêseis)’ which are produced by the unique principle of reality, and each of
which possesses its own ‘specific character (idiotês)’.108 This recalls Proclus’
terminology for the henads,109 as well as his theory that each henad, because of
its specific character, is one particular ‘excellence (agathotês)’.110 Further allusions
to the henads are perhaps to be found at in Int. 132,14–16,111 243,2–4112 and
249,17.113 In the digression on contingency and determinism (in Int. 130,27–
138,10), on the other hand, Ammonius had the opportunity of being somewhat
more explicit on the role of the henads, which, according to Proclus, are the
primary locus of providence, of pro-noia, understood as ‘that which precedes the
Nous’.114 In fact, Ammonius appears here to consider the gods as belonging not
to the first level of reality, but to the second, that of being and of intelligible
entities,115 although his characterisation of divine knowledge resembles that by
Proclus,116 whose henads are at the first level. An example of Ammonius even

105
Ammonius in Int. 24,24–9: ‘. . . since there are three originative orders above physical substances,
namely the divine, the intellective and moreover the psychic order, we say that things (pragmata)
are created by the divine, thoughts (noêmata) are brought about by intellects, and words (phônai)
are produced by those souls which are characterised by reason and have an essence separated from
all body.’
106
ibid. 213,1–5.
107
ibid. 213,8–10.
108
ibid. 135,28–32.
109
Proclus El. Theol. prop. 132, p. 116,30 Dodds (‘the ranks of the gods . . .’); prop. 151, p. 132,27 (‘the
divine ranks’).
110
ibid. prop. 133, p. 118,8–14.
111
Ammonius in Int. 132,14–16: ‘. . . or rather (if I have to say it more accurately) having established
their own true existence even above intellective particularity itself.’
112
ibid. 243,2–4: ‘. . . Aristotle thinks fit to call the gods “intellects”, and perhaps he indicates that they
possess a nature which is even more august than intellect.’
113
ibid. 249,17: ‘. . . the intelligible gods and the secret causes above them. . . .’
114
Proclus El. Theol. prop. 120, p. 104,32; ‘. . . the primary providence resides in the gods’ (Dodds).
115
cf. Ammonius in Int. 133,19–23 (‘things which really have being . . .’); 136,20–4.
116
Compare e.g. Ammonius in Int. 136,15–17 (‘It is necessary indeed that the gods have an undivided
and unextended knowledge of things divided, a unitary knowledge of things plurified, an eternal
knowledge (aiôniôs) of things temporal, and an ungenerated knowledge of things generable’) with
Proclus El. Theol. prop. 124, p. 110, 10–13 (‘Every god has an undivided knowledge of things
divided and a timeless knowledge of things temporal; he knows the contingent without contingency,
the mutable immutably . . .’) (Dodds). For Proclus, however, it would be impossible to use
Ammonius’ term ‘eternally (aiôniôs)’, which is appropriate to the second level of reality: the henads
transcend eternal being (Proclus Theol. Plat. I,99,6–7 Saffrey-Westerink). I have not been able to see
L. Obertello, ‘Proclus, Ammonius and Boethius on Divine knowledge’, Dionysius 5, 1981, 127–64.
 The metaphysics of Ammonius son of Hermeias 231

completely neglecting the henads is in Int. 248,7–249,26.117 However, I do not

think that Ammonius rules out the existence of henads here. Rather, I have the
impression that, generally speaking, he accepts the distinction between the
henads at the higher level and the intelligible gods at the lower. But the theory of
henads arises only when one goes into a kind of detail (akribeia)118 that is not
indispensable in all circumstances. Hence the references to the henads have only
a supplementary or even redundant character, either especially emphasising the
distance between the material world and the highest region of reality,119 or
adding a further rung for the ascent of our intellect.120 They are introduced only
occasionally when the context makes it necessary to introduce them from below.
For it is not the task of the in de Interpretatione to deduce the metaphysical
system in full completeness from the highest principle. The commentator merely
has to refer to this principle from time to time, to explain the conditions of the
possibility of the sensible world. I think it is legitimate to consider this economical
type of metaphysics as a Neoplatonic version of Aristotle’s ‘theology from the
standpoint of natural philosophy (phusikôs theologein)’.121 In principle, there is
no need for this ‘minimal’ metaphysics to meet the same requirements of detail
(akribeia) as a complete and systematic metaphysics. Obviously, the theory of
henads is an item for the latter rather than for the former. Hence, even Ammonius’
account of the relation between contingency and determinism can virtually do
without the henads: in the non-metaphysical context of in de Interpretatione it is
sufficient to set providence at a level where, in Proclus’ system, it is only secondary,
namely at the second level, that of the intelligible gods.122
Consequently, the way in which Ammonius, in his early in de Interpretatione,
deals with the henads is closely related to what we have already observed in
Asclepius’ in Metaphysica and in Zacharias’ Ammonius, namely the possibility of
unfolding or telescoping the concept of God. In Asclepius’ in Metaphysica the
henads have disappeared. This has rightly been noticed by Saffrey and Westerink.

117
The gods (ontôs theoi: Ammonius in Int. 248,31) are ‘pure acts’ (autoenergeiai: 248,18; 248,28);
further 248,17–18: ‘. . . the very first and supreme among beings, that is to say the intelligible and
divine ones. . . .’
118
ibid. 132,15: ei khrê to akribesteron phanai; cf. n. 111 above.
119
ibid. 132,14–16; 243,2–4.
120
ibid. 135,28–32.
121
The commentators often oppose Aristotle’s ‘theology from the standpoint of natural philosophy
(phusiologikôs theologein)’ to Plato’s ‘natural philosophy from the standpoint of theology (theologikôs
phusiologein)’ (cf. K. Kremer, Der Metaphysikbegriff, 189–95): e.g. Elias in Cat. 120,30–121,3;
124,21–3; Philoponus in Phys. 5,21–5.
122
cf. Proclus El. Theol. prop. 134, p. 118,20: ‘Every divine intelligence exercises intellection qua
intelligence, but providence qua god’ (Dodds).
232 Aristotle Transformed

They explain it as follows: ‘Despite the close links that existed between the
schools of Athens and of Alexandria, there was at Alexandria a more strictly
monotheist tradition than at Athens, and that excluded a priori the possibility of
teaching the theory of divine henads there with approval.’123 However, I think in
de Interpretatione shows that Ammonius at the beginning of his career still
accepted the doctrine of henads. Later he seems to have given it up. But this
should not, in my opinion, be interpreted as a consequence of any theistic
tradition. Rather, it is one element of Ammonius’ eventual simplification of
Neoplatonism in comparison with Proclus.

(c) Ammonius’ in Isagogen, in Categorias and in Analytica Priora

Like in de Interpretatione, the other logical commentaries contain only a few
indications concerning Ammonius’ theological views. Yet they are sufficient to
confirm what we have found so far. The Prooemium of the in Isagogen has some
references to a substantial, omniscient and provident God which sound very
theistic.124 However, since this prooemium is probably not authentic,125 the fact
does not demand particular attention. Things become recognisable again in the
opening words of the corpus of the commentary, where it is stated that the Good
is the principle of all reality.126 We also find the classical Neoplatonic doctrine of
the presence of paradigms in the demiurgic Intellect.127 The definition which
Ammonius gives of the goal (telos) of Aristotle’s philosophy in the Prooemium of
his in Categorias128 is characteristic of his Neoplatonic exegesis. The highest
principle is said to be the subject of negative theology.129 I think a formula like
‘there is nothing above the first beings (huper de ta prôta ouden)’130 should be
understood, not as a denial of the existence of the One,131 but merely as a

123
Proclus Theol. Plat. III , H.D. Saffrey and L.G. Westerink (eds), p. lxxxvi.
124
Ammonius in Isag. 3,9–15; 4,7–8; 11,11–13; cf. 9,16–17. See n. 12 (Praechter).
125
A. Busse, Praefatio, CAG 4.3, vi n. 4.
126
Ammonius in Isag. 24,2–5: ‘All beings by their very nature long for the Good, since there is one
principle of all beings, the Good. Therefore all things tend upwards to the Good as their own good,
and every being is perfected by participating in the Good in accordance with its own limitations.’
127
ibid. 41,20–42,26; 63,14–16.
128
Ammonius in Cat. 6,10–12: ‘. . . to raise oneself up to the common principle of all things, and to
come to know that it is unique, incorporeal, without parts, not to be comprehended, undefinable
and of infinite potency: absolute goodness . . . autoagathotês:’ cf. e.g. Proclus El. Theol. prop. 127,
p. 112,33; Elias Proleg. 17,2.
129
Ammonius in Cat. 36,25–6.
130
Ammonius in An. Pr. 25,20.
131
This is the opinion of K. Kremer, Der Metaphysikbegriff, p. 39: ‘The noetic things do not have a
ground any more to found their logos and existence on.’
 The metaphysics of Ammonius son of Hermeias 233

statement that there are no higher genera beyond the first substances,132 i.e. the
intelligible gods.133 Finally, Ammonius may have identified the absolute Good
with absolute Beauty.134

Ammonius’ interpretation of Aristotle’s concept of God

I have so far spoken of Ammonius’ own theology, but now we arrive at what is
undoubtedly the most famous, not to say the only tenet in Ammonius’ philosophy
that is generally known: his doctrine of the nature of Aristotle’s God. So far, an
erroneous interpretation of this doctrine has been the main source for the
conviction that Ammonius adopted the Christian identification of the supreme
God and the demiurgic Intellect. I do not pretend to be able to propose a
complete solution of the problem at the present stage of my research. For this
would require further investigations into its history before Ammonius and e.g.
into Simplicius’ ideas concerning the relation between physics and metaphysics.
But a first step can already be taken: I think it is possible to reconcile Asclepius’
and Simplicius’ apparently incompatible accounts of the theory of Ammonius.

(a) Simplicius
At both in Cael. 271,13–21 and in Phys. 1360,24–1363,24, Simplicius tells us that
to Ammonius the Aristotelian God is not only the final but also the efficient
cause of the universe:135 he mentions a monograph by Ammonius completely
devoted to this question. As late as the tenth century al-Farabi refers to this
monograph as the source of his own interpretation of Aristotle.136 Simplicius

132
Ammonius in An. Pr. 25,11–20.
133
Ammonius in Cat. 26,2–3; 35,20–1; 37,1–10.
134
cf. Ammonius in An. Pr. 5,19–22.
135
Most scholars discuss the issue just on the basis of Simplicius’ texts, without referring to Asclepius
at all: e.g. P. Tannery, ‘Sur la période finale de la philosophie grecque’, in P. Tannery, Mémoires
scientifiques, vol. 7, Toulouse and Paris 1925, 226–7; E. Zeller, Die Philosophie der Griechen III , 2,
894 n. 1; K. Praechter, ‘Simplikios’, col. 211; R. Vancourt, Les derniers commentateurs alexandrins
d’Aristote: l’école d’Olympiodore, Etienne d’Alexandrie, Lille 1941, 18–21; M. Wolff, Fallgesetz und
Massebegriff. Zwei wissenschaftshistorische Untersuchungen zur Kosmologie des Johannes Philoponus,
Berlin 1971, 129 n. 32; id., ‘Philoponus and the rise of preclassical dynamics’, in R. Sorabji (ed.),
Philoponus and the Rejection of Aristotelian Science, London and Ithaca NY 1987, 108.
136
E. Behler, Die Ewigkeit der Welt. Problemgeschichtliche Untersuchungen zu den Kontroversen
um Weltanfang und Weltunendlichkeit im Mittelalter, vol. 1, Munich, Paderborn, Vienna 1965,
74–5, 128, 137; M. Mahdi, ‘Alfarabi against Philoponus’, Journal of Near Eastern Studies 26, 1967,
236–7 n. 9.
234 Aristotle Transformed

notes that it was the purpose of the theory to harmonise Aristotle with Plato,137
and that it was opposed to orthodox Peripatetic exegesis, including Alexander’s.138
We may add that it was opposed to Proclus’ exegesis as well: Aristotle’s conception
of God, Proclus writes, contains only half of the truth.139 Ammonius by contrast
is credited by Simplicius140 with five instances to prove that Aristotle considered
God as an efficient cause:141 Phys. 2.3, 194b29–31; Cael. 1.4, 271a33; Cael. 1.9,
279a27–30;142 GC 1.3, 318a1–5;143 Metaph. 1.4, 984b15–22. However, a further
question is this: of what exactly is Aristotle’s God the efficient cause? No one
would deny that He is the efficient cause of sublunary movement, albeit indirectly,
by directly moving the heavens.144 According to Simplicius Alexander even
accepted that to Aristotle God was the efficient cause of heavenly movement.145
At the same time, however, Alexander denied that the unmoved Mover was the
efficient cause of heavenly substance.146 Ammonius, on the other hand, was
convinced that Aristotle’s God was the efficient cause of the whole universe (tou
pantos kosmou),147 which seems to mean: both of its movement (heavenly and
sublunary) and of its existence as a substance (heavenly and sublunary).
Therefore Simplicius, who agrees with Ammonius, adduces two arguments
against Alexander’s limitation of efficient causality to movement.148 The first
argument is this. That which receives its movement from elsewhere, also receives
its existence from elsewhere.149 Heaven receives its movement directly from the

137
Simplicius in Phys. 1360,28–31. The point was rightly emphasised by H.D. Saffrey, ‘Le chrétien Jean
Philopon et la survivance de l’école d’Alexandrie au VIe siècle’, Revue des études grecques 67, 1954,
400 n. 3. Cf. n. 184 below.
138
Simplicius in Cael. 271,13–14; in Phys. 1362,11–15.
139
Proclus in Tim. I, 266,28–267,12 (cf. L.G. Westerink, Anonymous Prolegomena, 19); in Parm.
842,26–35; cf. further in Parm. 788, 12–28 (cf. A.J. Festugière, Proclus, Commentaire sur le Timée II ,
Paris, 1967, 108, n. 4.); Philoponus Aet. 31,7–32,10. Probably the term ‘some men’ (tines) in
Simplicius in Phys. 1360,24 includes Proclus; cf. 258,14.
140
Simplicius in Phys. 1363,8–12. The entire passage 1361,11–1362,12, is translated in Richard Sorabji,
Matter, Space and Motion, London and Ithaca NY, 1988, 275–7.
141
Simplicius in Phys. 1361,12–1362,10.
142
cf. n. 191 below.
143
cf. Philoponus in GC 50,1–5. Cf. n. 189 below.
144
Simplicius in Phys. 1362,17–20.
145
ibid. 1361,31–3; cf. 258,13–25.
146
ibid. 1362,11–15. R. Vancourt, Les derniers commentateurs, 20–1 writes: ‘Simplicius devotes a special
paragraph to refuting Alexander’s interpretation, according to which God is the final cause and not
the efficient cause of heavenly movement, which is itself ungenerated and eternal.’ I do not think
this is what Simplicius’ text says. The sentence ascribed to Alexander is this: ‘The first mover is the
efficient cause of the movement of the divine body, which is ungenerated (to goun prôton kinoun tês
tou theiou sômatos kinêseôs aition esti poiêtikon ontos agenêtou)’, Simplicius in Phys. 1361,32–3.
Vancourt erroneously takes ontos agenêtou to refer to kinêseôs; if that were correct we would have
ousês agenêtou.
147
Simplicius in Phys. 1363,9–10; cf. 1360,30–1.
148
ibid. 1362,11–1363,8.
149
ibid. 1363,2–4.
 The metaphysics of Ammonius son of Hermeias 235

Unmoved Mover as efficient cause.150 Therefore, it receives its corporeal substance

from the incorporeal Unmoved Mover as well.151 The second argument is that at
Phys. 2.6, 198a5–13 where Aristotle says Intellect and nature are the causes of
‘this universe (tou ouranou toude; toude tou pantos)’.152 Accordingly, Ammonius’
ascription of an efficient causality to the Unmoved Mover represents a major
alteration, not only of Aristotle’s concept of God, but also of his concept of
efficient causality: God is the efficient cause of the universe, not only as the
principle of its movement (hothen hê arkhê tês kinêseôs),153 but also as the
principle which produces its very existence (hupostasis) as a substance (ousia).154
Moreover, the efficient causality at issue is an eternal relation. This is the reason,
Simplicius adds, why Aristotle was less explicit in calling God the efficient
cause of the world than he was in calling Him its final cause: the idea of
efficient causality is too readily associated with the idea of a temporal beginning
of the effect.155
The foregoing already suggests that the background of Ammonius’ theory is
Neoplatonic. Aristotle is said to conceive the relation between God and the
world as a double process: on the one hand, the world is produced by God as its
‘efficient’ cause (this is its procession – proodos), on the other hand it is directed
towards God as its final cause (this is its reversion – epistrophê). Yet, there is no
trace, in Simplicius’ account, of any corresponding ‘Neoplatonisation’ of
Aristotle’s concept of God itself.156 According to Simplicius, Ammonius simply
considered Aristotle’s God as the divine Intellect.157

150
ibid. 1362,17–19.
151
ibid. 1363,4–8.
152
ibid. 1362,20–1363,2.
153
Aristotle Phys. 2.6,198a3.
154
Simplicius in Phys. 1363,2–8.
155
ibid. 1363,12–24.
156
‘Neoplatonisation’ is meant here as the ‘distinction between the Good and the Intellect’. Simplicius
does not say anything here about the fact that Ammonius’ reinterpretation implies that Aristotle’s
God is no longer self-thinking Thought, but a demiurgic Intellect in the Neoplatonic sense, which
includes the Ideas in itself. Obviously, however, he is aware of this, and he himself is convinced that
for Aristotle also the ideas are immanent in the demiurgic Intellect (in Cael. 87,3–11).
157
e.g. Simplicius in Phys. 1360,25; 1362,7–8; 1362,16–17; 1362,32. It is true that Simplicius in Cael.
271,21–7 calls Aristotle’s God ‘the One (to hen)’. However, this predicate here concerns only the
uniqueness of the first Unmoved Mover, and as only ‘the object of desire (monôs epheton)’ at that.
Neither does the fact that Aristotle’s God at in Phys. 257,13–16 is said to be both ‘intellect (nous)’
and ‘good (agathos)’ mean that Simplicius here identifies the Good and the divine Intellect. Rather,
the term agathos here refers to the final causality of the Intellect with regard to the world. Simplicius
understands the goodness (agathotês) of the Platonic Demiurge (Tim. 29E1) as final causality
(e.g. in Phys. 7,17–19; 26,15–18; 43,6), and he puts Aristotle’s divine Intellect at the same level
as Plato’s Demiurge (cf. in Phys. 1360,28–34). For Simplicius’ own real view on Aristotle’s God cf.
n. 182 below.
236 Aristotle Transformed

(b) Asclepius158
In Asclepius the picture is different: Ammonius is said to have identified
Aristotle’s God with the highest principle of Neoplatonism, not with divine
Intellect. This interpretation is influenced, of course, by the fact that Aristotle
himself calls God the ‘first principle (arkhê)’ of the entire cosmos,159 and that, as
everyone agreed,160 he considers God the final cause,161 that is, the final cause of
being. He even calls Him ‘the Good (to agathon)’.162 To Ammonius, if we rely on
Asclepius, Aristotle’s God is ‘the Good, which everything is longing for’:163 it is
the Neoplatonic first cause.164 This God is said to be not only the final cause, but
also the absolute origin of all reality.165 Terming the latter aspect an ‘efficient
causality’166 is to speak rather loosely: Aristotle conceived efficient causality only
as an activity exerted upon an already existing substance.167
But what are we to do about the fact that Aristotle describes God as an
Intellect? In principle, Asclepius’ identification of Aristotle’s God with the
Neoplatonic first cause suggests the possibility that Aristotle’s God contains both
a supreme principle and, distinct from it, an Intellect. As far as I know, Asclepius
does not say anything like that explicitly. But an indirect indication of Ammonius
applying that distinction to Aristotle might be found in the following argument.
We are told by Asclepius168 that Ammonius ascribed to Aristotle the distinction
between Being and the One with regard to the lowest level of the ontological

158
According to L.G. Westerink, Anonymous Prolegomena, 19; The Greek Commentaries on Plato’s
Phaedo, vol. 1: Olympiodorus, 24, Ammonius’ interpretation of Aristotle’s concept of God is also to
be found in Philoponus (= Stephanus), in DA 571,1–5. However, it should be noted that the
beginning of that passage is not about the object of appetency (orekton) (a final cause of movement,
like the Unmoved Mover; cf. Proclus in Tim. I, 267,9) but about the orektikon, i.e. the impulsive
faculty of the soul. It is this faculty that is said to act as an efficient cause. At the end of the passage
the intellect is called ‘an efficient, not a final cause (poiêtikos kai ou telikos)’, which does not recall
Ammonius’ view on Aristotle’s theology either. I think that the text is purely psychological, and that
Stephanus does not allude here to the Aristotelian divine Intellect.
159
Aristotle Metaph. 12.7, 1072b13–14.
160
cf. Simplicius in Phys. 1361,11–12.
161
Aristotle Metaph. 12.7, 1072b1–3.
162
ibid. 12.10, 1075a11–12.
163
Aristotle EN 1.1, 1094a3, quoted by Asclepius, in Metaph. 28,26–7. Cf. Proclus El. Theol. prop. 12,
p. 14,18.
164
Asclepius in Metaph. 28,31–2 (prôton aition).
165
ibid. 148,10–11; 225,15–17; cf. 105,30–2 (where Ammonius is mentioned as ‘the philosopher’).
According to L.G. Westerink, Anonymous Prolegomena, p. 19, Ammonius’ interpretation of the
Aristotelian God is also reflected in Asclepius in Metaph. 151,15–32. I have the impression that this
text is not primarily concerned with Ammonius’ view on Aristotle’s God, but rather expounds
Ammonius’ own theology.
166
Asclepius in Metaph. 28,20–2; 103,3–4.
167
cf. Aristotle Phys. 2.3., 194b29–32.
168
Asclepius in Metaph. 233,25–31.
 The metaphysics of Ammonius son of Hermeias 237

hierarchy, namely matter. Aristotle’s matter is both being and non-being.169 It is

being, one is to understand, in so far as it is not nothing, and it is non-being in so
far as it is not anything in particular. This means, according to Asclepius, that for
Ammonius Aristotle himself accepted the transcendence of the One compared
to Being – ‘The One surmounts being (huperbebêke to hen to on)’: matter, as it is
non-being but not nothing, is the One, below all being. Now, the analogy of the
status of matter to that of the One is obvious. I refer to a text written by Ammonius
himself, where the different modes of non-being are enumerated: matter,
Ammonius writes, is non-being because of the absence of forms, and this implies
a similarity accompanied by dissimilarity (anomoios homoiotês) between matter
and the first principle.170 As a consequence, it is not too far-fetched, I think, to
infer from Asclepius’ text on matter that Ammonius interpreted Aristotle’s God
in an analogous way, as containing both being and a principle above being.
Moreover, in the case of Aristotelian matter, that which transcends Being is
called by Asclepius the One. Accordingly, Aristotle’s highest principle also is One.
This means a fortiori that it cannot be identical with the divine Intellect, since for
Ammonius Aristotle’s Intellect is not just a unitary self-thinking thought, but
contains within itself the plurality of Platonic Ideas.
In this last reference to Plato’s Ideas, it is again Ammonius’ intention to
harmonise Aristotle with Plato.171 Aristotle’s opposition to the theory of Ideas is
said to be directed, not against the existence of Ideas, but against those interpreters
of Plato who wrongly think that the Platonic Ideas are entities subsistent outside
the demiurgic Intellect. In Ammonius’ view both Plato and Aristotle hold that
the Ideas exist as demiurgic logoi within the divine Intellect.172 Asclepius
mentions four arguments in favour of this Neoplatonic interpretation of
Aristotle. In the first place he refers to Metaph. 12.10, 1075a12–17, where the
immanence and the transcendence of the Good in relation to the cosmos is
compared with the presence of the order of an army both in the general and in
the army itself.173 Secondly, at DA 3.4, 429a27–8 Aristotle calls the soul ‘the place
of forms (topos eidôn)’.174 At first sight it is not very clear how this text can be

169
cf. n. 69 above.
170
Ammonius in Int. 213,8–10; cf. Simplicius in Phys. 205, 8–15; Philoponus in Phys. 162,12–20.
171
Ammonius’ diligence in harmonising Plato and Aristotle on the subject of the doctrine of Ideas is
also mentioned by Zacharias Ammonius l.952.
172
Asclepius in Metaph. 44,32–7 (cf. K. Kremer, Der Metaphysikbegriff, 168); 69,17–27 (cf. Kremer, ibid.
185–6); 75,27–8; 167,14–34; 183,14–16; 233,38–40; 363,1–5; 393,34–394,2; 441,27–31; 442,1–2; cf.
K. Praechter, ‘Simplikios’, col. 211. To Proclus in Tim. I, 267,1 the Aristotelian God is still an ‘intellect
without plurality (nous aplêthuntos)’.
173
Asclepius in Metaph. 44,35–7.
174
ibid. 69,19–20; 167,29–31.
238 Aristotle Transformed

made relevant to the ontological status of Ideas. In my opinion the argument is

this: universal Ideas (the logoi) are present in God, in nature, and in the rational
soul. In God this presence is both cognitive and creative, in nature it is only
creative, in soul it is only cognitive.175 Now, if at the level of soul the logoi are
already contained within the unity of a ‘place’, this is true a fortiori for the divine
logoi. The third argument is Aristotle’s statement at DA 3.5, 430a19–20 (cf. 3.7,
431a1–2) that ‘the intellect in act is things’ (i.e. is identical with the objects of its
thought).176 Finally, Asclepius says, according to DA 3.5, 430a14–15 ‘the intellect
in potency is active, but the intellect in act is productive’ (sic).177 This interpretation
of Aristotle’s divine Intellect seems hardly compatible with the view that
Aristotle’s God is the Neoplatonic One, unless one accepts that Ammonius
implicitly ascribes to Aristotle the doctrine that the Intellect is subordinate to the
highest principle.

(c) The compatibility of Simplicius’ and Asclepius’ accounts

The hypothesis just mentioned (subordination of the Intellect to the One) is
necessary, I think, in order to prevent Asclepius from contradicting himself on
Ammonius’ interpretation of Aristotle’s theology. On the other hand there is
nothing in Simplicius’ texts on Ammonius to confirm it: if we rely on Simplicius,
Aristotle’s God for Ammonius simply was the divine Intellect. Therefore, the
conclusion that the two commentators contradict each other seems at first sight
inevitable. If we try to explain the contradiction, at least three possibilities may
be considered.
First of all, one might suppose that there was a difference between Ammonius’
esoteric exegesis of Aristotle within the school (represented by Asclepius) and
the monograph he published for a wider public (referred to by Simplicius). On
this hypothesis some influence of Christianity on the published work might
seem to be possible. However, one should not forget that Simplicius completely
agrees with the purport of Ammonius’ book, and, moreover, that he clearly
indicates the Neoplatonic character of Ammonius’ view on the causal relation
between Aristotle’s God and the world. This suggests that Simplicius regarded
Ammonius’ interpretation as fully in line with Neoplatonism, which would have
been impossible if there were any influence of Christian theism in it.

175
ibid. 81,2–4; 148,29–31.
176
ibid. 69,20; 167,31. M. Hayduck ad 69,20 also refers to DA 3.8, 431b21.
177
ibid. 69,20–1.
 The metaphysics of Ammonius son of Hermeias 239

Did Ammonius’ monograph perhaps represent his later position, a position

facilitating a convergence between Aristotelianism and Christianity by changing
in a theistic direction the original, Neoplatonically inspired interpretation of
Aristotle’s theology? I think that the objections just raised against the first
possibility hold against this one also.
The third possibility is that both Asclepius and Simplicius give an incomplete
account of the theory of Ammonius. If that is correct, the contradiction is only
apparent, and the two versions should be considered as two complementary
variants of the same theory. Both in Asclepius and in Simplicius, one might
suppose, Aristotle’s identification of the first principle and the Intellect is replaced
by an implicit subordination of the Intellect to the highest principle. In Asclepius,
as we have seen, this implication is rather obvious. In Simplicius it is not, but this
should not necessarily mean that it is not present at all. In my opinion the
difference between the two versions is indeed a matter of different emphasis. In
Asclepius, Ammonius’ ascription of a double causality to Aristotle’s God is
primarily related to the level of the supreme cause, the One. Consequently,
Asclepius neglects to mention explicitly that the same double causality belongs
likewise to the demiurgic Intellect. Moreover, it means that divine efficient
causality is styled as a process beyond thought and consciousness. In Simplicius,
on the other hand, Ammonius’ interpretation of Aristotle’s God is primarily
related to the level of divine Intellect. Hence Simplicius omits the fact that final
and efficient causality for Aristotle also belong to the supreme cause. I assume that
the difference between Asclepius and Simplicius is nothing but a shift of emphasis.
Maybe this shift itself can be explained. Asclepius is speaking as a commentator
of the Metaphysics, Simplicius as a commentator of the Physics. The distinction
between these two disciplines seems to offer, in the Alexandrian tradition, the
possibility of giving different emphases to Aristotle’s concept of God. In this
respect we have a very important piece of evidence in Elias’ in Categorias. Why,
Elias wonders, does Aristotle in his Metaphysics call the Intellect the very first
principle (hê prôtistê arkhê) and not, like Plato, the Good (to agathon)? The
answer is this: the Metaphysics is a theological treatise, but it immediately follows
the physical treatises.178 This explains why Aristotle does not ascend from the
Intellect to the Good in his Metaphysics: the Intellect is the more physical of
the two principles (hê phusikôtera arkhê). But the fact that in the Metaphysics the
Intellect is regarded as the ultimate principle does not mean that Aristotle was
not aware of the transcendence of the Good compared to the Intellect. The

178
ta phusiologika, broader than just the Physics.
240 Aristotle Transformed

beginning of the Nicomachean Ethics, a treatise which is further removed from

natural philosophy, shows that he was. In general, however, one can characterise,
according to Elias, Aristotle’s way of theological thinking as ‘theology from the
standpoint of natural philosophy (theologôn phusiologei)’.179
But how does this remark of Elias help us to understand the difference
between Asclepius and Simplicius? Would one not anticipate, relying simply on
Elias, that Asclepius, as a commentator of the Metaphysics, held the view that the
Intellect is the highest principle, rather than introducing the Good? Asclepius’
commentary proves that this conclusion would be erroneous. Therefore we
should concentrate on the following points in Elias’ account: (1) Aristotle knew
the difference between the Intellect and the Good; (2) he kept silent about the
Good in the Metaphysics all the same; (3) he could do so (a) because his
metaphysics (or theology) is limited to indicating the immediate transcendent
cause(s) of the physical world, and (b) because the divine Intellect is closer to the
physical world than the Good. The text of Elias shows that the Alexandrian
school conceived of Aristotle’s God as consisting of two ‘hypostases’, the Good
and the Intellect. Aristotle is said to have telescoped his concept of God in the
Metaphysics. This means that one can consider Aristotle’s God, according to one’s
point of view, either as the Neoplatonic Good or as the Neoplatonic divine
Intellect. When we apply this principle to the relation between Asclepius’ and
Simplicius’ versions of the Ammonian interpretation of Aristotle’s concept of
God, the result is as follows. At the metaphysical level (Asclepius) the Unmoved
Mover is primarily understood in its ultimate depth as the Good, the first origin
and the final cause of all reality. At the level of natural philosophy (Simplicius) the
Aristotelian God is primarily conceived as the demiurgic Intellect, which at the
same time is the final cause of cosmic movement. In this perspective the apparent
contrast between Asclepius and Simplicius gives no ground for the hypothesis
of a synchronic and/or diachronic duality in Ammonius’ interpretation of
Aristotle’s theology.
However, it remains somewhat curious that, in reporting Ammonius’ view on
Aristotle’s God, Simplicius does not say anything about the two ‘hypostases’. I
assume that this has to do with the fact that in Physica and in de Caelo belong to
natural philosophy. Simplicius was aware of the limits imposed on theological
speculation in natural philosophy.180 And in the very lines of in Physica where he
refers to Ammonius’ monograph he draws attention to the incompleteness of his

179
Elias in Cat. 120,23–30.
180
e.g. Simplicius in Phys. 148,22–4.
 The metaphysics of Ammonius son of Hermeias 241

account.181 More clarity – if any – should come from a complete study of

Simplicius’ own statements about Aristotle’s theology, scattered throughout his
commentaries. But I want to state right now that Simplicius too, speaking on his
own authority, credits Aristotle with the distinction between the divine Intellect
and the supreme principle, just like Elias.182 I think this fact considerably
increases the plausibility of my hypothesis that Ammonius already ascribed the
Neoplatonic distinction between the first two hypostases to Aristotle, despite
Simplicius’ silence about the distinction in his texts on Ammonius.

(d) Philoponus
The way in which both Simplicius and Asclepius deal with Ammonius’ ascription
of final and efficient causality to Aristotle’s God should have made clear (1) that
Ammonius’ theory has nothing to do with any attempt to ‘Christianise’ Aristotle
and (2) that it did not imply a theistic concept of God either. Since it was launched
by P. Tannery,183 the ‘Christian’ hypothesis has been accepted by several scholars.184
However, a number considerations should have raised suspicions about its
correctness. First, Simplicius would never have agreed with any theory that tried
to Christianise Aristotle. Secondly, Simplicius says clearly that the purpose of the
theory (in Ammonius’ mind also, we may add) was to harmonise Aristotle with
Plato.185 And this harmonisation was indisputably a concern of pagan
Neoplatonism, not of Christian philosophy.186 Thirdly, it should be remembered

181
ibid. 1363,10–12.
182
Simplicius in Cael. 485,19–22: ‘That Aristotle has the notion of something even above Intellect and
being is shown by his saying clearly, near the end of his book “On Prayer”, that God is either Intellect
or something even beyond Intellect’ (= Aristotle fr. 49 Rose; I adapt Ross’s translation).
183
P. Tannery, ‘Sur la période finale’, 226–7.
184
The ‘Christian interpretation’ has been accepted by L.G. Westerink, The Greek Commentaries on
Plato’s Phaedo, vol. 1: Olympiodorus, 24; M. Mahdi, ‘Alfarabi against Philoponus’, 237, n. 9; G.
Verbeke, ‘Some later Neoplatonic views on divine creation and the eternity of the world’, in D.J.
O’Meara (ed.), Neoplatonism and Christian Thought, Albany 1982, 46. It is also implied in a remark
by E. Ducci (‘Il to eon parmenideo nell’ interpretazione di Filopono’, Rassegna di Scienze Filosofiche
17, 1964, 272), who detects the Christian conviction of the author in the statement ‘God is the
efficient cause of all things (Deus causa efficiens omnium est)’ in ps.-Philoponus in Metaph. fol. 2rA
(Patrizzi). The opposite position is that Ammonius’ theory intended to harmonise Plato and
Aristotle; it was taken by H.D. Saffrey, ‘Le chretien Jean Philopon’, 400 n. 3; P. Merlan, ‘Ammonius
Hermiae, Zacharias Scholasticus and Boethius’, 201; A.C. Lloyd, ‘The Later Neoplatonists’, 317.
R. Vancourt, Les derniers commentateurs, 19, and É. Évrard, L’école d’Olympiodore et la composition
du ‘Commentaire à la Physique’ de Jean Philopon, diss., Liège 1957, 208–9 (and Westerink; cf. the
quotation p. 202) are of the opinion that the two positions are compatible.
185
cf. n. 137 above.
186
cf. e.g. A.C. Lloyd, ‘The later Neoplatonists’, 275; H.J. Blumenthal, ‘Neoplatonic elements in the de
Anima commentaries’, Phronesis 21, 1976, 65–7 (= Chapter 13 in this volume, pp. 306–8); I. Hadot,
Le problème du néoplatonisms alexandrin, 68, 73–6. In 529 ad Philoponus indirectly claims that the
242 Aristotle Transformed

that Ammonius does not question Aristotle’s belief in the eternity of the world.
Hence, even in the revised interpretation there remains a considerable distance
between divine causality in Aristotle and Christian creationism (although some
Christians, irrespective of Ammonius’ intentions, may have felt that an eternally
‘creating’ Aristotelian God was less un-Christian than the Mover usually ascribed
to Aristotle which is only a final cause). Finally – and in my opinion this is
conclusive – the fate of Ammonius’ theory in the philosophy of Philoponus
proves that it was regarded as closely related to pagan Neoplatonism.
One of the discontinuities between Philoponus’ earlier and later work, which
leads me to speak187 of Philoponus 1 and Philoponus 2, is precisely his about-
turn with regard to the interpretation of both Platonic and Aristotelian theology.
I must limit myself to a few essentials here. Philoponus 1, without mentioning
his master in this respect, accepts Ammonius’ view that, for Aristotle, God is
both the final and the efficient cause of the universe. He uses two types of
arguments. On the one hand, there are two passages in in de Generatione et
Corruptione where it is stated that Aristotle implicitly ascribes efficient causality
to ‘God’.188 On the other hand, Philoponus also finds explicit evidence in de
Generatione et Corruptione.189 As a result, it is a matter of course to Philoponus 1
that Aristotle’s God is also the efficient cause of the universe.190 The way in which
Philoponus understands this is very similar to what we found in Asclepius:
Aristotle’s God is the source of all reality, including matter,191 i.e. He is the One.
Just like Asclepius and Simplicius, Philoponus 1 believes that, for Aristotle, the
Platonic Ideas are logoi immanent in the demiurgic Intellect.192 His first piece of

theory that Aristotle’s God is an efficient cause as well (a theory which is implied in the harmonising
ascription of the doctrine of Ideas to Aristotle) is due to the fact that some recent philosophers were
ashamed about the disagreement between Plato and Aristotle (Philoponus Aet. 29,2–8; cf. 32,10–
13). (K. Praechter names the recent philosophers in question as ‘Ammonius Hermiae and his circle’,
‘Simplikios’, col. 211, but it escapes his attention that Philoponus 1 had been advocating Ammonius’
view himself). The underlying idea is this: the Neoplatonists try to harmonise Aristotle with Plato
out of shame over the discord within Greek philosophy compared to the consensus in Christian
doctrine. However, according to Elias in Cat. 123,2–3 Iamblichus was already of the opinion that
Aristotle and Plato did not contradict each other with regard to the doctrine of Ideas (cf. H.J.
Blumenthal, ‘Neoplatonic elements’, 66 (= p. 307 below)).
187
See Chapter 11 below.
188
Philoponus in GC 136,33–137,3; 152,23–153,2. According to E. Evrard, ‘Les convictions religieuses’,
p. 354, the first passage is redundant in its context: Philoponus, Evrard believes, goes especially out
of his way to expound his Christian view of Aristotle. I cannot enter into this any further now; the
mere reference to in GC 152,23–153,2 is sufficient, I think, to refute Evrard’s view.
189
GC 1.3, 318a1–5 (Philoponus in GC 50,1–5; cf. n. 143 above); GC 2.10, 336b31–4 (Philoponus in GC
297,15–24).
190
Philoponus in Phys. 298,6–10; 304,5–10.
191
ibid. 189,13–17, referring to Cael. 1.9, 279a27–30 (cf. p. 205 above) and Metaph. 12.7, 1072b13–14.
192
Philoponus in DA 37, 19–31.
 The metaphysics of Ammonius son of Hermeias 243

evidence is Metaph. 12.10, 1075a12–17, as in Simplicius and Asclepius.193 Two

further arguments (DA 3.4, 429a27–8 and 3.5, 430a19–20) are common to
Asclepius and Philoponus.194 Whereas Asclepius finds his fourth argument in
DA 3.5, 430a14–15, Philoponus may be referring to Metaph. 12.9, 1074b33–4,
where Aristotle, he says, identifies the divine Intellect’s self-contemplation
(heauton horôn) with its contemplation of reality (ta pragmata horâi).195 All this
means that the picture in Philoponus 1 is very much the same as in Asclepius. It
is most significant that Philoponus 2, wanting to dissociate himself publicly
from Alexandrian Neoplatonism and to present himself as a Christian
philosopher, immediately drops his master’s interpretation of Aristotle’s God.
Both in his de Aeternitate Mundi contra Proclum and in his in Analytica Posteriora
Philoponus argues that Aristotle rejected Plato’s theory of Ideas itself, not just an
erroneous interpretation which regarded the Ideas as substances outside the
demiurgic Intellect.196 If Aristotle only intended to criticise those Platonists who
hypostatised the Ideas outside the Demiurge, and not the very existence of the
Ideas, he would have made this clear, and he would have indicated which
interpretation of the theory of Ideas is the correct one.197 Instead, he calls the
Ideas mere ‘twittering (teretismata)’.198 Moreover, it is recorded that Aristotle was
already strongly opposing the doctrine of Ideas when Plato was still alive.199
Philoponus 2 no longer accepts the evidential value of the Aristotelian texts that
are adduced by those who try to harmonise Aristotle with Plato.200 The important
thing in all this is that Philoponus 2, by denying that Aristotle agreed with the
doctrine of Ideas, at the same time denies that Aristotle’s God is also the efficient
cause of the universe. Indeed, if Aristotle rejected the existence of Ideas ante res,
then his God cannot be anything but pure self-thinking thought. This means that
the divine Intellect to Aristotle is no longer a creative Intellect: an Intellect
without Ideas cannot be creative, it is merely the final cause of the universe.
In this way we arrive at an unexpected conclusion. Ammonius’ interpretation
of Aristotle’s concept of God is not only free of Christian influence, but is even

193
ibid. 37,20–6.
194
ibid. 37,26–8.
195
ibid. 37,28–31. Needless to say, Aristotle claims that God is pure self-consciousness.
196
Philoponus Aet. 26,24–32,13; in An. Post. 242,14–243,25.
197
Philoponus Aet. 29,8–30,6.
198
ibid. 30,8–12. The reference is to Aristotle An. Post. 1.22, 83a32–3, which is part of the lemma of
Philoponus in An. Post. 242,14–243,25.
199
Philoponus in An. Post. 243,20–1. This may be another hint by Philoponus at his own relation to
Ammonius as he wanted it to be interpreted after 529. Cf. Chapter 11 n. 181 below.
200
Philoponus in An. Post. 242,26–243,13. Three of the texts referred to are the same as in in DA
37,19–31: Metaph. 12,10, 1075a12–17; 12.9,1074b33–4 (?); DA 3.4, 429a27–8. The remaining
reference (in An. Post. 243,7) is perhaps to DA 3.8, 432a2 (intellect = form of forms).
244 Aristotle Transformed

regarded as an anti-Christian attempt to make pagan philosophy more harmonious

than it was. Philoponus 2 implicitly returns to the view that Aristotle’s God is
merely the final cause of the universe. This is a Christian reaction which denies the
harmony of Plato and Aristotle and rejects the idea of an eternal ‘creation’ implied
in the interpretation of Ammonius and Philoponus 1. Both to Ammonius and to
Philoponus 1 that idea was so obviously correct that they ascribed it to Aristotle as
well. To Philoponus 2, by contrast, ‘eternal creation’ has become something
inconceivable, a contradiction in terms. Consequently, Philoponus 2 can no longer
consider Aristotle’s eternal world to be created by the divine Intellect, just as,
conversely, he can no longer regard Plato’s created universe eternal a parte ante.

Ammonius and the problem of Alexandrian Neoplatonism

What does my analysis of the metaphysics of Ammonius contribute to the

discussion of ‘the problem of Alexandrian Neoplatonism’? Obviously, the result
of Mme. Hadot’s investigations on Hierocles and Simplicius are confirmed by
my findings on Ammonius, in so far as the Neoplatonic and non-Christian
character of Alexandrian philosophy is concerned. I do not think that Ammonius’
theology was essentially influenced by Christianity, nor that his interpretation of
Aristotle’s theology was inspired by Christian motives. His philosophy is no
return to pre-Plotinian Platonism either: it is clearly Neo-Platonism.
More delicate, however, is the question to what extent Ammonius remained
faithful in his mature years to the Neoplatonism of his master Proclus. Asclepius’
Commentary on the Metaphysics displays some characteristics which might
suggest that Ammonius tended to simplify the system: the henads disappear; the
Demiurge seems to be simply identified with the divine Intellect; there is not
much left of Proclus’ construction of innumerable triads; the articulation of the
intelligible world at levels between the divine Intellect and the sensible world has
been blurred. Moreover, Ammonius is inclined to remodel the hierarchy of
ontological levels into a dichotomy between the creative and the created. As a
result, the Intellect and the One are frequently taken together in the notion ‘God
(ho theos)’. Soul, on the other hand, which is actually intermediate between
intelligible (creative) and sensible (created) reality,201 is sometimes decoupled
from the first two levels of reality and considered as ‘caused by God’.202 All this

201
Asclepius in Metaph. 98,5–15; Ammonius in Cat. 37,17–18.
202
Asclepius in Metaph. 147,31–2.
 The metaphysics of Ammonius son of Hermeias 245

seems to mean that, with the appropriate qualifications, Praechter was not
completely wrong in ascribing to Ammonius ‘a tendency to retrogression’.203
But at this point another question arises: isn’t it possible that this simplification
is characteristic of Ammonius’ exegesis of Aristotle only, and that in his lost
commentaries on Plato he practised metaphysical speculation in a more
Athenian style? We know that in Athenian Neoplatonism the reading of Aristotle
was considered propaedeutic to the study of Plato.204 And we know also that
Ammonius, at some stage of his career, commented on the Gorgias and the
Theaetetus.205 Consequently, one cannot rule out a priori the possibility that he
reserved a fuller exposition of his Neoplatonic metaphysics for his commentaries
on Plato. If this is true, then the status of metaphysics in Ammonius’ commentaries
on Aristotle206 would seem comparable with the status of metaphysics in
Hierocles’ in Carmen Aureum and Simplicius’ in Epicteti Enchiridion. In all cases,
one might argue, the system is simplified and/or fragmented because of the
preparatory level of the texts and commentaries at issue.207 I think there can be
little doubt that this holds true for Ammonius’ commentaries on the Organon,
regardless of whether or not there was any fundamental difference between
Ammonius’ metaphysics in Asclepius’ commentary and in his commentaries on
Plato. Hence we must concentrate on Asclepius’ in Metaphysica: is what we have
there Ammonius’ last word in metaphysics? Or do we have solid reasons to
suppose that Asclepius too offers only a reduced version of Ammonius’
metaphysics? Obviously, any answer to this question must remain highly
speculative: one cannot compare existing with non-existing commentaries. But
in my opinion this does not necessarily mean that we are completely in the dark.
Let us first try to see what arguments can be adduced for affirming that
Ammonius’ Neoplatonism was not essentially different from contemporary

203
K. Praechter, ‘Richtungen und Schulen’, 145–6.
204
Marinus, Vita Procli XIII ; cf. I. Hadot, Le problème du néoplatonisme alexandrin, 149.
205
L.G. Westerink, Anonymous Prolegomena, p. xi. According to Westerink, The Greek Commentaries
on Plato’s Phaedo, vol. I: Olympiodorus, 83, Philoponus at in An. Post. 215,4–5 refers to a commentary
of Ammonius on the Phaedo. H.D. Saffrey, ‘Le chrétien Jean Philopon’, 406 n. 1, thinks the reference
is to the currently used school commentary; cf. W. Haase, ‘Ein vermeintliches Aristoteles-Fragment
bei Johannes Philoponos’, Synusia. Festgabe für W. Schadewaldt, Pfullingen 1965, 350 n. 47. Others
believe that Philoponus refers to a lost in Phaedonem of his own, e.g. A. Gudeman, ‘Ioannes
Philoponus’, in G. Wissowa and W. Kroll (eds), Realencyclopädie der klassischen Altertumswissenschaft
IX , 2, 1916, col. 1772 and C.M.K. MacLeod in G. Verbeke, Jean Philopon, Commentaire sur le De
Anima d’Aristote. Traduction de Guillaume de Moerbeke, Louvain and Paris 1966, xviii.
206
I think A.C. Lloyd, ‘The later Neoplatonists’, 315 is exaggerating somewhat when he says: ‘But when
the simplicity and hard-headedness of the Alexandrians’ Neoplatonism is contrasted with the
elaboration and “speculative” character of Athenian metaphysics, a simple but hard fact must be
remembered. We have not got their metaphysics.’
207
cf. above.
246 Aristotle Transformed

Athenian Neoplatonism. (1) The fact that Ammonius was a pupil of Proclus is
not sufficient in itself for supposing that he never departed from his master’s
system. Ammonius’ interpretation of Aristotle’s theology was different from
Proclus’; the doctrine of henads has disappeared from Asclepius’ in Metaphysica.
These examples show that Ammonius did not recoil from some originality. (2) A
better argument is, of course, the propaedeutic character of the study of Aristotle.
In principle the hypothesis that Ammonius may have expounded a fuller
‘Athenian’ system in his commentaries on Plato covers two possibilities. Either
one thinks of the Platonic commentaries as contemporary with or posterior to
the ‘Asclepian’ version of Ammonius’ metaphysics, or one considers them as
earlier. On the latter alternative Ammonius gave up the study of Plato in order to
concentrate on Aristotle. This would rather suggest that he moved away, to some
extent, from his Athenian background. However, Ammonius must have been
commenting on Plato until the end of his career, given the fact that Olympiodorus,
who was born probably after 495208 says he attended a lecture of Ammonius on
the Gorgias.209 But it is important to note that there is no mention of Ammonius
commenting on the Timaeus or the Parmenides. He is only known to have
lectured on more ‘elementary’ dialogues.210 It remains to be seen how far it is
possible to infer something more definite about the contents of Ammonius’
lectures on Plato on the basis of Olympiodorus’ commentaries. Meanwhile, one
cannot rule out a priori the possibility of a more elaborate system in these
Platonic commentaries. (3) A third argument is that in Asclepius’ in Metaphysica
the different ‘forces’ in the intelligible world are only referred to in a very vague
way,211 which might imply that it is in principle possible to give a more detailed
account of the structure of intelligible reality.
On the other hand there are several arguments for the assumption that
Ammonius’ later philosophy did indeed represent an intrinsic simplification of
Neoplatonism, in other words, that Asclepius gives a more or less complete
account of Ammonius’ system.
(1) The study of Aristotle’s Metaphysics is the final stage in the study of Aristotelian
philosophy.212 The goal (telos) of this philosophy is defined as raising oneself up to

208
L.G. Westerink, Anonymous Prolegomena, p. xiii.
209
Olympiodorus in Gorg. 199,8 Westerink; cf. L.G. Westerink, Anonymous Prolegomena, p. xi.
210
For the order of study of the Platonic dialogues cf. L.G. Westerink, op. cit., pp. xxxvii–xl; A.J.
Festugière, ‘L’ordre de lecture des dialogues de Platon aux Ve/VIe siècles’, Museum Helveticum 26,
1969, 281–96 (= A.J. Festugière, Études de philosophie grecque, Paris 1971, 535–50).
211
cf. above.
212
Ammonius in Cat. 6,17–20; cf. Philoponus in Cat. 6,3–16; Olympiodorus in Cat. 9,31–10,2; Elias in
Cat. 121,5–19; Simplicius in Cat. 6,15–18.
 The metaphysics of Ammonius son of Hermeias 247

the highest principle, Goodness-itself.213 The Commentary on the Metaphysics

contains a theology which derives everything from this principle and does not limit
itself to a metaphysical reference to the Intellect. Hence, one might argue, the
Commentary on the Metaphysics no longer represents a preparatory level of thinking.
On the contrary, it seems to aim at a radical and ultimate explanation of reality. In
his non-metaphysical commentaries, on the other hand, Ammonius sporadically
practises the so-called physical type of metaphysics. In this perspective, Neoplatonic
theology (in a reduced form) may be supposed to have its (or an) appropriate locus
in the Commentary on the Metaphysics. Consequently, there need not be a
fundamental difference between the metaphysical standpoints of this commentary
and of the Platonic commentaries respectively. Maybe, therefore, Ammonius had
not much more to say about metaphysics than what we read in Asclepius.
(2) A second argument for this conclusion might be that there is no great
difference, in Asclepius’ in Metaphysica, between the ways in which Ammonius
treats Plato’s and Aristotle’s theology respectively. Nothing indicates that, for
Ammonius, Plato is the metaphysician proper and that Aristotle’s metaphysics is
only preparatory. On the contrary, the attempt to ‘Neoplatonise’ Aristotle’s
theology rather points to an upgrading of Aristotle compared to Plato. The result
of this effort to bring both philosophers closer together is, so far as Plato’s
theology is concerned, a picture which is largely the same as that of Aristotle’s.
Plato’s Ideas too are logoi in the demiurgic Intellect.214 The Platonic Demiurge is
both the efficient and final cause of the cosmos: his goodness generates the
cosmos215 and is its goal (telos) as well.216 None the less, he is not the supreme
principle of reality. The Platonic Ideas are produced by the One,217 and, for Plato,
the One transcends being and form.218 Ammonius’ obvious desire to make Plato’s
and Aristotle’s theology converge does not suggest that he considered Plato as
the real thing and Aristotle merely as a propaedeutic.
(3) So far I have not said anything about one of the standard items in any
treatment of Ammonius, namely Damascius’ remark on Ammonius’ deal with
the Alexandrian bishop, to the effect that Ammonius made concessions in
exchange for financial compensation.219 H.D. Saffrey supposed, on the basis of

213
Ammonius in Cat. 6,10–16; cf. above n. 128.
214
Asclepius in Metaph. 44,32–5; 166,31–4; 167,24–6.
215
ibid. 21,18–21.
216
ibid. 52,21–4; cf. 55,25–6; 103,9–12.
217
ibid. 54,25–6.
218
ibid. 42,14–5.
219
Damascius Vita Isidori fr. 316, p. 251 Zintzen. According to L.G. Westerink, Anonymous
Prolegomena, pp. xii–xiii, Ammonius ‘even went through the formality of being baptised’. This
248 Aristotle Transformed

this remark, that Ammonius gave up teaching Platonic philosophy.220 However,

both Asclepius and Olympiodorus refer to Ammonius commenting on Plato.221
Therefore the agreement with the Church probably did not affect the teaching of
Plato. Moreover, the ecclesiastical authorities can hardly have had much greater
sympathy for Aristotle than they had for Plato.222 On the other hand, there is a
text of Simplicius implying that there was no more teaching of Plato in Alexandria
at the time Philoponus studied with Ammonius.223 Simplicius must be
exaggerating somewhat here, the message being probably that there was nothing
in Alexandria to compare with the Athenian exegesis of Plato’s great dialogues.
‘If nothing has survived of his (Ammonius’) interpretation of Plato, the reason is
no doubt that the work of the Athenian school was felt to be more important;
just as, on the other hand, Proclus’ commentaries on Aristotle have left hardly
any traces.’224 But why, one may ask, was Ammonius’ exegesis of Plato considered
inferior to that of the Athenians? The only possible answer seems to be that his
approach must have been more sober, less fantastic and less sophisticated than
Proclus’ or Damascius’. I think that this relative decrease of interest in Plato is the
other side of an increasing concentration on Aristotle. It remains uncertain
whether or not all this had anything to do with the agreement between
Ammonius and the Church; I feel rather sceptical about that. Probably
Ammonius was just not the type of the true Platonic hierophant by Athenian
standards. Instead, he developed a Neoplatonising exegesis of Aristotle.
(4) A further argument for the thesis that Ammonius’ metaphysics, compared
with Athenian Neoplatonism, underwent an intrinsic simplification, is that the
Neoplatonisation of Aristotle’s metaphysics is met by a corresponding

‘conversion’ was also accepted by P. Merlan, ‘Ammonius Hermiae, Zacharias Scholasticus and
Boethius’, 203 and apparently not excluded by A. Cameron, ‘The last days of the Academy at Athens’,
Proceedings of the Cambridge Philological Society 195 n.s. 15, 1969, 14–15. It is rejected e.g. by
É. Évrard, ‘Jean Philopon, son commentaire sur Nicomaque et ses rapports avec Ammonius’, Revue
des études grecques 78, 1965, 597–8; M. Mahdi, ‘Alfarabi against Philoponus’, 235 n. 6; L. Tarán,
Asclepius of Tralles, 10, n. 48; G.A. Lucchetta, ‘Aristotelismo e cristianesimo in Giovanni Filopono’,
Studia Patavina 25, 1978, 582. It is very unlikely indeed that the patriarch succeeded in altering
anything about the traditional paganism of the leading academic milieu in Alexandria. Cf. for this
R. Rémondon, ‘L’Egypte et la suprême résistance au christianisme (Ve-VIIe siècle)’, Bulletin de
l’Institut français d’archéologie orientale 51, 1952, 63–6.
220
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 400–1; likewise W. Wolska, La Topographie Chrétienne de
Cosmas Indicopleustès, Paris 1962, 149; M. Mahdi, ‘Alfarabi against Philoponus’, 234 n. 3; more
prudently I.P. Sheldon-Williams, ‘The Greek Christian Platonist tradition from the Cappadocians
to Maximus and Eriugena’, in A.H. Armstrong (ed.), The Cambridge History of Later Greek and Early
Medieval Philosophy, Cambridge 1970, 478.
221
cf. L.G. Westerink, Anonymous Prolegomena, p. xi; É. Évrard, L’école d’Olympiodore, 207–8.
222
S. Pines, ‘An Arabic summary of a lost work of John Philoponus’, Israel Oriental Studies 2, 1972, 338
n. 266.
223
Simplicius in Cael. 84,11–13; cf. Chapter 11 below.
224
L.G. Westerink, Anonymous Prolegomena, pp. xi–xii.
 The metaphysics of Ammonius son of Hermeias 249

‘Aristotelianisation’, if you will pardon the expression, of the Neoplatonic system.

Ammonius’ metaphysical system has a cosmological orientation. The criteria
according to which the ontological hierarchy is constructed pertain largely to
the Aristotelian concepts of matter and form.225 Once the principle is accepted
that the hierarchy of substances must primarily be conceived as a gradation
involving both matter and form,226 there is little room left for a differentiation of
intelligible reality at the level of pure form. Whereas in Proclus’ system the
sensible world is only the modest foothill of an impressive intelligible
superstructure, the centre of gravity in Ammonius’ system lies in cosmic reality.
Obviously this shift of interest constitutes a major difference between Alexandrian
and Athenian Neoplatonism. This need not mean that Ammonius came to reject
a great part of Proclus’ system. The few vague allusions in Asclepius to the
possibility of a more detailed elaboration of the picture of intelligible reality
rather seem to indicate that Ammonius was prepared in principle to unfold
some more elements of the ‘Athenian’ metaphysics if necessary, but that he felt no
particular inclination to do so. His preference for the study of Aristotle is in line
with this tendency.
(5) Finally, one should not forget that Philoponus 1, whose philosophy in all
probability is very close to that of his master, also fails to suggest that, for
Ammonius, the study of Aristotle’s metaphysics was just propaedeutic to the
study of Plato’s. It is generally accepted that the substance of the Aristotelian
commentaries of Philoponus (Philoponus 1) is based on Ammonius’ lectures,
even if the formula ‘from the lectures of Ammonius, son of Hermeias (ek tôn
sunousiôn Ammôniou tou Hermeiou)’ does not appear in all titles.227 Ammonius’
legacy was primarily Aristotelian and consisted of three main parts: logic, natural
philosophy (including psychology) and metaphysics (Asclepius). I cannot enter
here into a detailed comparison of the pictures of Plato’s philosophy in
Philoponus 1 and Asclepius respectively. Roughly speaking, there is a great
similarity in their approaches to the relation between Plato and Aristotle. In
Philoponus 1 too there are a few excursions into Platonic-Pythagorean
philosophy,228 but as in Asclepius they do not suggest that Ammonius was much
interested in a large-scale multiplication of entities in the manner of Proclus.
The conclusion from all this has to be, in my opinion, that with Ammonius
something like a specifically Alexandrian Neoplatonism emerges, characterised

225
cf. Chapter 11 below.
226
cf. Ammonius in Cat. 35,18–36,4.
227
There is no mention of Ammonius in the titles of in Categorias, in Physica and in Meteorologica.
228
e.g. Philoponus in DA 116,1–121,10 (on Plato’s psychogony).
250 Aristotle Transformed

by a tendency to simplify the complicated system of Proclus. Due to the loss of

Ammonius’ Platonic exegesis we cannot know to what extent this simplification
is only a feature of his exegesis of Aristotle. Yet the Aristotelian commentaries
themselves point rather in the opposite direction. Regardless of whether or not
what we have contains the essence of Ammonius’ metaphysics, this much is clear,
that the study of Aristotle became more important than the study of Plato: the
‘Neoplatonisation’ of Aristotle was Ammonius’ main achievement. Platonic
exegesis in the Athenian manner became less relevant to this purpose. Or the
other way around: Ammonius may have felt that it was useless to rival Proclus in
Platonic exegesis, and instead preferred to develop his Neoplatonic and
harmonising exegesis of Aristotle. This ‘Ammonian’ Neoplatonism is to be found
in Asclepius, of course, and also in Philoponus 1. Further investigation should
teach us how far Ammonian Neoplatonism survives in the philosophy of
Olympiodorus and, beyond, in the commentaries of Elias, David and Stephanus.
It is only if it does that we shall be fully entitled to call it ‘Alexandrian’
Neoplatonism.229

229
I thank H.J. Blumenthal, A.C. Lloyd and R.R.K. Sorabji for their helpful and stimulating comments
on earlier versions of this chapter.
 11

The development of Philoponus’ thought and

its chronology*
Koenraad Verrycken

Since the end of the nineteenth and the beginning of the twentieth century,
when the Berlin corpus of Greek commentaries on Aristotle was published,
historians of ancient philosophy have had at their disposal the materials for
studying not only the commentators’ exegetical merits, but also their own
philosophical views. But the availability of the texts has not yet resulted in
research to match their volume and historical importance.1 In particular, the
commentaries from the late fifth and sixth centuries ad have not (Simplicius’
writings apart) received anything like the attention they deserve. Two main
factors are responsible for this situation. In the first place, the commentaries
are usually not read as a whole: they are consulted in connection with certain
passages of Aristotle or in relation to a special problem. Consequently, scholars
are generally less interested in their autonomous philosophical background,
a background which clearly shows that the commentaries have their roots
in late Neoplatonism. In the second place few historians who study late
Neoplatonism for its own sake pay much attention to the Aristotelian
commentators, although they make an exception for Syrianus and Simplicius,
who both belong to the tradition of Athenian Neoplatonism. For this Athenian
philosophy based mainly on Platonic exegesis is, at least as a metaphysics, much
more important than the philosophy of Alexandria. Because of this neglect, the
study of late fifth- and sixth-century commentaries on Aristotle, taken as a
whole, is one of the areas where the task is not simply one of adding to existing
interpretations.

* This chapter summarises some historical and philological conclusions drawn from the presence of
two different systems in the philosophical works ascribed to Philoponus. I am preparing further
publications on the subject.
1
Richard Sorabji’s translation project, The Ancient Commentators on Aristotle, will remedy this.

251
252 Aristotle Transformed

The foundations of Alexandrian Aristotelian exegesis were established by a

pupil of Proclus, Ammonius son of Hermeias. Among Ammonius’ own pupils
were Asclepius, Philoponus and Olympiodorus. Olympiodorus himself is
considered the originator of the last group of Alexandrian commentaries: those
of Elias, David and Stephanus. What is noteworthy about Philoponus is that
besides his commentaries he wrote autonomous philosophical treatises of a
clearly Christian tendency as well as a great number of dogmatic works of
theology. This poses our main problem about Philoponus in its most elementary
form: what is the relation between the commentaries attributed to the author
and his Christian writings? The first attempt at a global solution was undertaken
in 1916 by A. Gudeman.2 According to Gudeman, Philoponus was not a Christian
by birth (so, originally, his name was not John). As a pagan, ‘John’ became a pupil
of Ammonius. In the first period of his creative career the author, still a pagan,
wrote his grammatical works, his commentaries on Aristotle and most of his
other non-theological works. About 520 ad, according to Gudeman, our
philosopher was converted to Christianity, and from then onwards his writings,
including the more philosophical ones, clearly show the Christian conviction of
the author. This picture of Philoponus’ career was and still is accepted by several
scholars.3 In 1953, however, É. Évrard argued convincingly that Philoponus’ in
Meteorologica (at least in its present form) is posterior to the de Aeternitate
Mundi contra Proclum.4 Moreover, Évrard believes, the presence of several
passages in the in Physica, arguing against the eternity of the world, proves that

2
A. Gudeman, ‘Ioannes Philoponus’, in G. Wissowa and W. Kroll (eds), Realencyclopädie der
klassischen Altertumswissenschaft IX , 2, 1916, cols. 1769–72.
3
See W. von Christ, W. Schmid and O. Staehlin, Geschichte der griechischen Literatur II , 2 (Handbuch
der Altertumswissenschaft VII , 2, 2), 6th ed. Munich 1924, 1067; M. Meyerhof,‘Joannes Grammatikos
(Philoponos) von Alexandrien und die arabische Medizin’, Mitteilungen des deutschen Instituts für
ägyptische Altertumskunde in Kairo 2, 1931, 2–3; id., ‘La fin de l’école d’Alexandrie d’après quelques
auteurs arabes’, Archeion 15, 1933, 3; B. Tatakis, La philosophie byzantine, Paris 1949, 39–40; O.
Perler, Patristische Philosophie (I.M. Bochenski (ed.), Bibliographische Einführungen in das Studium
der Philosophie no.18), Bern 1950, 36; L. Bréhier, La civilisation byzantine, Paris 1950, 357; M.V.
Anastos, ‘Aristotle and Cosmas Indicopleustes on the Void. A note on theology and science in the
sixth century’, Hellênika 4, 1953 (repr. in his Studies in Byzantine Intellectual History, London 1979),
48; K. Axelos, ‘Les lignes de force de la spiritualité byzantine’, Bulletin de l’Association Guillaume
Budé, 1957, 4; A. Abel, ‘La légende de Jean Philopon chez les Arabes’, Acta Orientalia Belgica
(Collection Correspondance d’Orient no. 10), Brussels 1966, 252; H. Dörrie, Platonica Minora,
Munich 1976, 187, n. 64; G. Reale, Storia della filosofia antica, vol. 4, Milan 1978, 694. The thesis that
Philoponus was a convert was already advanced by G. Cave, Scriptorum ecclesiasticorum historia
literaria a Christo nato usque ad Saeculum XIV, vol. 1, Oxford 1740, art. Johannes Philoponus, 567.
I have not been able to see E.G.T. Booth, ‘John Philoponos: Christian and Aristotelian conversion’,
Studia Patristica vol. 17, part 1, 1983, 407–11.
4
É. Évrard, ‘Les convictions religieuses de Jean Philopon et la date de son Commentaire aux
“Météorologiques” ’, Bulletin de l’Académie royale de Belgique, Classe des lettres, sciences morales et
politiques, Série 5, 39, 1953, 303, 339, 345.
 The development of Philoponus’ thought and its chronology 253

by 517 ad Philoponus was already a Christian.5 All this means that Gudeman’s
thesis, according to which Philoponus’ Aristotelian commentaries were written
by a non-Christian author, has to be rejected: Philoponus is not a convert, he was
always a Christian.6 Since Évrard’s and Saffrey’s7 articles this ‘unitarian’ approach
to Philoponus’ philosophical works has become predominant.8 Nevertheless,
Évrard was aware of some of the problems arising from his views: for instance,
what are we to think about Philoponus’ affirmation of the pre-existence of the
human rational soul?9 And how are we to reconcile, within the limits of one and
the same commentary (in Physica), arguments against the eternity of the world
and the acceptance of Aristotle’s fifth essence and, consequently, of the eternity
of heaven?10 Évrard’s solutions to such problems are basically of two types. On
the one hand, doctrines such as that of a rational soul which is eternal a parte
ante are said not to be necessarily incompatible with Christianity (Évrard, rather
arbitrarily, suggests that Philoponus may have been influenced in this matter by
Origen).11 On the other hand, passages which imply an affirmation of the eternity
of the world are supposed to have only an exegetical significance: Philoponus

5
ibid., 352–4.
6
ibid., 345–7, 356–7.
7
H.D. Saffrey, ‘Le chrétien Jean Philopon et la survivance de l’école d’Alexandrie au VIe siècle’, Revue
des études grecques 67, 1954, 402, 408.
8
See P. Joannou, ‘Christliche Metaphysik in Byzanz, Johannes Philoponos (VI . Jh.)’, Wissenschaftliche
Zeitschrift der Martin-Luther-Universität Halle-Wittenberg 10, 1961, 1389; id., ‘Le premier essai
chrétien d’une philosophie systématique, Jean Philopon’, Studia Patristica vol. 5, 1962, 508; L.G.
Westerink, Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962, p. xii (modified);
C.M.K. MacLeod, ‘Jean Philopon’, Commentaire au De Intellectu. Traduction latine de Guillaume de
Moerbeke, diss. Louvain 1964, vol. 1, 15–16; J.L. Kraemer, ‘A lost passage from Philoponus’ Contra
Aristotelem in Arabic translation’, Journal of the American Oriental Society 85, 1965, 323; G.A.
Lucchetta, ‘Aristotelismo e cristianesimo in Giovanni Filopono’, Studia Patavina 25, 1978, 583–4; R.
Sorabji, ‘John Philoponus’, in R. Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science,
London and Ithaca NY 1987, 4–5. According to some authors the issue of a possible conversion
remains open: W. Wieland, ‘Die Ewigkeit der Welt (Der Streit zwischen Joannes Philoponus und
Simplicius)’, in D. Heinrich, W. Schulz and K.H. Volkmann-Schluck (eds), Die Gegenwart der
Griechen im neueren Denken. Festschrift H.G. Gadamer, Tübingen 1960, 294; S. Sambursky, The
Physical World of Late Antiquity, London 1962, 155 (but in his article ‘Philoponus, John’, in
P. Edwards (ed.), The Encyclopedia of Philosophy, vol. 6, New York and London 1967, 156, Sambursky
thinks that Philoponus is likely to have been a Christian by birth); W. Böhm, Johannes Philoponos,
Grammatikos von Alexandrien. Christliche Naturwissenschaft im Ausklang der Antike, Vorläufer der
modernen Physik, Wissenschaft und Bibel. Ausgewählte Schriften, Munich, Paderborn, Vienna 1967,
27; ‘The Greek Christian Platonist Tradition from the Cappadocians to Maximus and Eriugena’, in
A.H. Armstrong (ed.), The Cambridge History of Later Greek and Early Mediaeval Philosophy,
Cambridge 1970, 478–9.
9
É. Évrard, ‘Les convictions religieuses’, 349–50; e.g. Philoponus, in DA 16,10–26; in Phys. 128,22–31
(soul versus hê genêtê kai phthartê ousia).
10
É. Évrard, op. cit., 324–5 gives the following instances of Philoponus’ acceptance of Aristotle’s
theory of aether in the in Physica: in Phys. 9,23–10,2; 219,19–22; 220,20–5; 340,31. One can add: in
Phys. 15,29–30; 16,2; 16,8; 156,10–12; 438,9–10. See further n. 60 below (eternity of heaven).
11
É. Évrard, op. cit., 350.
254 Aristotle Transformed

could not always expound or even indicate his own point of view12 when
expounding Aristotle’s.
The position I should like to defend is to some extent intermediate between
that of Gudeman and that of Évrard. I think Évrard is right in rejecting the
hypothesis of Philoponus’ conversion. But I also think Gudeman was right in
assuming – more or less conjecturally – a duality in Philoponus’ philosophical
work. Both Gudeman and Évrard, however, pose the problem wrongly in terms
of ‘religious conviction’ only. If Philoponus did not develop a Christian
philosophy in his first philosophical period, that does not show that he must
have been a pagan at that time. And if he was born a Christian, that does not
establish that his philosophy must always have been Christian in character.
Philosophy is one thing, religion another. In my opinion, the problem should
first be posed on the purely philosophical level: what does the author say? Only
afterwards can one try to ‘project’ the results of the philosophical analysis onto
the levels of biography and psychology. This is the method I employ. To start
with, I shall outline very briefly the main characteristics of the philosophical
systems of ‘Philoponus 1’ and ‘Philoponus 2’, as I shall call them. Then I shall try
to piece together something of what can reasonably be said about Philoponus’
biography. Thirdly, I shall propose the first sketch of a new solution to the
problem of the chronology of the author’s Aristotelian commentaries. I shall
finish with some remarks on the development of Philoponus 2.

Philoponus 1 and Philoponus 2

The philosophical system of Philoponus 1 (i.e. of Philoponus before 529 ad ) can

be labelled an ‘Alexandrian Neoplatonism’. In theory it does not accept a personal
God as supreme, but is instead Neoplatonic in character, and like orthodox
Neoplatonism, it postulates above the divine Intellect a still higher principle of
reality, called the One, or the Good.13 Creation is a necessary, not a deliberate,
process.14 Typical of the Alexandrian form of Neoplatonism (Ammonius,
Asclepius, Philoponus) is the thesis that Aristotle’s God is the cause, both final

12
ibid., 355, n. 3: ‘The brief allusions which we find in the commentaries to the eternity of the world
should not give us pause. A commentator is inevitably led to repeat some theses of the author he is
expounding, without being able to criticise them every time. What clearly proves that they have
scarcely any importance is that they are found in the Physics commentary (e.g. 490,14), a work
which, as already seen, attacks the thesis of the eternity of time, of movement and of the world.’
13
e.g. Philoponus in Phys. 22,13–15; 163,2–12; 187,6–9; in GC 296,19–21; in DA 119,22–4.
14
Philoponus in Cat. 145,8–11.
 The development of Philoponus’ thought and its chronology 255

and efficient, of the existence of the universe,15 not merely the final cause of its
motion and order. Aristotle’s efficient causation is supposed to have worked
through the presence of Platonic Ideas as creative logoi in the demiurgic
Intellect.16 In the ontology of the early Philoponus being consists of a hierarchy
of substances (ousiai).17 Substances are ranked according to a list of properties:
proximity to pure form, degree of actuality, activity, unity, and identity. The basic
relations among the members of this ontological hierarchy are participation,
imitation and mediation. All this results in the following structure of being:
the One (above being); the demiurgic Intellect; transcendent intellects;18
undescended rational soul; the heavens (the celestial spheres have rational
souls);19 sublunary life (rational, irrational, vegetative); the four elements; three-
dimensionally extended substance; form (eidos); matter (below being). Evidently
this structure had no beginning in time, but is eternal: the sensible world, with
its different levels, has being to a lesser degree than pure form, but has never
lacked being. It is a necessary part of the totality. All levels of this totality possess
different degrees of substantiality, and imitate with decreasing perfection the
fullness of being which exists in the demiurgic Intellect.
The system of Philoponus 2 (after 529) is no longer Neoplatonic. First of all, it
clearly has the concept of a supreme personal God: the creative Intellect is the
highest, not the second highest, principle. Creation is no longer a necessary
process: it is the result of God’s free decision.20 Aristotle is said now to have
rejected Plato’s theory of Ideas.21 This means that the Aristotelian God is again
considered merely as the final cause of the world’s motion and order. The
ontology of the later Philoponus has become dualistic instead of hierarchical:
God is independent of the world, the world is no longer divine. This ‘secularisation’
of the world applies to intellects and souls as well as to sensible reality. Philoponus
2 argues at length against the eternity of movement, time and the world, and
against Aristotle’s theory of the eternal fifth element as the matter of the heavens.
He even goes so far as to deny the possibility of an eternal matter.22

15
Philoponus in GC 136,33–137,3; 152,23–153,2. See Chapter 10 above.
16
Philoponus in DA 37,19–31.
17
Philoponus in Cat. 49,23–51,21 (cf. Ammonius in Cat. 35,18–36,4); 67,7–10; in de Intell. 22,21–5
(=Sophonias in DA 126,17–19).
18
Philoponus in DA 35,1–2; in de Intell. 84,63–9; cf. Ammonius in Int. 248,17–31; 250,34–251,1.
19
Philoponus in DA 260,18–25.
20
Philoponus Aet. 78,11–17; 566,4–10.
21
Philoponus Aet. 26,20–32,13; in An. Post 242,14–243,25.
22
Philoponus Aet. 458,5–7; 469,6–10.
256 Aristotle Transformed

The philosophical career of John Philoponus

There seems no reason to doubt that our philosopher had the Christian name
‘John’ right from his birth, and that he was born in Alexandria. That the name
‘Philoponus’ refers to his (temporary) membership of a group of philoponoi
(some kind of militant Christian brotherhood) is not impossible, but by no
means certain.23 Probably he was born about 490,24 perhaps a few years earlier.
He studied ‘philology’ first, and turned to the study of philosophy only afterwards.
That could be the reason why Simplicius calls him a late learner (opsimathês) in
philosophical matters.25 When did this transition to philosophy take place?
Saffrey’s proposal (c. 510) is quite likely.26
What exactly was Philoponus’ relation with his teacher Ammonius? According
to Saffrey, Philoponus became an ‘assistant’ of Ammonius: his function was that
of an ‘official editor’ of the master’s commentaries. For fear of the ecclesiastical
authorities in Alexandria, Saffrey thinks, Ammonius used his Christian assistant
to cover him, not daring to publish his commentaries under his own name.27
There are, however, some objections to this view. First of all, nothing proves that
Philoponus’ project to edit a series of Ammonius’ commentaries was inspired or
even supervised by Ammonius himself. In the case of the Commentary on
Nicomachus’ Introductio Arithmetica, Philoponus seems to have based his version
not on notes he had made himself from the voice of Ammonius (apo phônês
Ammôniou), but on Asclepius’ redaction of Ammonius’ lectures.28 And one
cannot rule out the possibility of a similar situation for (some of) the
commentaries on Aristotle. The in Categorias and the in Analytica Priora ascribed

23
The suggestion was first made by S. Pétridès, ‘Spoudaei et Philopones’, Echos d’Orient 7, 1904, 343–4;
still hesitating: J. Maspéro, Histoire des patriarches d’Alexandrie depuis la mort de l’empereur Anastase
jusqu’à la réconciliation des églises jacobites (518–616), Paris 1923, 197–8, n. 4 and M. Meyerhof,
‘Joannes Grammatikos’, 2–3; affirmative: e.g. H.D. Saffrey, ‘Le chrétien Jean Philopon’, 403–4; P.
Joannou, ‘Christliche Metaphysik’, 1389; id., ‘Le premier essai chrétien’, 508; G.A. Lucchetta,
‘Aristotelismo e cristianesimo’, 584; M. Wolff, Geschichte der Impetustheorie. Untersuchungen zum
Ursprung der klassischen Mechanik, Frankfurt a. M. 1978, 104–13; prudent: e.g. C.M.K. MacLeod,
Jean Philopon, vol. I, 13–14; W. Böhm, Johannes Philoponos, 26–7; sceptical: A. Segonds, ‘Jean
Philopon, Traité de l’astrolabe, Paris 1981, 10–11; H.J. Blumenthal, ‘John Philoponus: Alexandrian
Platonist?’, Hermes 114, 1986, 317–18.
24
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 403.
25
Simplicius in Cael. 159,3; 159,7; in Phys. 1133,10; cf. in Cael. 29,7–8; 140,5.
26
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 403.
27
ibid., 408; cf. M. Mahdi, ‘Alfarabi against Philoponus’, Journal of Near Eastern Studies 26, 1967, 234–5.
28
L.G. Westerink, ‘Deux commentaires sur Nicomaque: Asclépius et Jean Philopon’, Revue des études
grecques 77, 1964, 530 (reprinted in his Texts and Studies in Neoplatonism and Byzantine Literature.
Collected Papers, Amsterdam 1980, 108); L. Tarán, Asclepius of Tralles, Commentary to Nicomachus’
Introduction to Arithmetic, Transactions of the American Philosophical Society, n.s. 59, 4, Philadelphia
1969, 10–13.
 The development of Philoponus’ thought and its chronology 257

by the modern editors to Ammonius are in fact students’ redactions comparable

to Asclepius’ in Nicomachi Introductionem Arithmeticam and in Metaphysica.29
So Philoponus may have made his own versions of in Categorias and in Analytica
Priora in the same way as he probably wrote his in Nicomachi Introductionem
Arithmeticam. All this points in a direction other than Saffrey’s hypothesis. One
rather gets the impression that it was not Philoponus, but pupils like Asclepius
who first received or took over the task of ‘officially editing’ Ammonius’ lectures.
But from a certain moment onwards (when? how?) Philoponus, though only a
philologist (grammatikos), got himself recognised, or appointed himself as the
main philosophical heir to Ammonius, and started the publication project all
over again, beginning with his lost *in Isagogen. None of this is to deny that
Philoponus was a pupil of Ammonius, and that he could rely on his own notes
for (some of) his Aristotelian commentaries. A second reason for doubting
Saffrey’s view on the relation between Philoponus and his master is that in the in
Physica Philoponus openly criticises the doctrine of the eternity of the world, a
doctrine which we know Ammonius adhered to without any hesitation.30 He
even goes so far as to reject tenets which he explicitly ascribes to Ammonius.31
This is not exactly the type of behaviour Ammonius would have expected from a
mere editor of his commentaries.32 In the third place, one wonders why
Ammonius, who had agreed with the Alexandrian bishop, according to Saffrey,
to teach only Aristotle, not Plato,33 would still have needed to hide behind a
Christian collaborator when publishing precisely his Aristotelian commentaries.
In my opinion Philoponus, after studying for some years under Ammonius,
somehow came to be the leading figure de facto of the Ammonian school,
without, however, at first teaching philosophy himself.34 The only absolute
chronological indication concerning this period is the date to be found in the in

29
The in de Interpretatione was written by Ammonius himself; cf. L.G. Westerink, Anonymous
Prolegomena, p. xi; id., The Greek Commentaries on Plato’s Phaedo, vol. 1: Olympiodorus, Amsterdam,
Oxford, New York 1976, 19; L. Tarán, Asclepius of Tralles, 14.
30
Ammonius’ eternalism is the main subject of Zacharias Scholasticus’ dialogue Ammonius.
31
In the relevant passages Ammonius is called ho philosophos: Philoponus in An. Post. 47,24–48,11;
158,7–11; 177,19–29; 258,26–30; 260,29–261,8; in Phys. 583,13–585,4 (Coroll. de loco).
32
cf. E. Evrard, L’école d’Olympiodore et la composition du ‘Commentaire à la Physique’ de Jean Philopon,
diss. Liège 1957, 213–14. I thank É. Évrard for allowing me to read his unpublished works Philopon,
Contre Aristote. Livre premier, Liège 1943, and L’école d’Olympiodore.
33
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 400–1.
34
According to A. Gudeman, ‘Ioannes Philoponus’, col. 1769, Philoponus stopped teaching after 529.
As we shall see, exactly the opposite may have been the case. After Ammonius died or retired, the
mathematician Eutocius seems to have held the chair of philosophy de iure; cf. L.G. Westerink, ‘Elias
on the Prior Analytics’, Mnemosyne, Series IV, 14, 1961, 129, 131 (=Collected Papers 62, 64); id.,
Anonymous Prolegomena, p. xiii; A. Cameron, ‘The last days of the Academy at Athens’, Proceedings
of the Cambridge Philological Society 195, n.s. 15, 1969, 26.
258 Aristotle Transformed

Physica: 10 May 517.35 We do not know whether Ammonius was still active or
even alive at that time. Given that he was born between 435 and 445,36 he would
have been at least 72 years old in 517. It is not very likely that his active career
lasted a long time after that date. And it is quite possible that by then others
already controlled things in the Alexandrian school. Consequently, the date in
the in Physica probably refers to Philoponus’ own redaction of this commentary
(the first redaction, as we shall see), rather than to lectures of Ammonius at that
time. On the other hand in Physica was not Philoponus’ first commentary: at
least *in Isagogen and in Categorias were written earlier.37 Hence we can assume
that in 517 our philosopher had been involved in the publication project for
some time. However, is it not a contradiction to say that Philoponus was a
Christian by birth and at the same time that he became the most important
public spokesman of what proves to be Ammonius’ Neoplatonic system? I do not
think so. It seems quite conceivable that under the influence of Ammonius the
Christian John himself became a Neoplatonic philosopher, and that his faith was
more or less superseded by Neoplatonism. After all, would this be the only case
of a young adult losing or changing his conviction? Moreover, Zacharias of
Mytilene tells us that Ammonius had just such a reputation of dislodging young
Christians from their faith.38 If it is true that Philoponus drifted away from his
Christian background (which should not mean that there are absolutely no
traces of Christianity in his early philosophical works), then it is also true that
his later transition to Christian philosophy was not the result of a conversion,
but rather a return (on the level of doctrine) to Christianity.39

35
Philoponus in Phys. 703,16–17.
36
L.G. Westerink, Anonymous Prolegomena, p. x.
37
cf. nn. 104–105 below.
38
Zacharias Ammonius ll.27–32 Colonna: ‘Well then, tell me: how is his (=Ammonius’) school
(phrontistêrion) and the assembly of his hearers? Is the school still being frequented by honourable
youths, pure in spirit? Indeed, “I am stricken with mortal fear”: I am anxious that he is filling the
youths with his nonsense. For he is clever in ruining their souls by removing them from God and
the truth (deinos gar ho anêr diaphtheirai neôn psukhas, aphistôn theou te kai tês alêtheias)’. In the
‘Hypothesis’ of the dialogue (p. 93 ll.1–2 Colonna) Zacharias likewise refers to Ammonius’ skill in
‘converting’ young Christians to ‘Hellenism’: Phoitêtês tis Ammôniou, tou dêthen philosophou,
genomenos kai êrema pros Hellênismon apoklinas. . . . This ‘slight inclination’ to Hellenic philosophy
was to become considerably stronger in Philoponus’ case.
39
According to Zacharias, Ammonius ll.998–1002, Ammonius prevented some of the youths he had
‘converted’ from hearing the arguments of the Christians: they might be seduced to become
Christians again (. . . hôs an . . . mê . . . khristianizein authis anapeistheien)! I must dissent here from
Mrs Minniti Colonna, who translates: ‘for fear . . . that they . . . might let themselves be convinced to
profess Christianity thereafter’ (p. 177). I think authis has its full value here (‘again’). This is made
clear by ll.999–1000: the students at issue are ‘Hellenised’ Christians (cf. the formula ‘fill with his
nonsense’, repeated from l.30).
 The development of Philoponus’ thought and its chronology 259

Towards the end of the 520s there is a radical change, a rupture in Philoponus’
thought. One has only to compare in de Anima with de Aeternitate Mundi contra
Proclum, written in 529,40 to realise the depth of the gap between Philoponus 1
and Philoponus 2. Why is this? By his editions of Ammonius’ commentaries
Philoponus had declared himself the new coryphaeus of the Alexandrian school.
Why, then, did he dissociate himself radically from pagan philosophy in 529?
According to Saffrey it was the author’s intention to save the Alexandrian school
by means of a Christian manifesto from being closed down like the Athenian
school.41 But this is not very likely. First of all, one should not forget that the
Alexandrian school as such was not really ‘Christianised’. Olympiodorus
undoubtedly was a pagan scholarch.42 And the so-called ‘Christian’ members of
the Olympiodorus school (Elias, David, Stephanus) did not oppose the traditional
doctrine of the eternity of the world.43 Moreover, at the end of the 530s
Philoponus disappeared from the philosophical scene, a rather strange fate for
the ‘saviour’ of the Alexandrian philosophical school. Everything suggests that
within the school Philoponus’ recantation on the eternity of the world was
essentially a one-man affair. What were his motives? Roughly speaking, there are
two possibilities. The first is that the Christian Philoponus became conscious of
the true philosophical implications of the faith of his youth. On this hypothesis
our author renounced his philosophical past out of honest conviction. The
second is that he acted against his own conviction, out of opportunism. I feel
more inclined to the latter than to the former of these interpretations, although
the evidence may not seem to be entirely conclusive. Probably we would be in a
better position if we still had the beginning (and end) of de Aeternitate Mundi
contra Proclum. The issue is not in itself crucial to my analysis, since whatever
biographical interpretation one prefers to give to the rupture in Philoponus’
philosophical thought, that duality itself just remains what it is. So I shall
postpone further discussion until later on.

40
Philoponus Aet. 579,14–15.
41
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 408–9; followed e.g. by A. Tuilier, ‘La tradition
aristotélicienne à Byzance des origines au VIIe siècle. La formation de la scolastique byzantine’,
Association Guillaume Budé. Actes du Congrès de Lyon 8–13 Septembre 1958, Paris 1960, 190–1; W.
Wieland, ‘Die Ewigkeit der Welt’, 315 n. 24; C.M.K. MacLeod, Jean Philopon, vol. 1, 38; J.L. Kraemer,
‘A lost passage’, 322; M. Mahdi, ‘Alfarabi against Philoponus’, 235; M.V. Anastos, ‘Byzantine
philosophy’, in P. Edwards (ed.), The Encyclopedia of Philosophy, vol. 1, 437; P. Lemerle, Le premier
humanisme byzantin. Notes et remarques sur enseignement et culture à Byzance des origines au Xe
siècle, Paris 1971, 24; G.A. Lucchetta, ‘Aristotelismo e cristianesimo’, 582–3.
42
L.G. Westerink, Anonymous Prolegomena, pp. xiii–xx; A Cameron, ‘The end of the ancient
universities’, Cahiers d’histoire mondiale 10, 1966, 671; id., ‘The last days of the Academy’, 9, 26–7.
43
L.G. Westerink, Anonymous Prolegomena, pp. xx–xxiv; A. Cameron, ‘The last days of the Academy’,
27; G. Verbeke, ‘Some later Neoplatonic views on divine creation and the eternity of the world’, in
D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Albany 1982, 46.
260 Aristotle Transformed

It is not unlikely that from 529 onwards Philoponus taught philosophy de

facto for some time, although de iure he was only a grammatikos, i.e. a professor
of philology.44 In this period he may have made the revised version of the in
Physica as well as the present redaction of the in Analytica Posteriora and the in
Meteorologica. But here again evidence is very scanty. Gudeman indicates only
two passages which might prove that the texts at issue belong to an oral
exposition.45 However, it is striking that both passages occur in the in
Meteorologica, a work which everyone now dates after 529. Moreover, it is
possible to complete Gudeman’s data.46 Characteristic of many passages in the in
Meteorologica are the strong self-consciousness of its author and the frequency
of references to his own points of view and writings.47 These characteristics are
also present in the texts I take to represent the second redaction of the in
Physica,48 whereas they are far less prominent in e.g. in de Anima or in de
Generatione et Corruptione. I am inclined to believe that both the present version
of the in Meteorologica and the texts which Philoponus 2 inserted in the original
redaction of the in Physica were composed in connection with the actual
teaching activity of their author after 529. As to the chronological limits of this
activity, it seems to have started with the publication of de Aeternitate Mundi
contra Proclum. First of all, this treatise is clearly the programmatic work by
which Philoponus 2 announces himself as a Christian philosopher. Besides, not
only the in Meteorologica but probably also the second redaction of the in Physica
was written after de Aeternitate Mundi contra Proclum.49 There are reasons to
suppose that towards the end of the 530s Philoponus stopped teaching
philosophy. With the publication of a third, non-polemical work about the
creation of the world (probably between 535 and 540)50 we seem to have reached

44
For this meaning of the term grammatikos (against M. Wolff, Geschichte der Impetustheorie, 108–9)
see L.G. Westerink, Anonymous Prolegomena, p. xiii; A. Segonds, Jean Philopon, Traité de l’astrolabe,
10; cf. already P. Tannery, ‘Sur la période finale de la philosophie grecque’, in P. Tannery, Mémoires
Scientifiques, vol. 7, Toulouse and Paris, 1925, 227 n. 2.
45
A. Gudeman, ‘Ioannes Philoponus’, col. 1766: Philoponus in Meteor. 53,26–7 and 71,3–4.
46
See Philoponus in Meteor. 63,8–9: ‘we shall hear the philosopher saying . . .’; and probably also
118,25–6: ‘and that is why we too should terminate this discussion here and move on to the next
issue, making a new start.’
47
For these references to his other writings see M. Hayduck, Index nominum, CAG 14.1 s.v.
‘Philoponos’ (153–4) and É. Évrard, ‘Les convictions religieuses’, 339–45. Self-confident utterances
occur in the first person singular (e.g. in Meteor. 8,13; 9,8; 15,23; 22,8; 43,7; 75,30; 79,31; 110,13 . . .)
and in the first person plural (e.g. in Meteor. 23,22; 43,37; 47,30; 48,14; 61,16; 77,23; 78,31; 82,17;
86,18–19; 96,22. . .).
48
cf. below. The frequent criticism of Ammonius in the in Analytica Posteriora (cf. n. 31 above) is
probably another striking example of the self-confidence of Philoponus 2.
49
cf. below.
50
cf. below.
 The development of Philoponus’ thought and its chronology 261

the end of the ‘philosophy of Philoponus 2’ strictly speaking, or at least of its

exposition in autonomous treatises.51 On the other hand, Olympiodorus’
appointment to the chair of Philosophy in succession to Ammonius and Eutocius
has to be set in the same period at the latest, if not earlier.52 So we can affirm,
without great risk, that Philoponus’ second philosophical period ends after
Olympiodorus’ appointment as a scholarch. We do not know whether it was a
long time afterwards. But it is reasonable to suppose that the transition to an
anti-eternalist doctrine of creation had isolated Philoponus within the
Alexandrian school. At a certain moment Eutocius, perhaps a rather neutral
figure, was succeeded by Olympiodorus, overtly a pagan. This indicates that,
maybe from as early as the beginning of the 530s onwards, opportunities for
pagan philosophy must have opened up again, and that Philoponus’ Christian
alternative was soon threatened or superseded by a new impulse of the
Ammonian tradition. Undoubtedly, under Eutocius also, the Alexandrian school
remained essentially true as an institute to the spirit of Ammonius. Hellenic
tradition in Alexandria recovered its full rights in the person of Olympiodorus.
I assume that this meant, in one way or another, the end of Philoponus’
philosophical role: the ‘traditionalists’ eventually succeeded in eliminating him
as a philosopher. Philoponus retired from ‘academic philosophy’,53 which does
not necessarily mean that he gave up his chair of philology. The rest of his life
was to be devoted to dogmatical theology.

The chronology of Philoponus’ commentaries on Aristotle

So far hardly anything has been said on the order of writing of the commentaries.
But everybody now accepts two basic facts: that ‘the’ in Physica was written in or

51
Philoponus Opif. 1,6–17; H.D. Saffrey, ‘Le chrétien Jean Philopon’, 408: ‘After the 530s our John
appears to follow a theological rather than a philosophical career. He does not seem to play any
further role in the university.’
52
On Olympiodorus’ start as a scholarch: L.G. Westerink, ‘Elias on the Prior Analytics’, 128, 131
(Collected Papers, 61, 64), where it is stated that Olympiodorus must still have been relatively young;
id., Anonymous Prolegomena, p. xiii: ‘His appointment to the chair of philosophy took place a good
many years before 541, if it is true that Elias was his pupil . . .’. According to Westerink, Olympiodorus
was born between 495 and 505. Cf. also id., The Greek Commentaries on Plato’s Phaedo, vol. 1:
Olympiodorus, 20.
53
cf. the term ‘solemn philosophy (semnê philosophia)’ at Philoponus Opif. 1,9. That Philoponus did
indeed take his leave of philosophy seems indicated also by the opening sentence of a later treatise,
the Tractatus de totalitate et partibus ad Sergium presbyterum, where he speaks of his return, just for
once, to philosophical discussion (. . . rursus ad rationalia certamina nos incitat. . .) (A. Sanda,
Opuscula Monophysitica Ioannis Philoponi, Beirut 1930, 126).
262 Aristotle Transformed

about 517, and that ‘the’ in Meteorologica is posterior to 529. The solution I propose
consists in dividing the surviving commentaries into two groups. The first group
embraces in Categorias, in Analytica Priora, in de Generatione et Corruptione, and
in de Anima-in de Intellectu. These writings imply the philosophical system of
Philoponus 1, without capital contradictions. For the moment it is of only
secondary importance to know their order of composition. The second group
embraces in Analytica Posteriora, in Physica, and in Meteorologica. These
commentaries contain passages which are obviously incompatible with the system
of Philoponus 1. However – and this seems to be the key to the whole problem –
the in Physica clearly has elements of both Philoponus 1 and Philoponus 2. Hence
my hypothesis that the commentaries of the second group were originally
composed by Philoponus 1, but were revised by Philoponus 2. But why, one may
ask, did Philoponus 2 limit himself in the in Physica to a revision which left the
earlier version intact, so as to produce plain contradictions? The answer might be
that he needed a commentary (for didactic purposes), but had neither the time
nor the desire to make a completely new one. The sporadic insertion of elements
from his new system would have had the effect of making the old text seem to have
a purely exegetical purpose, when it was no longer consistent with his views, while
the small number of insertions publicised the author’s own ideas on critical issues.
Why was there no revision of the first group of commentaries? We do not know.
But this problem as such is no greater than the problems that arise from a unitarian
approach, e.g.: Why does Philoponus criticise eternalism in the in Physica, but not
in in de Generatione et Corruptione and in de Anima-in de Intellectu, especially
since the latter two commentaries were composed ‘with certain observations of his
own (meta tinôn idiôn epistaseôn)’?54 I fully realise that my hypothesis creates
some new problems, and that a large amount of philological work remains to be
done. But I also think that it is less problematic than the solutions advanced
hitherto; maybe it can serve as a first basis for further investigation.

(a) The date of the in Physica

Most scholars regard the year 517 as the date of ‘the’ redaction of the Commentary
on the Physics.55 Others, somewhat more prudently, consider 517 as the date at

54
Philoponus in GC 1,3; in DA 1,3. The formula occurs also in the title of in Analytica Posteriora (1,3).
55
e.g. M. Wallies, Berliner Philologische Wochenschrift 36, 1916, col. 587; É. Évrard, ‘Les convictions
religieuses’, 299; S. Sambursky, The Physical World of Late Antiquity, 154; W. Wolska, La topographie
chrétienne de Cosmas Indicopleustès, Paris 1962, 149; C.M.K. MacLeod, Jean Philopon, vol. 1, 19, 22;
W. Böhm, Johannes Philoponos, 28, 31.
 The development of Philoponus’ thought and its chronology 263

which Philoponus was working on the redaction of his commentary.56 But one
most significant fact has not yet been sufficiently taken into account: Philoponus
refers three times in the surviving part of the work to his lost critical commentary
on Physica 8 as to an already existing text.57 This can be explained in two different
ways. The first possibility is that the anteriority of *in Physica 8 was just a matter
of succession within one and the same redaction: Philoponus then must have
written Book 8 before he wrote Books 1–4.58 But why should he have done so?
And if he did, why didn’t he simply refer to it in the future tense? Therefore, it
seems better to think of a discontinuous anteriority. On the other hand, it is
impossible to consider the fragments of in Physica 8 edited by H. Vitelli as
belonging to the anti-eternalist *in Physica 8, since some of them clearly accept
the eternity of the world.59 This points in the direction of two different versions
of in Physica 8: a first one in which the author had not yet challenged Aristotelian
eternalism, and a later, separate one in which he criticised Aristotle’s arguments
for the eternity of movement. Moreover, those passages where Philoponus refers
to his anti-eternalist *in Physica 8 must be still later again than this separate
commentary. Consequently, we should distinguish at least three successive stages
in Philoponus’ work on the Physics. The distinction between the second (*in
Physica 8) and the third stage (in Physica2: the second redaction of at least in
Physica 1–4) is clear. But what exactly are my reasons for distinguishing a first
stage (in Physica1) from the second and the third?
In the introduction to this chapter I mentioned an indication of a doctrinal
duality in the in Physica, namely the inconsistency with regard to the eternity of
the world. Similar inconsistencies can be found in Philoponus’ accounts of place
and void. I think that a brief survey of the contradictions which occur on these
three subjects will be sufficient to show that, in the present text of in Physica,
there are at least two strata, representing the systems of Philoponus 1 and
Philoponus 2 respectively.
It is well known that the Commentary on the Physics contains several passages
in which Philoponus challenges the doctrine of the eternity of the world: in Phys.

56
H.D. Saffrey, ‘Le chrétien Jean Philopon’, 401; I. Hadot, Le problème du néoplatonisme alexandrin, 25;
cf. W. Haase, ‘Ein vermeintliches Aristoteles-Fragment bei Johannes Philoponos’, Synusia. Festgabe
für W. Schadewaldt, Pfullingen 1965, 345, n. 1. ‘About 517’: A. Gudeman, ‘Ioannes Philoponus’, col.
1779; M. Wolff, Fallgesetz und Massebegriff. Zwei wissenschaftshistorische Untersuchungen zur
Kosmologie des Johannes Philoponus, Berlin 1971, 110; id., Geschichte der Impetustheorie, 67, n. 2.
57
Philoponus in Phys. 458,30–1; 639,7–9; 762,7–9. Cf. H. Vitelli, Index nominum, CAG 17 s.v.
‘Philoponos’, 995.
58
This seems to be the opinion of É. Évrard, ‘Les convictions religieuses’, 353.
59
cf. below.
264 Aristotle Transformed

54,8–55,26; 191,9–192,2; 428,23–430,10; 456,17–458,31; 467,1–468,4; 762,2–9.

On the other hand the commentator affirms the eternity of heavenly substance,60
of the sublunary cosmos,61 of movement62 and of time.63 The excerpts from
Books 5–8 are eternalist as well: heavenly bodies are eternal;64 sublunary
substances possess an eternity by way of creation and destruction (aidiotês kata
genesin kai phthoran);65 the earth is said to be eternally immobile;66 the eternity
of movement is repeatedly affirmed;67 time is without beginning or end.68

60
cf. n. 10 above (fifth element) and Philoponus in Phys. 1,17; 1,23–4; 15,30; 152,5–7, ‘Some things,
though natural, are ungenerated, I mean celestial things’; 362,21–5 and 497, 8–9 (only local
movement); 601,12–13 (implicitly); 777,11–12 (implicitly).
61
Philoponus in Phys. 236,29–237,4; 303,18–25 (read, ll.24–5 . . . aidia einai kata arithmon, têi
diadokhêi . . .) 405,3–7; 484,15–19.
62
Philoponus in Phys. 298,6–12; 303,1–5; 438,5–6; 747,1–3.
63
ibid., 410,21–4.
64
ibid., 898,2–4.
65
ibid., 838,14–15; 893,6–28.
66
ibid., 859,30.
67
ibid., 812,22–3; 820,30–821,4; 824,22–5; 870,3–8; 894,8–10; 897,23–7; 898,15–16; 905,19–21;
906,38–40 (eternity of heavenly circular movement); 832,17–18, ‘[the motion] must be eternal as
proved at the beginning (. . . anankê aidion einai [sc. tên kinêsin] hôs dedeiktai en arkhêi)’; 838,20–1;
838,28–33; 873,1–2 (eternity of movement). According to M. Wolff, Fallgesetz und Massebegriff,
94–5, Philoponus at in Phys. 797,23–6 rejects the necessity of the eternity of movement. However,
the passage only concerns the finite power of intra-mundane generated substances, not of the
cosmos as a whole. Moreover, in Aristotle’s eternalist view the principle that no finite substance has
infinite power is valid even for the cosmos as a whole: the cosmos is not naturally eternal (cf. Phys.
8.10, 266a24–b6). In the fragments of the commentary on Phys. 8.6–8 Philoponus explicitly
associates himself with Aristotle’s argument, e.g. in Phys. 889,17–23, ‘we proceeded for the rest to
prove that there are some ever-moving things . . .’; 893,6–9, ‘For above he established the eternal
mover from the fact that there is something eternally in motion, whereas here he infers something
eternally in motion from the eternal mover. But this was not unreasonable.’ He affirms the
correctness of Aristotle’s demonstration, e.g. in Phys. 894,26; 897,15–17, ‘There must always be
motion, for he established this at the beginning of this book (. . . anankê aei kinêsin einai (touto
gar en arkhais tou bibliou toutou kateskeuasen . . .)’; 906,38–9, ‘But if motion must be eternal, as he
proved (ei d’anankê aidion einai kinêsin, hôs edeixe . . .)’. I do not see any reason to suppose that
the commentator dissociates himself from Aristotle. And it is most unlikely that in Phys. 890,7 ‘As
was said there (. . . hôs en ekeinois elegeto)’ refers to the anti-eternalist *in Physica 8, as H. Vitelli,
Index nominum, CAG 17 s.v. ‘Philoponos’ (995) suggests. The formula just takes up 890,2–3 ‘He
reminded us of the problem (aporia) raised at the beginning of this book (. . . anemnêsen hêmas tês
en arkhêi tou bibliou toutou aporias . . .)’, which is a reference to Phys. 8.2, 253a7–21 (and not, as
Vitelli ad loc. thinks, to 8.1, 250b11). At Phys. 8.2, 253a7–21 Aristotle himself refers to a later
passage, namely 8.6, 259b1–20, which is exactly the text Philoponus is commenting on here.
According to G.A. Lucchetta, finally, in Phys. 891,33–892,24 must be connected with Opif. 28,20–
29,9, where God is said to have given an initial impetus to the heavenly bodies (G.A. Lucchetta,
‘Ipotesi per l’applicazione dell’ “impetus” ai cieli in Giovanni Filopono’, Atti e Memorie dell’
Accademia Patavina di Scienze, Lettere ed Arti 87, 1974–5, 341–7; id., ‘Aristotelismo e cristianesimo’,
588; Una fisica senza matematica: Democrito, Aristotele, Filopono, Trento 1978, 137–50). This would
mean that the anti-eternalist doctrine of creation of Philoponus 2 is already to be found at in Phys.
891,33–892,24. It would lead me too far afield to refute in detail Lucchetta’s analysis here. Suffice it
for the moment to say that he wrongly takes ‘kinetic power’ (kinêtikê dunamis, in Phys. 891,35),
‘unmoved causes’ (akinêtoi aitiai, 892,7) and ‘lives’ (hai zôai, 892,18) to refer to impetus. But what
Philoponus is talking about is not an external impetus, but a principle of movement that is
immanent in heavenly bodies: soul.
68
Philoponus in Phys. 812,23; 820,30–2; 823,16–20; 870,2–9.
 The development of Philoponus’ thought and its chronology 265

In his Corollarium de loco (in Phys. 557,8–585,4) Philoponus rejects Aristotle’s

definition of place as the first immobile limit of [a thing’s] surroundings (to tou
periekhontos peras akinêton prôton),69 which had still been accepted by
Ammonius.70 In his subsequent commentary on Phys. 4.4, 212a7ff Philoponus
likewise disagrees with the Aristotelian definition.71 To him, the place of physical
bodies is their empty, incorporeal, three-dimensional extension. From this
definition, according to Philoponus, there follow none of the absurdities which
Aristotle, Phys. 4.4, 211b14–29 alleges.72 In his exegesis of Physica 3 and Physica
4,1–4 (down to 212a7), on the other hand, Philoponus does not identify place
and extension, but accepts the Aristotelian definition without criticism. He
himself defines place as the limit of the surroundings (peras tou periekhontos);73
having a place is being limited.74 He uses the Aristotelian concept of place
without any reserve,75 and even appears positively to agree with it.76 It is
erroneous, we are told, to identify place with three-dimensional extension.77
Such a concept of place entails absurdities which do not follow from Aristotle’s
definition of place as the limit of the surroundings.78
In Physics 4,6–9 Aristotle denies the existence of void as a place deprived of
body:79 there is no void, either outside or in bodies.80 Philoponus begins by
commenting on Aristotle’s text down to Phys. 4.9, 217a21,81 not without some
critical interventions.82 The Corollarium de inani (in Phys. 675,12–695,8)
expounds Philoponus’ own theory on void. In the first part (677,9–689,25) the
commentator attacks Aristotle’s refutation of the existence of external void. In

69
Aristotle Phys. 4.4, 212a20–1.
70
Ammonius in Cat. 58,16–17. Cf. ibid., 29,9–10; Ammonius in Isag. 41,1–2; 110,22.
71
e.g. Philoponus in Phys. 592,16–32.
72
ibid., 563,20–5; cf. 567,29–33.
73
ibid., 454,23–4; 536,6–7; 539,5–6; 555,25–7.
74
ibid., 447,18–20; 448,20–1.
75
ibid., 444,5–6; 463,3–4; 499,26–7; 542,17–18; 543,1–4; 546,16–17.
76
ibid., 499,26, ‘we will say (eroumen)’; 536,6, ‘I mean (legô dê)’; 542,18, ‘I mean (legô dê)’; 542,36, ‘so I
say (phêmi oun)’.
77
ibid., 526,20–3: ‘. . . Once this has been shown, I mean what place is, we have the solution of all the
aporias which made it seem that there is no place, and especially of those which arose from
erroneous conceptions of what place is, like the aporias which assume that place is three-
dimensionally extended.’
78
ibid., 555,25–7: ‘Hence none of the absurdities that result for those who hold that place is extension
follow for those who say that place is the limit of the surrounding (body).’ One can hardly imagine
a sharper contradiction than that between in Phys. 555,25–7 and 563,22–5: ‘The argument has
sufficiently proved, I think, that for those who hold that place is extension which is empty according
to its own definition and different from the bodies that enter it, none of the so-called absurdities
mentioned by Aristotle follows.’
79
cf. Aristotle Phys. 4.7, 214a16–17.
80
cf. ibid., 4.9, 217b20–1.
81
Philoponus in Phys. 606,25–675,11.
82
ibid., 632,4–634,2; 639,3–643,8.
266 Aristotle Transformed

the second part (689,26–694,27) he argues that local movement is possible in the
void, and that it is even impossible without a void. This does not mean, however,
that Philoponus considers real void as something existing on its own, without
corporeal filling; it is always a filled void.83 The Corollary is followed by the
exegesis of the rest of Phys. 4.9.84 But the author does not defend the doctrine of
the Corollary in the entire commentary. Although one should not, of course, put
every statement down to the commentator’s personal conviction, there are a few
texts which show that Philoponus must originally have shared Aristotle’s view on
the existence of void. At in Phys. 440,16–17 he writes: ‘. . . that there is no void
among the things there are, he (Aristotle) will prove in the next Book.’85 Elsewhere
Philoponus seems to dissociate himself clearly from ‘certain physicists’ according
to whom movement takes place through the void,86 which is precisely the view
he will himself defend in the Corollarium de inani.87 In this respect, there is even
a contradiction, at least verbally, between the Corollary88 and its immediate
exegetical sequel.89 Outside the Corollary the idea of a necessary concomitance
of body and void90 is rather emphatically ascribed to ‘some physicists’.91 And a
reserve with regard to the reality of void may also be implied in the frequent
attribution to others of the definition of void as a place deprived of body.92
The only explanation I can see for the inconsistencies just mentioned is that
in Physica was originally written by Philoponus 1 and later revised. The
Corollaries, although they are not necessarily contemporaneous with the anti-
eternalist passages, do not in my opinion belong to the first redaction.93 The year

83
ibid., 675,21–9.
84
ibid., 695,9–702,9.
85
. . . kenon de hoti ouk estin en tois ousin, deixei en tôi ephexês bibliôi. Cf. in Phys. 2,25–7: ‘Aristotle
proves (deiknusin) in the fourth book of this work that there cannot be void anywhere (all’ hoti men
oudamou dunatai einai to kenon deiknusin en tôi tetartôi tês prokeimenês pragmateias ho Aristotelês
. . .)’. There can be no doubt that Philoponus was aware of the full significance of (deiknunai): cf. in
Phys. 762,8–9: ‘He tries to prove in the eighth book of this work that there is always motion, but in
fact proves anything but this (Hoti de hê kinêsis aei estin, en tôi ogdoôi tautês tês pragmateias peiratai
men deiknunai, pan de mallon deiknusin ê touto . . .)’. Cf. also 584,14.
86
Philoponus in Phys. 340,7–9; 619,1–4; 661,9–12; cf. 346,8–9.
87
ibid., 677,9–686,29; 689,26–694,27.
88
ibid., 693,30: ‘that it is altogether impossible for there to be motion, if there is no void.’
89
ibid., 697,26: ‘And it is very obvious that there will be motion, even if there is no void.’
90
ibid., 689,4–16.
91
ibid., 2,21–2; 347,9–11.
92
Philoponus in Phys. 340,12–13; 347,10–11; 405,28–9; 408,14–15; 500,28–9; 611,16.
93
The question whether or not the Corollaries fit perfectly with the system of Philoponus 2 requires
further investigation. For the moment I shall discuss only one relevant point. According to M. Wolff,
Fallgesetz und Massebegriff, 136, n. 42, Philoponus in Phys. 686,16–17 ‘for things unlike in kind
cannot be compared (ou gar sumblêta ta anomoiogenê . . .)’ distinguishes the fifth element which
makes up the heavens from the sublunary elements down here. If this were correct, it would mean
that at least one feature of Philoponus 1 survives in the Corollarium de inani. I do not exclude that,
on detailed analysis, other texts might prove the presence or persistence of Philoponus 1 in the
 The development of Philoponus’ thought and its chronology 267

517 may very well have been the time at which Philoponus was writing this first
version, as a part of the project of editing Ammonius’ lectures on Aristotle. It is
possible that this text was a complete commentary on Physica 1–8.94 Later on
Philoponus expounded his changed view about the eternity of movement and
the nature of artificial movement95 in a separate *in Physica 8. Eventually he
reworked at least Books 1–4 to their present form. Apart from the contents, there
are also formal indications of the digressive and supplementary character of the
anti-Aristotelian passages in the in Physica. In the case of the Corollaries textual
autonomy is evident. But in addition the closing words of some of these personal
arguments show their excursive nature. For Philoponus twice refers in closing to
his *in Physica 8,96 once to a work of his which he does not identify,97 and once
to a future, separate account of the problem of the infinity of time.98 Finally there
is a hidden reference to another work or other works in the statement that the
complete argument about the origin of matter is too long to be incorporated in
the commentary on Phys. 1.9, 192a25–33.99 This striking frequency of references
to other writings in the anti-eternalist passages seems to suggest that the author
was in somewhat of a hurry, an impression reinforced by his emphasis on the
limitations imposed by the context100 and on the greater completeness of other

Corollaries, but I do not think in Phys. 686,16–17 proves that Philoponus still accepted the
Aristotelian theory of the fifth element when he wrote this text. It seems to me that the author is not
talking here about a heterogeneity between the heavenly and the sublunary body. The passage
concerns the question of a possible proportionality between movements, the times they need, and
the media in which they take place. It does not follow, Philoponus says, if there is a certain ratio
between the time during which a body moves in a straight line and the time of the revolution of the
lunar sphere, that there should be a proportionate ratio between the respective ‘media’. Times are
always mutually comparable magnitudes, but the same does not necessarily hold true for
movements and their media. Movements not comparable to each other are, e.g., qualitative and
quantitative change, or understanding and local movement. Media of movement not comparable to
each other are, e.g., intellect and body, body and void, and body and surface (in Phys. 685,23–
686,29). In other words, the heterogeneity between the sublunary world and the lunar sphere is not,
in this case, a matter of different substance. In my opinion, it simply means that the medium of
rectilinear movement is a body, the ‘medium’ of heavenly movement, however, a spherical surface.
This seems to be the appropriate way to understand in Phys. 686,12–17: ‘. . . nor, again, I presume, if
the duration of every rectilinear movement through a body has a certain ratio to the duration of the
movement of the lunar sphere, is it necessary on that account that the “media” of the respective
movements, e.g. the air (i.e. a body) and the surface surrounding the (sphere of the) moon, should
also have the same ratio to each other, or should be comparable at all; for things of different kind
are not comparable (to each other). . . .’
94
This can be inferred, of course, from the remaining fragments of the commentary on Physica 5–8.
See also J.L. Kraemer, ‘A lost passage’, 318–19.
95
cf. Philoponus in Phys. 639,7–10.
96
ibid., 458,30–1 and 762,8–9.
97
ibid., 55,24–6; cf. below.
98
ibid., 430,9–10.
99
ibid., 191,34–192,2.
100
ibid., 55,25; 191,34–192,1; 458,15–16.
268 Aristotle Transformed

demonstrations.101 Apparently he was aware of the perplexity he caused his

readers by contradicting himself on the eternity of the world. The references to
other works are intended to assure the reader that anti-eternalism is indeed the
position of the author, notwithstanding the eternalism of in Physica1.
In my opinion, the following texts at least do not belong to the first redaction
of in Physica: in Phys. 54,8–55,26; 191,9–192,2; 428,23–430,10; 456,17–459,1;
467,1–468,7; 557,8–585,4 (Corollarium de loco); 592,16–32; 619,10–13; 632,4–
634,2; 639,3–643,8;102 675,12–695,8 (Corollarium de inani); 761,34–762,9.
Let us return once again to the date of 517. If we were to accept the communis
opinio that the whole text as it now stands was written in or about 517, this
would mean that by an early age Philoponus would already have written quite a
lot of works (which, of course, is not impossible in itself): grammatical treatises,103
and certainly also *in Isagogen,104 in Categorias,105 the *Summikta Theôrêmata106
and *in Physica 8. On the other hand, as we shall see, neither in de Generatione et
Corruptione nor in de Anima-in de Intellectu show any trace of the progressive
ideas Philoponus develops at several places in the in Physica. So we would have
to locate these two commentaries too before 517, thereby implying a radical
change in the author’s thought before 517 (cf. *in Physica 8). And for the period
517–529 there would remain at most: in Analytica Priora, *in Topica107 and in
Analytica Posteriora. If, however, we are right in connecting the date of 517 with
the first, ‘Ammonian’ redaction of in Physica, these difficulties disappear: *in
Physica 8 can be put after in de Generatione en Corruptione and in de Anima-in
de Intellectu without any problem; we will no longer need to set these two
commentaries before 517; and in Physica1 becomes a member of a group of early
commentaries representing the philosophy of Philoponus 1.
If this solution to the chronological problem contains some truth, it is perhaps
possible to say something about the date of Philoponus’ later revision of in
Physica 1–4 as well. First: when did he write the anti-eternalist *in Physica 8,
whether or not this was a separate book? Maybe there are reasons to suppose

101
ibid., 55,26; 191,1–2; 430,9–10.
102
In this passage Philoponus develops his theory of impetus. It contains a reference to *in Physica 8
(639,7–9) and a typical remark concerning the limits imposed by the context (640,23–5).
103
Philoponus was a late starter in philosophy (cf. n. 25 above). Since, according to Simplicius in Cael.
26,21–3, he studied philology first, it is reasonable to set his grammatical writings at the beginning
of his career.
104
Philoponus in Phys. 250,28. Cf. H. Vitelli, Index nominum, CAG 17 s.v. ‘Philoponos’, 995.
105
Philoponus in Phys. 414,21–2; 705,21–2. Cf. H. Vitelli, loc. cit.
106
cf. below.
107
That Philoponus was the author of a lost in Topica may perhaps be inferred from Philoponus in An.
Post. 3,2–4.
 The development of Philoponus’ thought and its chronology 269

that this happened after the publication of de Aeternitate Mundi contra Proclum
in 529. On two occasions indeed de Aeternitate Mundi contra Proclum refers to
the future contra Aristotelem, where we should expect a reference to *in Physica
8 if in 529 that work existed at all.
(1) First, there is a brief argument on the impossibility of the eternity of
movement in Aet. 7,6.108 Philoponus does not say anything about his *in Physica
8 here. Everyone will agree that it would be rather strange if the author completely
forgot here a book he mentions three times in his in Physica. For even if one
takes account of the possibility that *in Physica 8 only criticised Aristotle’s
arguments for the eternity of movement, without proving that this eternity is
impossible, the relevance of its contents to Aet. 257,4–259,6 is clear. No kind of
movement, Philoponus argues, can be eternal: substantial change necessarily has
a beginning and an end, and so do qualitative and quantitative changes. The only
form of movement to remain is local movement. Rectilinear movement cannot
be eternal. This, says Philoponus, has been proved by Aristotle.109 For the circular
movement of heavenly bodies the author refers to the contra Aristotelem. Now, a
statement about the value of Aristotle’s arguments for the eternity of circular
movement (Phys. 8.8–9) would exactly parallel the mention of Aristotle with
regard to rectilinear movement. But there is no such statement. What would
have been more natural here than a reference to *in Physica 8, like that at in Phys.
762,7–9? It is difficult to imagine that Philoponus would not have referred to his
*in Physica 8 (as he did in the ultimate version of in Physica) if this work already
existed in 529. Instead, he promises to refute the eternity of circular movement
in contra Aristotelem. He must indeed have done so in contra Aristotelem Book
6.110 At the same time he promises the publication of the third, non-polemical
work on the creation of the world.111 And the very words he uses might imply
that *in Physica 8 also lay in the future: ‘. . . when we have got rid of the annoyance
resulting from all the aporias.’112 The task of refuting all eternalist arguments in
529 is still ahead in part, namely in so far as Aristotle is concerned. Apart from
indicating that the de Aeternitate Mundi contra Proclum is probably Philoponus’
first anti-eternalist work, the quoted passage reveals the logic of the author’s
enterprise. First he wants to refute the arguments of the eternalists (de Aeternitate

108
Philoponus Aet. 257,4–259,6.
109
Aristotle Phys. 8.8, 261b31–262b8.
110
Simplicius in Cael. 201,3–4. ‘Since he (Philoponus) tries also to refute the arguments in the eighth
book of the Physics which prove the circular movement to be eternal.’ Cf. É. Évrard, Philopon, Contre
Aristote, 87–9.
111
Philoponus Aet. 259,3–5.
112
cf. ibid., 9,22–6: . . . pasas tas peri tou aidion einai ton kosmon aporias elenxantes . . .
270 Aristotle Transformed

Mundi contra Proclum, contra Aristotelem), i.e. to preserve the possibility that the
world is not eternal a parte ante; afterwards he will prove that this possibility is a
necessity. But already in the first two treatises there are arguments which should
strictly speaking have been reserved for the third work.113 Whether or not the
lost in Physica 8 tried to prove the actual necessity of the non-eternity of
movement, we can say that it is, together with the contra Aristotelem, intermediate
in Philoponus’ project between the de Aeternitate Mundi contra Proclum and the
non-polemical work.
(2) A second text to prove, or at least to make it very likely, that *in Physica 8
was not yet written in 529, nor even planned, is Aet. 10,6 (397,25–400,3). This
chapter is also, like part of 7,6, about the eternity of circular movement. Whereas
in 7,6 Philoponus postponed the subject to contra Aristotelem, in 10,6 he argues
against the eternity of circular movement, without, however, going any further
than denying the necessity of this eternity. At the end of the chapter the author
writes:

But if, relying on the attempted proofs of Aristotle, they say that circular
movement has no beginning or end, we shall see whether any of those reasonings
is cogent when, with God’s help, we arrive there. For there is no need to debate
those matters now, since the philosopher [Proclus] in his own arguments has not
put forward anything like that. Let us see also whether there are any arguments
by means of which we can show that it is impossible for circular movement to be
eternal. For the moment we have only set ourselves to object to Proclus’ attempted
proofs. [Hence] we have adduced only, to the best of our ability, what is germane
to refuting his arguments.114

Why would Philoponus only refer to a future refutation of Aristotle’s

arguments for the eternity of circular movement, if such a refutation were
already contained in his *in Physica 8? Reference to an existing work would
better have served the purpose of assuring the reader that the author of de
Aeternitate Mundi contra Proclum was able to cope with Aristotelian eternalism
than a promise not yet fulfilled.
I think it is difficult to avoid the conclusion that Philoponus in 529 did not
even have the intention of composing an anti-eternalist in Physica 8 distinct
from contra Aristotelem 6. On the other hand, contra Aristotelem 6 did contain,

113
Impossibility of eternal movement: Philoponus Aet. 257,4–259,6; contra Aristotelem 6 (Simplicius
in Phys. 1178,7–1179,26). Impossibility of infinite time: Aet. 619,3–620,7. Impossibility of an eternal
cosmos: Aet. 9,1–11,23.
114
Philoponus Aet. 399,20–400,3. Unlike 258,22–6, this text does not mention the title of the contra
Aristotelem.
 The development of Philoponus’ thought and its chronology 271

so far as we can gather from Simplicius, a number of arguments against the

eternity of movement. Some of these arguments refuted Aristotle, others even
tried to prove that eternal movement is impossible. Thus at least the first group
of arguments is identical with what we surely know to have been (part of) the
content of *in Physica 8. As already mentioned, we no longer possess the
refutation of Aristotle’s proofs of the eternity of circular movement (Phys. 8.8–
9).115 But we do have Philoponus’ arguments against Phys. 8.1, where Aristotle
proves the eternity of movement as such: against Phys. 8.1, 251a8–b10. Philoponus
defends the possibility of a first movement;116 against Phys. 8.1, 251b10–28, he
argues that the concept of time does not imply the eternity of movement;117 and
against Phys. 8.1, 251b28–252a5, he defends the possibility of a last movement.118
It appears that *in Physica 8, so far as Philoponus’ personal views are concerned,
must have been very close in content to contra Aristotelem 6. One might even be
tempted simply to identify the two books. But the threefold recurrence of the
term skholai with regard to *in Physica 8 seems to leave no other possibility but
that this text was a real commentary. It must have anticipated, paralleled, or
repeated in exegetical form (part of) the content of contra Aristotelem 6. Maybe
its composition had to do with Philoponus’ teaching activities in those years.
If *in Physica 8 was indeed written after 529, the same holds true for the
revised redaction of in Physica 1–4, in which Philoponus mentions his separate
Commentary on Book 8. In my opinion it is not unlikely that in Phys. 55,24–6
refers to the de Aeternitate Mundi contra Proclum:

Let this for the present purpose be a sufficient reply (eirêsthô Diels) from us also
to (the proposition) that being did not originate, since we have discussed that
kind of proposition adequately elsewhere (en allois hêmin gegumnasmenôn
hikanôs tôn toioutôn theôrêmatôn).

A. Gudeman considered this passage to be a reference to a lost in de Caelo of

Philoponus,119 a view which was sufficiently refuted by É. Évrard.120 Évrard
himself is convinced that Philoponus here refers to his *Summikta Theôrêmata,
and he was followed in this by M. Wolff. To Évrard first: the only argument in
favour of his supposition is the occurrence of the term theôrêma at in Phys.

115
cf. above.
116
Simplicius in Phys. 1130,7–1131,7; 1133,17–1135,15; 1147,10–1149,4.
117
ibid., 1157,6–1159,4; 1164,11–30; 1166,37–1167,16.
118
ibid., 1171,30–1175,26 (with insertions by Simplicius).
119
A. Gudeman, ‘Ioannes Philoponus’, col. 1779.
120
É. Évrard, ‘Les convictions religieuses’, 340, n. 3.
272 Aristotle Transformed

55,26.121 But this is not very conclusive, Évrard is not entirely consistent indeed
in his understanding of theôrêma: he first takes it to mean ‘thèse’, and further
understands it as ‘discussion’. Apparently, however, the theôrêmata at issue are
‘propositions’, not of Philoponus, but of his opponents.122 By the formula ‘that
kind of proposition’ the author probably means views ‘of the same kind’ as the
theory that being is without beginning, views which constitute the subject matter
of at least one of his former works. It is only through a shift of meaning (from
‘thèse’ to ‘discussion’) that Évrard can take the term theôrêma to refer to the title
of the *Summikta Theôrêmata. Wolff123 adduces another argument, namely the
thematical connection he supposes to exist between the question of the
immutability of the three-dimensional second substrate (according to in Phys.
156,16–17, the *Summikta Theôrêmata dealt with that question) and the problem
of the eternity of the world. This leads Wolff to the conclusion that the *Summikta
Theôrêmata most probably expounded in extenso the point of view of in Physica
with regard to the eternity of the world. I cannot enter here into a discussion of
Wolff ’s construction of a ‘theory of substance of 517’. I do not believe that the in
Physica contains a single point of view on matter at all. But even apart from that,
why should a work whose title already makes clear that its subject matter is
heterogeneous (‘Miscellaneous Theôrêmata’) have developed a coherent theme?
We only know that it dealt with an optical124 and a geometrical125 subject, and
with extension as a substrate of bodies.126 One gets the impression of a real
diversity of subjects, as we might expect on the basis of the title. Consequently, if
the work referred to at in Phys. 55,24–6 is not the *Summikta Theôrêmata, we
have to look for other possibilities. In principle there are three: de Aeternitate
Mundi contra Proclum, Contra Aristotelem (de Aeternitate Mundi contra
Proclum), or the third, non-polemical work (+ de Aeternitate Mundi contra
Proclum + contra Aristotelem). To start with the last one, the final redaction of in
Physica is not likely to be posterior to the third treatise. There is indeed a good
chance that in Phys. 430,9–10 refers to the third work as still to be written: Ton
de peri toutôn entelesteron logon idiâi episkepsômetha. This promise concludes an

121
ibid.
122
cf. e.g. Philoponus in Phys. 639,7–9: ‘Against this theôrêma I have spoken in due measure in the
comments on the eighth book of this work (Kai eirêtai men moi pros touto to theôrêma metria en
tais skholais tou ogdoou tautês tês pragmateias)’; Aet. 399,24–6: ‘For there is no need to rehearse
those arguments now (oudemia gar nun ekeina gumnazein anankê. . .)’.
123
M. Wolff, Fallgesetz und Massebegriff, 121–2, n. 27, 133.
124
Philoponus in An. Post. 179,10–11.
125
ibid., 265,5–6.
126
Philoponus in Phys. 156,16–17. It is to be noted that in these three cases the title Summikta
Theôrêmata is explicitly mentioned.
 The development of Philoponus’ thought and its chronology 273

argument on the impossibility of infinite time, based on the impossibility of an

eternal world (in Phys. 428,23–430,10). It parallels the announcements of the
third treatise in Aet. 9,20–7 (. . . en heterôi men proïontes . . . entelesteron
deixomen), 11,14–17 and 259,3–5 (hôsper oun kai entelesteron idiâi deikhthêsetai
. . .).127 And we know that the third work tried to prove not only the impossibility
of the eternity of the world,128 but also, separately, the impossibility of the infinity
of time.129 I am therefore inclined to believe that in Phys. 430,9–10 also announces
the non-polemical treatise, and, consequently, that the final redaction of in
Physica is anterior to this treatise. Is it possible that in Phys. 55,24–6 refers to
contra Aristotelem (or rather both to contra Aristotelem and de Aeternitate Mundi
contra Proclum)? I do not think that this can be ruled out (that is, at the present
stage of investigation). After all, *in Physica 8 (written before in Physica2) was
very close in content to contra Aristotelem 6. But even if in Phys. 55,24–6 was
written after the contra Aristotelem, it refers primarily to the de Aeternitate
Mundi contra Proclum. For contra Aristotelem 1–5 was mainly devoted to a
refutation of the theory of aether, contra Aristotelem 6 mainly to a refutation of
the eternity of movement. Problems like that of the eternity a parte ante of being
are actually dealt with in de Aeternitate Mundi contra Proclum.130
As a result of my analysis I advance the hypothesis that in Physica was first
written about 517, that a separate *in Physica 8 was composed after 529, and that
the Commentary took its present form (through the insertion of at least the anti-
eternalist texts) after 529 as well, either before or after contra Aristotelem,131
almost certainly before the composition of the third treatise.132

(b) The date of in de Generatione et Corruptione and in de

Anima-in de Intellectu
The inconsistencies which allowed us to distinguish earlier and later texts in the
in Physica concerned the eternity of the world, the definition of place and the
existence of void. It soon emerges that, on the first of these subjects, in de

127
For these references see É. Évrard, Philopon, Contre Aristote, 16–19.
128
cf. S. Pines, ‘An Arabic summary of a lost work of John Philoponus’, Israel Oriental Studies 2, 1972,
330–6.
129
ibid., 325–30; Philoponus Aet. 117,20–1: ‘We will show elsewhere that time has a beginning.’
130
Such problems are: ‘Whether body has been generated or not, whether the world has been generated
or not, and, if it has, whether from things that are, or from things that are not’, Philoponus in Phys.
55,20–2. In Aet. 9,8–10 (338,21–344,26) Philoponus argues that it is possible that the world has
come into existence from absolute non-being.
131
The terminus ante quem for the contra Aristotelem is the date of Simplicius’ in de Caelo.
132
The terminus ante quem for the third treatise is the date of Simplicius’ in Physica.
274 Aristotle Transformed

Generatione et Corruptione and in de Anima-in de Intellectu agree with in

Physica1.
There can be no doubt that in in de Generatione et Corruptione and in de
Anima-in de Intellectu Philoponus still accepts the eternity of movement, time
and the world. As to in de Generatione et Corruptione: sublunary substantial
change is a perpetual process of generation (aeigenesia), the efficient cause of
which is the perpetual motion (aeikinêsia) of the heavenly bodies.133 Movement
as a whole is eternal,134 change of place is eternal in so far as it is circular.135 In
Philoponus’ exegesis of de Generatione et Corruptione 2.11, where Aristotle
argues that an absolutely necessary, eternal movement can only be cyclical,
nothing betrays even the slightest doubt about the correctness of Aristotle’s
doctrine.136 Time is infinite.137 Heavenly bodies are ‘numerically’ eternal,138
sublunary living substances are ‘specifically’ eternal.139 And of course, in de
Generatione et Corruptione agrees with Aristotle’s theory of the fifth element.140
In de Anima also views movement as eternal141 and time as infinite.142 Heavenly
bodies are eternal,143 and so are sublunary species.144 Here again the fifth element
is evidently accepted.145 In in de Intellectu, finally, Philoponus likewise appears to
believe in the eternity of heavenly bodies146 and mankind,147 as well as in the
theory of aether.148
Compared with the statements about the eternity of the world, Philoponus’
position with regard to the definition of place and the existence of void in in de
Generatione et Corruptione and in de Anima-in de Intellectu is not so

133
Philoponus in GC 49,12–13; 49,23; 288,19–289,22; 290,12–13; 290,24–5; 291,21–3; 296,14–17;
299,18–19.
134
ibid., 299,22; 299,25; 300,1–3; 300,7–8.
135
ibid., 312,17–18: ‘It has been demonstrated in the eighth book of the Physics that only circular
motion is eternal.’
136
cf. R.B. Todd, ‘Some concepts in physical theory in John Philoponus’ Aristotelian commentaries’,
Archiv für Begriffsgeschichte 24, 1980, 159, n. 34: ‘The commentary on the GC was probably an early
work, since in discussing 2.10 and 11 Philoponus does not question Aristotle’s theory that the
movement of the heavenly bodies was eternal; such criticism . . . was to come later.’
137
Philoponus in GC 45,5–6.
138
ibid., 1,9–12; 3,1–4; 134,4–5; 206,2–4.
139
ibid., 296,14–298,8.
140
e.g. ibid., 129,6–14; 135,1–2; 135,23–4; 139,20–1; 144,21–2; 147,19–25; 199,4–6; 282,23–7; 283,
1–10.
141
Philoponus in DA 21,1–2; 132,32–133,3.
142
ibid., 133,1–2.
143
ibid., 18,27; 24,27–9; 141,2–4; 259,9–10; 324,15–16.
144
ibid., 7,12–19; 228,16–17; 265,30–4; 268,6–9; 268,37–269,1; 269,26–7; 270,2–4; 270,29–30; 272,
31–2; 279,11–12; 286,19–21; 286,32–4.
145
ibid., 9,6–7; 9,14–15; 56,2–4.
146
Philoponus in de Intell. 78,2–4.
147
ibid., 52,21–3; 59,21–4.
148
e.g. ibid., 32,67–8; 36,57–9.
 The development of Philoponus’ thought and its chronology 275

evident. However, what is clear is that the author has not yet developed in
these commentaries the views he defends in the Corollaries of in Physica. Let us
first examine the concept of place, beginning with in de Generatione et Corruptione.
It is true that some passages in this work are somewhat ambiguous, because
they try to define place on the basis of a connection between the concepts of a
limit (peras) and an extension (diastêma).149 But nowhere does one find the
categorical opposition to the Aristotelian doctrine of place which is characteristic
of part of the in Physica. On the contrary, the general impression is that Philoponus
seems to agree with Aristotle, e.g. with GC 1.6, 322b33–323a1 (‘position’ belongs
to those things which also have a ‘place’, tr. Forster):150 on this occasion he recalls
explicitly the Aristotelian concept of place.151 Elsewhere place is defined as a
boundary (horos),152 as a relation of the surrounding to the surrounded.153 In in de
Anima the Aristotelian concept of place is not contested.154 And in in de Intellectu
there is a passage which implies that the non-Aristotelian definitions of place are
false.155
The existence of void is rejected more than once in in de Generatione
et Corruptione.156 And one passage in in de Anima denies the possibility of
movement through the void.157
I think that the materials gathered here are sufficient for us to affirm that in de
Generatione et Corruptione and in de Anima-in de Intellectu were composed by
Philoponus 1. As I have already said, it is less important for the moment whether
these commentaries are earlier or later than in Physica1. What is essential is that
they allow us to complete the system we find in in Physica1, without the risk,
which is in principle present in my analysis of in Physica, of taking texts with a
purely exegetical intention to contain Philoponus’ personal views. The fact that

149
Philoponus in GC 73,28–30; 157,5–6.
150
ibid., 132,12–13; 132,19.
151
ibid., 134,1–4. The formula ‘according to the argument on place handed down by Aristotle’ does not
seem to imply any reserve; rather it indicates the consistency of Aristotle’s argument.
152
ibid., 280,21–4.
153
ibid., 301,28–9.
154
Philoponus in DA 100,6–16.
155
Philoponus in de Intell. 42,91–6: ‘It is said in the fourth Book of the Physics, since there was doubt
among the ancients about place, that it is necessary to give such an account of place that all the
characteristics which it possesses according to its very concept are entailed in it, the doubts which
are raised about it are solved, and the reason why the ancients went astray on the subject is revealed
to us (Dictum est autem in quarto De Naturali auditu, quoniam apud antiquos de loco ideo
dubitabatur, quod oportet talem assignare de loco rationem, cui omnia quae secundum propriam
intelligentiam insunt assequuntur, et dubitationes quae feruntur circa ipsum solvantur, et causam
nobis manifestam faciat propter quam antiqui de ipso erraverunt)’.
156
Philoponus in GC 4,17; 73,8; 156,20–5; 178,21–2. ‘For it was proved in the Physics that it is absolutely
impossible for there to be void.’
157
Philoponus in DA 359,10.
276 Aristotle Transformed

e.g. the doctrine of the eternity of the world in in de Generatione et Corruptione

and in de Anima-in de Intellectu is not only not contested, but even accepted by
Philoponus in his own person, entitles us to consider the eternalist elements in
in Physica as not merely exegetical, but representative of an in Physica1 written
according to the principles of Alexandrian Neoplatonism.

(c) The date of the Commentaries on the Organon

I have already mentioned that in Analytica Posteriora belongs to the same
group as in Physica and in Meteorologica. With regard to the eternity of the
world, it fits in with the commentaries of Philoponus 1.158 However, at in An.
Post. 242,14–243,25 Philoponus rejects the thesis that Aristotle accepted Plato’s
theory of Ideas. This means that he agrees here with de Aeternitate Mundi contra
Proclum.159 Hence in An. Post. 242,14–243,25160 was written by Philoponus 2.
According to Philoponus 1, on the contrary, Aristotle did accept the existence of
transcendent Ideas as logoi in the demiurgic Intellect.161 So we have to conclude
that in Analytica Posteriora, just like in Physica, contains elements of both
Philoponus 1 and 2, and that it received its present form at least partially around
or after 529. This particular volte-face is set in a wider context in my chapter on
Ammonius.
As far as I can see at present, in Categorias and in Analytica Priora do not
contain anything which clearly belongs to the system of Philoponus 2. In the in
Categorias Philoponus refers approvingly to Physica 8 and explains the eternity
of movement by reference to the infinite power of the first unmoved cause.162
The world is eternal in its present structure; God creates it by necessity,163 from
all eternity.164 At the time Philoponus wrote his in Categorias he had not yet
changed his mind on the definition of place. More than once he refers explicitly
to Aristotle’s definition,165 he alludes to it166 or uses it.167 And at least one text

158
Philoponus in An. Post. 67,17–18 (implicitly); 110,14–15; 135,11–15.
159
cf. nn. 16, 21 above.
160
See also Philoponus in An. Post. 133,30–134,2.
161
cf. n. 16 above.
162
Philoponus in Cat. 50,23–31.
163
ibid., 145,10–146,2.
164
cf. Philoponus in An. Post. 243,17–24; 244,18–25.
165
Philoponus in Cat. 33,22–6; 87,7–10, ‘As we said before, place is the boundary of the surroundings
insofar as it surrounds what is surrounded’; 165,1–2.
166
ibid., 32,20–2.
167
ibid., 99,29–30.
 The development of Philoponus’ thought and its chronology 277

shows that the author had not yet developed the doctrine of the Corollarium de
inani on the possibility of movement through the void.168

(d) The date of in Meteorologica

É. Évrard has proved that in Meteorologica in its present form has to be set after
de Aeternitate Mundi contra Proclum.169 However, I have the impression that the
text shows traces of an earlier version, that it is the result of a later revision which,
just as in the in Physica, has produced some inconsistencies. At present I have to
limit myself to one provisional remark, which poses the problem rather than
solving it. It is well known that in his in Meteorologica Philoponus vigorously
rejects Aristotle’s theory of the fifth element.170 On the other hand, there are a few
passages where the author still seems to accept a fundamental difference between
heavenly and sublunary reality (which is typical of the system of Philoponus 1).
At in Meteor. 11,24–37 the heaven is considered as intermediate between
intelligible and perishable being. At in Meteor. 12,24–32 also, heaven is ranked on
a higher ontological level than the sublunary world.171 Closer investigation is
needed to answer the question whether the commentary presents still other
anomalies which might reinforce the supposition that it is a revision by Philoponus
2 of an older version.

The motives for Philoponus’ volte-face

I return now to the motives for Philoponus’ volte-face. In Arabic philosophy

there is a tradition that he ‘deceived’ his public. Let us start with the important
findings of J.L. Kraemer on this subject.172 The section on Philoponus in the
S.iwân al-H.ikma of as-Sijistânî (tenth century) deals with the Alexandrian’s
intentions when writing his de Aeternitate Mundi contra Proclum and contra
Aristotelem. According to one of as-Sijistânî’s sources, Philoponus undertook his

168
ibid., 86,23–7.
169
cf. n. 4 above. Évrard is convinced that in Meteorologica was written between de Aeternitate Mundi
contra Proclum and contra Aristotelem. According to C. Wildberg, ‘Prolegomena to the Study of
Philoponus’ contra Aristotelem’, in R. Sorabji (ed.), Philoponus and the Rejection of Aristotelian
Science, 202–8, the in Meteorologica is later than contra Aristotelem.
170
See É. Évrard, ‘Les convictions religieuses’, 328–34.
171
Philoponus in Meteor. 12,30–1: ‘And that (the heaven) by its essence transcends the bodies it
encircles and governs everything by its activities.’
172
J.L. Kraemer, ‘A lost passage’, 321–5.
278 Aristotle Transformed

refutations of both pagan philosophers (I quote Kraemer) ‘in order to pacify the
wrath of his fellow Christians, aroused by his preoccupation with the exegesis of
Aristotle’s works, and to protect himself against their threats of diverse forms of
duress’.173 According to another source he received money from the Christians
for writing both treatises.174 ‘Both reports have in common’, Kraemer writes, ‘the
view that Philoponus did not express his true convictions in the contra Aristotelem
and in the contra Proclum but acted, according to the first, in the interest of self-
protection or, according to the second, in the interest of material gain. Are these
stories unfounded or do they contain a grain of truth? There are grounds for
preferring the second alternative.’175 Kraemer’s references to other Arabic texts
show that this view was not uncommon. Even Avicenna was convinced that
Philoponus had deceived the Christians by trying to give the impression that he
disagreed with Aristotle. And Avicenna adds the simple but invaluable remark
that whoever studies Philoponus’ other works, e.g. his in de Generatione et
Corruptione, will not fail to notice his agreement with Aristotle.176
Thanks to the work of M. Mahdi we know that al-Farabi, in his Against John
the Grammarian, likewise held that Philoponus’ contra Aristotelem did not
express the Alexandrian’s real conviction. I quote, in Mahdi’s translation, the end
of al-Farabi’s Introduction to his treatise:

Then, in many of his objections and in many of [the things] by means of which
he (Philoponus) seeks to free himself from the doubts that he raises against
himself [or his own position], he makes use of the opinions [or doctrines] laid
down [or legislated] in [various] religions [or religious communities] and
whatever consequences follow from them [that is, from these religious opinions].

173
ibid., 322. There is a certain parallel here with what Philoponus himself says in the Preface of de
Opificio Mundi. This work too appears to have been solicited by other Christians. The author begins
by recalling that he has already devoted several books to the problem of creation, in order both to
refute the pagan eternalists and to prove positively that the world had a beginning (1,6–14). Yet
since then people have been continually putting a gentle pressure upon him (sunekhôs hêmin
ênôkhloun êrema pôs), blaming him a little (hupôneidizon) for limiting himself to the refutation of
philosophical arguments and not having done anything yet to defend the Mosaic creation story
against its detractors (1,14–2,4). The person who urged Philoponus most of all was Sergius, the
patriarch of Antioch, 2,4–8: ‘You pressed on me heavily, urging me and almost even forcing me to
contribute my best efforts to the cause (. . . polus epekeiso protrepôn, mononoukhi kai biazomenos, ta
eis emên suneispherein tôi pragmati dunamin)’. At the moment he wrote these words (546–9
G. Reichardt, W. Wolska or 557–60 É. Évrard, H.D. Saffrey) Philoponus’ anti-eternalist treatises and
commentaries had already been published for some time. If the Christians were not yet completely
satisfied with de Aeternitate Mundi contra Proclum, contra Aristotelem etc., it is not difficult to
imagine what their attitude must have been about 529, when Philoponus had produced mainly
Aristotelian-Neoplatonic philosophical writings, without any contribution to the Christian cause.
174
J.L. Kraemer, ‘A lost passage’, 323, draws attention to the resemblance between this allegation and
Damascius’ famous remark about the agreement of Ammonius with the ecclesiastical authorities.
175
J.L. Kraemer, op. cit., 322.
176
ibid., 323–4.
 The development of Philoponus’ thought and its chronology 279

It is unlikely that he did not understand how far removed these opinions are
from the nature of things. Therefore one may suspect that his intention from
what he does in refuting Aristotle is either to defend the opinions laid down in
his own religion about the world, or to remove from himself [the suspicion] that
he disagrees with the position held by the people of his religion and approved by
their rulers, so as not to suffer the same fate as Socrates.177

Maybe – but this is just a prudent suggestion – a glimpse of the same

accusation is to be found in Simplicius. As we shall see,178 there are reasons for
suspecting that Simplicius knows more about Philoponus than he claims. If this
is correct, at least one of his sneers at Philoponus might be taken to allude to the
insincerity of his adversary. ‘No one,’ he says, ‘aimed in such an unserious way,
because of the prevailing worthless conceptions about the demiurge of the
world, only at seeming to oppose those who demonstrate the eternity of the
world.’179
Obviously, some elements of the Arabic tradition concerning Philoponus’
deception of the Christians (e.g. Avicenna’s remark on in de Generatione et
Corruptione) are nothing more than the result of an interpretation of the doctrinal
duality in his philosophy. But one cannot for that reason reject the tradition in its
entirety. After all, the facts are that Philoponus started his philosophical career as
an adherent of Ammonius and that in 529 he wrote de Aeternitate Mundi contra
Proclum, a book in which he radically abjured his philosophical past. For it is
impossible to conceive the relation between the basic tenets of e.g. in de Anima
and in de Generatione et Corruptione on the one hand, and those of de Aeternitate
Mundi contra Proclum on the other as a gradual process of moving away from
Alexandrian Neoplatonism. There is an evolution (in the proper sense) from de
Aeternitate Mundi contra Proclum via contra Aristotelem through de Opificio
Mundi.180 But 529 is a clear rupture with the past, and for this rupture we have to
find an explanation. There can be little doubt that it had something to do with the
hostile attitude of Justinian towards the teaching of pagan philosophy. In 529 at
the latest Philoponus must have felt the necessity of publicly renouncing his
earlier philosophy. Maybe, as the Arabic tradition suggests, the Christians, in one
way or another, were powerful enough to force him to do so. On this hypothesis

177
M. Mahdi, ‘Alfarabi against Philoponus’, 256–7.
178
cf. p. 263 below.
179
Simplicius in Cael. 59,13–15: . . . oudeis mentoi kakoskholôs houtôs eis monon apeblepsen to
antitetakhthai dokein tois aidion ton kosmon apodeiknusi dia tas kratousas euteleis ennoias peri tou
ton kosmon dêmiourgêsantos. On the meaning of formulas like hai kratousai ennoiai cf. below.
180
cf. below.
280 Aristotle Transformed

(which covers several possibilities: serious threats, moral pressure, promise of

remuneration etc.), the Christians put pressure on Philoponus because he must
have been a famous philosopher by then, and because they had never ceased to
claim him as a Christian and/or because he himself continued to play that role
(notwithstanding his adherence to the philosophy of Ammonius). Or else they
considered him a renegade to whom, for lack of a better protagonist, they offered
another chance. Anyway, Philoponus was cooperative. This may have been due to
fear, but it is not impossible either that there was something more. As we have
seen, there is some evidence that Philoponus (continued or) started teaching
philosophy after 529. Apparently, he wished to remain active as a philosopher,
and considered his volte-face as the beginning of a new period in his philosophical
career. Perhaps his fellow Christians and he himself tried to use the new situation
in order to establish Christian philosophy at the Alexandrian school. Launching
an attack on Proclus for Philoponus had a double advantage. First, it associated
him with the spirit of Justinian’s action against Athenian philosophy. Secondly, it
offered him the possibility of attacking Ammonius only indirectly, of diverting
the public’s attention from the fact that he was a former adherent of Ammonius.181
Moreover, it is reasonable to suppose that, in the lost Introduction to de Aeternitate
Mundi contra Proclum, Philoponus justified his sudden change of position in one

181
It is noteworthy that Ammonius is not mentioned by name in de Aeternitate Mundi contra Proclum
(at least not in the surviving text). By contrast, in Analytica Posteriora and in Physica contain
passages overtly polemicising against ho philosophos (cf. n. 31 above), and in the in Meteorologica
Ammonius is mentioned as ho hêmeteros didaskalos (91,3) and ho hêmeteros didaskalos Ammônios
Hermeiou (106,9). A possible explanation of this difference might be that Philoponus wrote the
commentaries just mentioned in connection with a teaching activity of his (within the school there
would have been no sense in trying to conceal his philosophical past). De Aeternitate Mundi contra
Proclum, on the other hand, was written for a larger audience, and Philoponus may have been
prudent enough not to draw special attention to his relation with the pagan scholarch Ammonius.
I think there is a text in de Aeternitate Mundi contra Proclum that might substantiate this view. In
Aet. 9,11 (344,27–365,11) the author argues that the eidos of material beings comes to be out of and
perishes into absolute non-being. In order to reach this conclusion he rules out, one by one, seven
other possibilities (cf. 316,3–14). The last of these alternative solutions is introduced by the
following sentence: ‘Driven into a corner by these and similar arguments of ours, and trying, not
clumsily at all, to rescue his own opinion, he invented an argument like this’ (359,14–17). This text
suggests that Philoponus is referring here to an oral discussion (whether fictitious or not), a
discussion which, of course, he cannot remember literally. After the ‘quotation’ he writes: ‘Such was
his (ekeinou) argument, and we refuted the plausible element in the argument (Ta men oun
par’ekeinou toiauta. Kai hêmeis de tou logou toutou to pithanon apelenkhontes)’, 360,8–10. Who is ‘he
(ekeinos)’? Obviously it is not Proclus. This might also be inferred from the fact that Philoponus,
when he quotes Proclus, habitually indicates the work from which the quotation comes (cf. in the
present context Aet. 364,5–6). In my opinion ekeinos might very well be Ammonius. It has to be
someone on the ‘Hellenic’ side. Moreover, it has to be someone whom Philoponus’ readers in 529
know or can assume to have had personal contact with Philoponus. Finally, it must be someone
who both is very well known (otherwise there would be no sense in using ekeinos) and whose name
Philoponus, apparently by some damnatio memoriae, does not want to mention publicly. This
interpretation, if correct, implies that in 529 Philoponus tried to assure his readers that he had
already been opposing eternalism at the time he was still a pupil of Ammonius.
 The development of Philoponus’ thought and its chronology 281

way or another. Even in the remaining corpus of the treatise there are a few
allusions which may offer an indication of the manner in which the author
wanted his audience to understand his transition from Neoplatonic to Christian
philosophy. It appears that he turned things upside down, and that he was trying
to convince his readers that his Neoplatonic works (up to 529) did not express his
personal views, but were produced under some kind of external pressure. This
seems to be implied in Philoponus’ references to Plato’s fear of the masses. In
order not to suffer the same fate as Socrates (one should remember what al-
Farabi says about Philoponus himself) Plato often conformed to popular
mythology and was silent about his real conviction: . . . phobôi polla tôn deontôn
aposesigêken.182 In the same way, Philoponus wants to imply, I myself, going along
like Plato with the prevailing opinion (têi tote kratousêi doxêi hepomenos),183 and
in order not to become a victim of pagan fanaticism,184 have been expounding a
philosophy with which I did not agree. Most significant in this respect is what
Philoponus, contra Proclum 2,2 (26,20–33,5) writes about Aristotle’s criticism of
Plato’s theory of Ideas, although the suggested excuse is less strong here. Quoting
EN 1.6, 1096a11–17,185 Philoponus implicitly parallels Aristotle’s relationship to
Plato with his own relationship to Ammonius: Plato (Ammonius) is a friend, a
greater friend the truth (amicus Plato (Ammonius), magis amica veritas). And
when he refers to Aristotle’s ‘respect for his teacher’ (hê aidôs hê pros ton
didaskalon) he may first of all have in mind the effect of this formula on the way
his readers will appreciate his own ‘emancipation’ from Ammonius. The
philosopher, we are told, must be prepared, not only to correct the errors of others

182
Philoponus Aet. 331,17–332,23, quoting (332,1–11) Plato Epist. 5, 322A–B. The reference to silence
out of fear, combined with the opposition ‘secret belief in a personal god versus apparent agreement
with paganism’ seems to leave no room for the possibility that Philoponus’ allusion is to his being
threatened by the Christians himself. Yet, there is no fundamental objection either, I think, to
crediting Philoponus’ story about Plato with a double basis.
183
Philoponus Aet. 635,22–3. Similar formulas are often used by pagan authors to refer to the dominant
position of Christianity. A. Cameron, ‘The last days of the Academy’, 18 gives four instances from
Damascius’ Vita Isidori (including the fragment on Ammonius: fr. 316, 251 Zintzen). The term ‘the
present (ta paronta)’ is used in a similar sense by Proclus (‘the present time, ho parôn khronos’),
Damascius, Simplicius and Olympiodorus (A. Cameron, op. cit., 15). On the other hand such
formulas also occur in Christian authors: cf. I. Hadot, ‘Le problème du néoplatonisme alexandrin’,
23–4, n. 31.
184
Philoponus may have hoped that the allusion reminded his readers of the fact that in a not too
distant past Christian students were not always accepted in Alexandrian academic life. Zacharias,
Vita Severi, 22–7 Kugener tells the story (c. 485–7 ad ) of an aspirant-Christian, Paralios, beaten up
by his pagan fellow-students with the Grammarian Horapollo the Younger; cf. H.I. Marrou,
‘Synesius of Cyrene and Alexandrian Neoplatonism’, in A. Momigliano (ed.), The Conflict between
Paganism and Christianity in the Fourth Century, Oxford 1963, 136–7.
185
Philoponus Aet. 30,25–31,7. On Aet. 2,2 in its entirety cf. L. Tarán, ‘Amicus Plato sed magis amica
veritas. From Plato and Aristotle to Cervantes’, Antike und Abendland, 30, 1984, 112–14.
282 Aristotle Transformed

for the sake of truth, but also to revise his own suppositions if these prove to be
incorrect.186 Philoponus is no longer implying here (as he is when talking about
Plato’s fear of the masses) that he has been forced to hide his real conviction.
Rather, he seems to concede that he has been erring himself, and he is prepared
to abjure his earlier philosophy (elenkhein tas oikeias hupotheseis). Doesn’t all this
sound like an attempt to create the impression that he feels guilty about his past?
I think the frequency of announcements of the anti-eternalist books still to be
written, announcements that culminated in the lost proanaphônêsis tôn hexês187
at the end of the work, points in the same direction. What is the reason for this
strongly marked concern to remove all possible doubt about his position? Doesn’t
it seem as if the author feels the need to compensate? Promise is debt, but isn’t the
reverse true as well? In my view all these announcements are to be explained by
the fact that the de Aeternitate Mundi contra Proclum was a very surprising
publication to the Alexandrian public, and that no one expected such a book
from Philoponus. Consequently, the author had a great interest in presenting his
new theories as part of a larger programme, a new cosmology with which he
wanted to be identified in the future. He, or those who commissioned the book,
must have felt that the de Aeternitate Mundi contra Proclum alone did not justify
forgetting and forgiving the entire ‘Ammonian’ production of Philoponus. That is,
I think, why he tried to diminish the incredulous amazement of the people who
knew him as a philosopher. By promising the publication of the contra Aristotelem
and of the third, non-polemical work about the creation of the world, he obviously
wanted to get credit for the future. The fact that Philoponus offered a programme
for a new philosophy and his assurance of personal reliability and predictability
might both point to an attempt at establishing Christian philosophical teaching
in Alexandria.

Simplicius’ acquaintance with Philoponus

Simplicius’ contention that he never met the Grammarian (in Cael. 26,18–19) is
mostly believed to be sincere.188 Consequently, an explanation is sought for this

186
Philoponus Aet. 30,15–25: . . . kai dikaion . . . , philosophous ontas, mê monon tas allôn hupotheseis epi
sôtêriâi tês alêtheias elenkhein alla kai tas oikeias hekaston, ei mê têi alêtheiâi sumphônoi heuretheien.
187
ibid., 611,25–6.
188
cf. recently P. Hoffmann, ‘Simplicius’ polemics: some aspects of Simplicius’ polemical writings
against John Philoponus: from invective to a reaffirmation of the transcendency of the heavens’, in
R. Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science, 68.
 The development of Philoponus’ thought and its chronology 283

‘fact’.189 However, one cannot rule out the possibility that Simplicius is indulging
here in a haughty lie: that newcomer in philosophy (nearos: in Cael. 42,17), that
philologist, professional philosophers did not even take notice of him. It might be
significant that Simplicius does not write ‘whom I have never even seen (hon oude
etheasamên pôpote)’. For there may have been, at the time he published his in de
Caelo, enough former pupils of Ammonius among the philosophical public to
know that this statement (and, at the same time, Simplicius’ ‘depersonalisation’ of
the controversy) was a lie (if it was a lie indeed). He writes: ‘whom I am not
conscious of ever having even seen (hon oude (‘not even’) theasamenos oida
pôpote)’ which, I presume, is not the same as ‘whom I know I never saw (hon
oudepôpote theasamenos oida)’. This might imply the following calculation: maybe
there are readers who know for sure that Simplicius and Philoponus were in
Ammonius’ school at about the same time. Well, in that case they certainly will get
the message, namely that for real philosophers the Grammarian was simply
negligible. The sincerity of Simplicius’ claim not to know Philoponus has recently
been questioned again by H.J. Blumenthal.190 On the other hand one should not go
so far as to suppose that Simplicius’ hostility towards Philoponus was primarily
inspired by a personal grudge, for instance about the succession to Ammonius. In
this I agree with K. Praechter against Gudeman and Wieland.191 Yet there are
enough elements to venture the hypothesis that Simplicius did know Philoponus.
The mere fact that he claims three times not to know his opponent (in Cael. 26,19;
49,24–5; 90,12) is already sufficient in itself to raise suspicion. Further, if Simplicius
does not know Philoponus, where did he get his information about him? It is true
that much of what Simplicius says about Philoponus’ lack of philosophical
education is based just on what Philoponus wrote. Some of it even could be taken
to be deduced a priori from the title ‘Grammarian’ (in Cael. 119,7: houtos ho
Grammatikon heauton epigraphôn). But how does Simplicius know that Philoponus
never had a teacher in Platonic philosophy? At in Cael. 84,11–13 he writes: ‘But
since – I don’t know how – Plato’s views seem to please this individual (Philoponus),
although he had no teachers in that matter, as they say (hôs phasi), and although
he has not investigated himself with (real) love of knowledge the true meaning of

189
cf. e.g. H.D. Saffrey, ‘Le chrétien Jean Philopon’, 402, n. 4: ‘It would be explicable that Simplicius did
not meet Philoponus at Alexandria, if he was a philosophy student, when Philoponus was still a
philology student.’
190
H.J. Blumenthal, ‘529 and its sequel: what happened to the Academy?’, Byzantion 48, 1978, 379–80,
and ‘John Philoponus: Alexandrian Platonist?’, 318.
191
K. Praechter, ‘Simplikios’, in G. Wissowa and W. Kroll (eds), Realencyclopädie der klassischen
Altertumswissenschaft III A, 1, 1927, col. 204; A. Gudeman, ‘Ioannes Philoponus’, cols 1766–7; W.
Wieland, ‘Die Ewigkeit der Welt’, 301.
284 Aristotle Transformed

(what) Plato (said) either . . . ’. The formula hôs phasi might very well be a rather
transparent attempt to hide the truth. Or is the statement that Philoponus had no
teachers in Platonic philosophy just an inference from the fact (which, one might
argue, could be verified in the libraries) that the Grammarian never seriously
devoted himself to Platonic exegesis, if that is the meaning of ‘neither having
himself examined Plato’s intention with any care (mête auton philomathôs ezêtêkota
ton tou Platônos noun)’? Regardless of whether or not Simplicius is completely
sincere here, the text seems relevant to a possible limitation of Ammonius’ teaching
to Aristotelian philosophy (cf. Chapter 10 above).

The evolution of Philoponus 2

I shall sum up my chronological conclusions briefly. The publication of the de

Aeternitate Mundi contra Proclum in 529 marks a decisive break in Philoponus’
philosophical production. Before that date the author was what we might call an
‘Alexandrian Neoplatonist’. On the basis of the teaching of Ammonius he edited
several commentaries which display the philosophical system that I have called
‘Philoponus 1’. Among the surviving texts the following belong to this first group:
in Categorias, in Analytica Priora, in de Generatione et Corruptione and in de
Anima-in de Intellectu. To these one should add in Physica1, probably the earlier
stratum of in Analytica Posteriora, and maybe some older parts of in Meteorologica.
The de Aeternitate Mundi contra Proclum was probably Philoponus’ first anti-
eternalist work. It was followed by the contra Aristotelem and by at least one other
work on the creation of the world. At the same time Philoponus 2, probably for
teaching purposes, revised some of the earlier exegetical material. The most
obvious case of revision is the in Physica. There were at least two successive stages
in this revision. Philoponus first wrote an anti-eternalist commentary on Physica
8 (*in Physica 8). Afterwards he inserted, in the original text (in Physica1), a
number of anti-eternalist passages (in Physica2). I assume that both *in Physica 8
and in Physica2 are later than the de Aeternitate Mundi contra Proclum and earlier
than the third, non-polemical treatise. In Analytica Posteriora has a passage on
the doctrine of Ideas which is clearly Philoponus 2. In Meteorologica in its present
form is Philoponus 2 as well. I have limited myself in this chapter to the problem
of distinguishing Philoponus 1 from Philoponus 2. My chronology involves only
one new element for relative dating within one group, namely the posteriority of
in Physica2 to the de Aeternitate Mundi contra Proclum. The relative chronology
of in Categorias, in Analytica Priora, in de Generatione et Corruptione and in de
Anima-in de Intellectu awaits further investigation.
 The development of Philoponus’ thought and its chronology 285

Obviously, there are many problems left. On p. 243 I indicated two of them,
and proposed an answer to the first: the failure to rewrite in Physica completely.
As to the second, the failure to revise the other early commentaries, we can only
speculate. The non-revised commentaries pertain to logic (in Categorias, in
Analytica Priora), natural philosophy (in de Generatione et Corruptione) and
psychology (in de Anima-in de Intellectu). We just don’t know why in Categorias
(one might plausibly argue that in Analytica Priora hardly needed any revision)
was not revised, whereas in Analytica Posteriora was, nor why Philoponus 2
rewrote in Physica and in Meteorologica, but not in de Generatione et Corruptione.
Only with regard to in de Anima-in de Intellectu could one reasonably propose
an explanation. The Neoplatonic inspiration of this commentary, one might say,
was so dominant that in this case a dichotomy between mere exegesis of Aristotle
and criticism of him (such as is found in the in Physica) was not possible.
Moreover, the ontological place of the human rational soul was a major difficulty
for Philoponus 2, as I shall explain below. If it is true that Philoponus 2 made his
revisions for teaching purposes, then it is legitimate, I think, to suppose that his
lectures must have been primarily about natural philosophy. After all, the
philosophy of Philoponus 2 was primarily a cosmology.
The difference between the systems of Philoponus 1 and Philoponus 2 cannot
be explained by supposing a gradual evolution from Alexandrian Neoplatonism
to the Christian doctrine of creation to be found in the de Aeternitate Mundi
contra Proclum etc. In the first section of this chapter I summarised the main
discontinuities between Philoponus 1 and Philoponus 2. One can assume an
evolution only when there is room for some intermediate stage. But what could
possibly serve as the middle term between Philoponus 1 and 2 on such questions
as the eternity of the world, divine freedom, the harmony of Plato and Aristotle,
the principle ‘nothing comes from nothing’, etc.? One of the most striking
examples of Philoponus’ volte-face is the radical change in his interpretation of
Plato’s psychogony (Tim. 34–6).192

192
This is clear from a comparison between in DA 115,22–121,10 and Aet. 6,24 (195,7–200,3). I shall
expound the results of this comparison in a separate paper. There is no doubt that the opposition
Philoponus 1–Philoponus 2 appears likewise in the interpretation (eternalist versus literal) of
Plato’s cosmogony. I must reserve this matter too for later. But it has to be mentioned here that
Philoponus in his in Nicomachi Introductionem Arithmeticam gives up silently Ammonius’ (and his
own earlier) eternalist understanding of Plato’s cosmogony. This was established by L. Tarán,
Asclepius of Tralles, 10–12, 74 on the basis of a comparison between Asclepius in Nicom. I, 3, 68–9
and Philoponus in Nicom. I, 3, 54–8. On the other hand, there are elements in Philoponus’
Commentary on Nicomachus that recall ‘Philoponus 1’ rather than ‘Philoponus 2’. This could mean
that in this case too there was a later revision and adaptation of an earlier text.
286 Aristotle Transformed

By contrast, there certainly is an evolution properly so-called from the de

Aeternitate Mundi contra Proclum via the contra Aristotelem through to the
de Opificio Mundi. For, although this last (a commentary on the Hexaemeron) is
a theological work, it represents in certain respects the final stage of Philoponus’
philosophical thinking. I shall briefly indicate a few points to substantiate my
thesis. It is not unlikely that the de Aeternitate Mundi contra Proclum was written
rather quickly: the situation no doubt required prompt action. As a result, the
book is not free from inconsistencies and concealed aporias.
(1) For instance, there is some hesitation with regard to the problem of the
unity of divine will: is it a simple, unique will,193 or is it rather to be conceived as
invariance through successive acts of will?194 In principle Philoponus affirms the
former alternative, but actually he is forced to use the latter.
(2) A similar ambiguity is to be found in his concept of God’s intellective
activity. In principle divine thinking is conceived as a static ‘thinking all things at
once (panta hama noein)’;195 on the other hand, it is said not to exclude a
succession of ‘earlier (proteron)’ and ‘later (husteron)’,196 if that succession is not
understood as a temporal one.197
(3) A major aporia in de Aeternitate Mundi contra Proclum – I come to the
evolution of Philoponus 2 now – is the place of angels (transcendent intellects)
in the order of creation. There is no doubt, Philoponus says, that they have been
created by God:198 God is the efficient cause of intelligible substances.199 On the
other hand, one important new principle on which Philoponus 2 bases his
system is the principle of the posteriority of the effect to its (efficient) cause. This
means that an eternal being cannot have a cause. A being is either eternal or
created: eternal creation is impossible.200 Only God has no cause.201 The aporia is
clear: angels have been created; hence they cannot be eternal a parte ante. But in
Philoponus 1 the eternity of transcendent intellects had been so obvious202 that
in 529 he does not yet dare to deny their eternity outright. However, he no longer

193
e.g. Philoponus Aet. 81,12–25; 568,6–16; 573,13–574,12.
194
e.g. ibid., 80,26–7; 574,13–578,9; 583,23–584,3.
195
e.g. ibid., 575,21–2; cf. 568,21–8; 570,18–571,2. Philoponus 1 uses the formula to characterise the
activity of the demiurgic Intellect: in DA 126,26–7; 132,30–1.
196
Philoponus Aet. 583,12–23; cf. 114,19–116,1.
197
cf. ibid., 584,11–18.
198
ibid., 110,3–4; 575,13–14.
199
cf. ibid., 171,4–6.
200
ibid., 14,14–20; cf. 23,20–1; 473,23–5.
201
Philoponus ap. Simplicium in Phys. 1141,19.
202
Intelligible substances are autoenergeiai, i.e. pure acts. Cf. e.g. Philoponus in DA 35,1–2; 63,12–14;
94,18; 216,33–5; 297,10–11; in de Intell. 84,63–9 (autoactus, autoenergia); Ammonius in Int.
248,17–31; 250,34–251,1.
 The development of Philoponus’ thought and its chronology 287

explicitly calls them eternal either.203 It seems that in the de Aeternitate Mundi
contra Proclum he deliberately avoids deciding. In the contra Aristotelem his
position may have become clearer.204 In the de Opificio Mundi, he finally solves
the problem by siding with Basil’s view that creation is divided into two stages: a
purely angelic phase precedes the creation of the sensible world.205 Accordingly,
angels are not eternal a parte ante. God creates them at a certain ‘moment’ before
the creation of the sensible world;206 their coming into existence is a
‘supramundane coming into being (huperkosmios genesis)’.207
(4) Obviously, the problem with eternal created beings returns at the level of
rational soul in the broad sense of the term. I will first give some brief indications
concerning celestial souls. Philoponus 1 was convinced that in the case of
heavenly bodies (as opposed to sublunary bodies, both simple and composite)
there is an identity of natural and psychogenic movement. This theory had the
function of harmonising Plato and Aristotle, and Philoponus 1 ascribed it to
Aristotle himself.208 Actually he gave two different versions of it in the de
Anima.209 The first variant, taken from Alexander of Aphrodisias,210 was this:
heavenly soul just is the nature of heavenly bodies (phusis estin hê psukhê), i.e. it
is the only moving principle in the realm of the fifth element and it moves
automatically. The second variant was this: heavenly bodies have both a bodily
nature and a soul which differs from it: accordingly they possess a double
movement, one caused by their nature and another caused by their soul; however,
the two movements are identical, they coincide in a single, circular, movement.211
In the first case the identity natural movement-psychogenic movement212 is only
‘subjective’, i.e. it is we who say that the soul of the heavenly body is its nature or
principle of movement: there is no principle of movement other than soul, and

203
Philoponus Aet. 576,19–21: ‘It is possible, or rather necessary, that intellective substances should
exist, even when the world does not’, strictly speaking affirms only the pre-existence of intellective
substances compared to the material world, not their eternity. But the latter is, of course, implied.
204
Simplicius in Phys. 1159,18–19: ‘The eternity which he (Philoponus) thinks fitting only to God (. . .
tês aidiotêtos, hên houtos [i.e. Philoponus] monôi tôi theôi prosêkein oietai. . .)’.
205
Philoponus Opif. 16,19–17,3, quoting Basil Hom. in Hexaemeron PG 29, 13C. Philoponus defines
Basil’s position as follows: ‘That there was some condition of reality older than the creation of the
world, suited to the intellective powers above the cosmos’, 16,20–2.
206
Philoponus Opif. 26,21–27,6; 34,21–35,10; cf. 33,6–25; 45,10–12.
207
ibid., 25,22–3.
208
Philoponus in DA 138,21–3.
209
ibid., 101,34–102,31.
210
ibid., 102,1; cf. e.g. Alexander ap. Simplicium, in Phys. 1219,3–5; H.A. Wolfson, ‘The problem of
the souls of the spheres from the Byzantine commentaries on Aristotle through the Arabs and
St Thomas to Kepler’, Dumbarton Oaks Papers 16, 1962, 69, 72–5.
211
Roughly speaking this was the view of Simplicius; cf. e.g. in Cael. 387,12–19; H.A. Wolfson, op. cit.,
75–6.
212
Philoponus in DA 102,17.
288 Aristotle Transformed

consequently no movement other than psychogenic movement either. In the

second case the identity natural movement-psychogenic movement is objective:
the heavenly body has two different principles of movement, nature and soul,
which together produce a single movement. Philoponus 1 does not seem to have
a clear preference for either of these versions. This preference is only a matter of
secondary importance for him in comparison with what is common to both
versions, namely the conviction that the heavenly body and heavenly soul are
inseparably united.
In de Aeternitate Mundi contra Proclum the author maintains the existence of
heavenly soul(s). However, his view on the relation heavenly body-heavenly soul
is not the same anymore in this treatise as in in de Anima.213 The second version
of in de Anima has disappeared. Philoponus’ rejection of Aristotle’s theory of
aether214 made it impossible for him at this stage to consider the circular
movement of heaven as an effect of the peculiar nature of the heavenly body
itself any longer. On the other hand, the first version of in de Anima, paradoxical
as this may seem, could be maintained. Already in the earlier work it was said
(first version) that circular movement is natural for the heavens, while at the
same time the principle of this movement, namely soul, is a force superior to the
heavenly body itself (kreittôn dunamis).215 Hence heavenly movement is natural
(kata phusin), but not caused by nature (oukh hupo phuseôs, mê phusei).216 This
view was still usable in 529. It is argued now that circular movement is natural
for the heavens, not in so far as it is a body, but in so far as it is a living being,
animated by soul.217 As a body, the heaven is motionless;218 it has no principle of
movement on its own. It is moved only by soul, a force superior to body (kreittôn
dunamis).219 Although the principle of circular movement is supernatural for the
body of the heaven (huper tên tou sômatos autou phusin),220 the movement itself
is natural for the heaven as a living being.221 This view, which in in de Anima had
nothing to do with any doubt about the eternity of celestial movement, in de
Aeternitate Mundi contra Proclum unexpectedly offered the opportunity both for
affirming and for denying the eternity of heavenly soul, as I shall immediately

213
I thank Richard Sorabji for helping me to clarify my argument. cf. his Matter, Space and Motion,
London and Ithaca NY 1988, 241, 246.
214
Philoponus Aet. 492,5–493,24; 517,7–519,17.
215
Philoponus in DA 102,2.
216
ibid., 102,1; 102,18.
217
Philoponus Aet. 485,5–10 (Plato and Philoponus); 487,17–20; cf. 489,28–490,1.
218
ibid., 489,24–490,1.
219
ibid., 492,26–7
220
ibid., 492,27–493,1. For the whole argument in Aet. cf. É. Évrard, Les convictions religieuses, 309–14.
221
cf. n. 217 above.
 The development of Philoponus’ thought and its chronology 289

explain. The second version of in de Anima is deliberately cut out: in 529

Philoponus had not yet reached the point where he could conceive a nature
of the heavenly body itself which was (1) different from heavenly soul,
(2) responsible on its own for the circular movement of the heaven, (3) yet
different from the nature of Aristotle’s fifth element. Only in contra Aristotelem
will he be able again to regard the circular movement of the heaven as the effect,
not only of the presence of heavenly soul, but also of the bodily nature of the
heaven. In 529, by contrast, he is still of the opinion that the elements in their
natural places are motionless by themselves.222
In contra Aristotelem Philoponus still holds to the theory of heavenly soul, a
fact that provokes Simplicius’ surprise.223 As I have already indicated, he can fall
back now on the second view of his in de Anima: heavenly movement is caused
both by the nature of the heavenly body and by heavenly soul;224 there is an
identity between the natural inclination (phusikê rhopê) of the body and the
impulse caused by the soul (hê tês psukhês hormê), as in the case of a bird flying
straight down.225 In contrast to in de Anima, however, contra Aristotelem finds
that Aristotle is contradicting himself on the subject.226 In contrast to de
Aeternitate Mundi contra Proclum, on the other hand, Philoponus has now
succeeded in conceiving the nature of the heavenly body as something which is
different from heavenly soul, which is responsible on its own for the circular
movement of the heaven, and yet is different from the nature of Aristotle’s fifth
element. Apart from being caused by the soul, the circular movement of heaven
is also the effect of the very nature of its body. In this last respect there is no
difference between heavenly movement and the circular movement of the sphere
of fire and of the upper air, a movement which is natural for those elements as
totalities, in their natural place.227
In the de Opificio Mundi, finally, Philoponus completely gives up the doctrine
of celestial soul. What is the moving principle of heavenly bodies? It is impossible,
he argues, to say that they are moved by angels. For how could bodies of that size
continue to be moved over such distances by an unnatural (para phusin) moving
force? Why would it be impossible for God, who created heavenly bodies, at the
same time to put in them a natural force to make them move in a circle, a force

222
cf. É. Évrard, loc. cit.
223
Simplicius in Cael. 91,17–19.
224
ibid., 78,17–21.
225
ibid., 79,8–13.
226
ibid., 78,21–79,8; 79,13–14.
227
ibid., 34,5–11; cf. E. Evrard, op. cit., 318–20.
290 Aristotle Transformed

similar to the heaviness and lightness in bodies that move in a straight line?228 De
Opificio Mundi 6,2 is largely devoted to the refutation of the presence of souls in
heavenly bodies.229
The evolution of the later Philoponus with regard to the problem of heavenly
soul is clear: de Aeternitate Mundi contra Proclum accepts its existence, but does
not identify it with the nature of the heavenly body itself; contra Aristotelem both
accepts the existence of heavenly soul and identifies the effect of its activity with
the movement caused by the nature of the corresponding body itself; de Opificio
Mundi rejects the existence of heavenly soul and reduces the capacity for circular
movement to an inanimate nature.
Now, what has all this to do with the question of the eternity of celestial souls?
I think that the evolution just outlined is relevant to this problem as well. De
Aeternitate Mundi contra Proclum, first, in considering heavenly soul as a
principle which is supernatural to the heavenly body, clearly uncouples the
question of the eternity of the former from that of the eternity of the latter. This
uncoupling creates the opportunity of denying the eternity a parte ante of the
heaven without necessarily denying the eternity a parte ante of heavenly soul. I
cannot enter here in detail into Philoponus’ position with regard to the eternity
of world soul and heavenly souls in 529. It may be sufficient at present to point
out that the main problem in this respect is to determine how far Philoponus’
interpretation of Plato’s cosmogony and psychogony can be taken to reflect his
own views. There is some ambiguity. On the one hand Philoponus assumes that
for Plato the world soul is eternal a parte ante, which implies that it existed as a
pure intellect before it became the animating principle of the material world.230
This is the view which is predominant in de Aeternitate Mundi contra Proclum; it
is the guiding principle of the interpretation Philoponus 2 gives of the psychogony
of the Timaeus. On the other hand there is at least one text that creates the
impression that Philoponus does not exclude a possible coming into existence of
the Platonic world soul.231 Does this exegetical hesitation reflect indecision on
the part of Philoponus himself? I think it does. Anyhow, the uncoupling of
heavenly soul and heavenly body fortunately leaves room both for the eternity
and for the non-eternity of heavenly souls. Obviously, Philoponus’ problem in

228
Philoponus Opif. 28,20–29,9. Movement is either natural or unnatural (29,5–7). Both movement by
immanent souls (formerly ‘supernatural’) and movement by external intellects (angels) are
considered now ‘contrary to nature (para phusin)’.
229
Philoponus Opif. 231,3–233,17.
230
Philoponus Aet. 195,13–196,19.
231
ibid., 267,16–268,10.
 The development of Philoponus’ thought and its chronology 291

this case was the same as in the case of transcendent intellects (angels). The
revised concept of efficient causality excluded the eternity of heavenly souls. But
in the de Aeternitate Mundi contra Proclum Philoponus still seems to have
recoiled from drawing this conclusion at once. In the contra Aristotelem, by
contrast, he may have cut the knot. The psychogenic movement of heaven is
again identified here with the natural movement of the heavenly body. Obviously
this movement cannot be eternal a parte ante. But what about the eternity of
heavenly soul itself? Strictly speaking the soul might be eternal without animating
the body from eternity. Yet the union between body and soul is closer now than
it was in de Aeternitate Mundi contra Proclum. The reintroduction of the nature
of the heavenly body as the co-principle of its movement clearly tends to make
heavenly soul superfluous,232 to reduce it to a redundant inheritance from the
past. This depreciation must have affected its ontological status as well, including
its eternity, if we can rely on Simplicius’ remark that Philoponus reserved eternity
for God.233 From here it was only one step to the final solution in de Opificio
Mundi: it is just the nature of the heavenly body, a nature created by God, which
is the principle of its movement.
(5) Only with regard to the human rational soul does Philoponus venture to
draw the inevitable conclusion in 529: it cannot be eternal a parte ante. The
reason referred to is that it is subject to evil (kakia),234 i.e. that it has the capacity
for turning to a state contrary to its nature,235 that it is characterised by natural
weakness.236 Everything that is capable of evil is perishable, and therefore it must
have come into existence.237 This means, of course, a decisive break with the
Neoplatonic psychology of Philoponus 1: just like the sensible world, the soul
has been created by God ex nihilo. However, another important question remains
unanswered in the de Aeternitate Mundi contra Proclum: if the soul is generated,
does it come to be together with or after the body, or is it pre-existent to it? The
author discusses the subject in de Opificio Mundi 1,10, in connection with the

232
In the system of Philoponus 1, the function of heavenly souls was both kinetic and hierarchical.
233
cf. n. 204 above.
234
Philoponus Aet. 468,26–469,5: ‘First it is unclear – to say only this much now – whether each of the
things mentioned (i.e. matter and soul) did not have a beginning of its existence. For, as far as I
know, there is no argument necessitating (the assumption) that they had no beginning. In another
Book (i.e. Book 9: cf. 336,1–337,13) we have sufficiently discussed the fact that no being which by
its nature is capable of deficiency, such as our soul, can have a beginningless existence.’
235
ibid., 302,25–303,2; 336,5–8.
236
ibid., 302,25–303,25 passim; 336,1–337,13 passim.
237
ibid., 336,1–337,13. Philoponus does not enter into the obvious problem of how to save, on these
premises, the immortality of the rational soul.
292 Aristotle Transformed

generation of angels. The human soul is certainly created after the creation of the
world.238 It does not come to be together with the body.239 But is it created after
the conception of the body or is it pre-existent to it? Philoponus’ view on this
point is not clear, as the text seems to contain arguments for both interpretations.
(6) Another issue on which Philoponus 2 later evolved a view different from
that of de Aeternitate Mundi contra Proclum is the end of the world. In the de
Aeternitate Mundi contra Proclum he seems to draw no distinction yet between
the end of the cosmos sensu stricto (i.e. of the present formal structure of the
sensible world) and the end of the world sensu lato (i.e. form and matter): the
world will perish.240 How does he account for this statement? The ultimate
ground for it is not cosmological. In spite of Simplicius’ contention,241 Philoponus
does not infer the necessity for the world to perish on the basis of its intrinsic
finiteness, neither in the de Aeternitate Mundi contra Proclum nor in the third,
non-polemical work. To Philoponus this finiteness only proves that the world
may perish, it does not prove that it will.242 De Aeternitate Mundi contra Proclum
has indeed a few passages which, merely hypothetically, envisage the possibility
that the world will not perish for a supernatural reason, namely the will of the
Creator.243 This hypothesis is to be found in the third treatise as well.244 Yet
Philoponus is convinced that the world must perish, not on cosmological, but on
theological grounds. If the Creator, he writes, is able to dissolve the world without
ever actually dissolving it, he will always be in potentiality in this respect, never
in actuality. This means that in this respect he will always be imperfect, which is
impossible. An eternal being cannot forever remain in potentiality. Consequently,

238
Philoponus Opif. 26,23–4.
239
ibid., 25,6–9; 26,5–6.
240
Philoponus Aet. 312,7–16: ‘So we acknowledge both that the world was generated and that it will
perish, having Plato’s vote as well as the truth in support of its having been generated, but in support
of saying that having been generated it will perish, come what may, and will not be undissolved and
immortal, as Plato thinks . . . (Hôste hêmeis kai gegonenai ton kosmon kai palin phtheiresthai
homologountes, hoti men gegonen, sumpsêphon ekhomen meta tês alêtheias kai Platôna, hoti de
genomenos pantôs phtharêsetai kai oukh, hôs Platôni dokei, alutos estai kai athanatos . . .)’. There is no
doubt that kosmos here refers to the world sensu lato, since the term gegonenai obviously means
‘generation ex nihilo’.
241
Simplicius in Phys. 1327,15–16: ‘He (Philoponus) thought that what has finite power is at once
shown to be perishable.’
242
[For a different analysis of Philoponus, see Chapter 9 – Ed.]
243
Philoponus Aet. 300,17–21 (at 300,20 read athanasian for astheneian Rabe; cf. 300,11; 301,1);
312,21–6; 630,12–17. W. Wolska, La Topographie Chrétienne, 186 from Aet. 312,21–6 erroneously
infers: ‘In the final analysis, Philoponus falls in with his Greek opponents and acknowledges the
indestructibility of the world, with qualifications . . . He infers the indestructibility of the heavens,
but attributes it to the intervention of a force majeure, the will of God.’ It is true that Philoponus
holds in the contra Aristotelem that the world is unperishable. But he is talking then about another
‘cosmos’, not about this one, as the Hellenes do.
244
Philoponus ap. Simplicium in Phys. 1333,23–7.
 The development of Philoponus’ thought and its chronology 293

God must dissolve the world.245 The cosmological version of this theological
argument would be: the world contains in itself the potentiality for being
dissolved, and this potentiality has to actualise itself. But Philoponus nowhere
defends this proposition. Cosmology cannot prove that the world will never
perish,246 nor that it must perish either. On Philoponus’ premises, the necessity
for God to dissolve the world at some time should apply to matter as well.247 In
the contra Aristotelem, on the other hand, Philoponus was convinced that the
end of the present cosmos would not be the end of the world: it would only be
the transition to a more perfect, a more divine sensible world.248 We are left to
conclude that the view of contra Aristotelem is different from that of de Aeternitate
Mundi contra Proclum, namely that the sensible world (form and matter) will
necessarily perish. The present cosmos, the components of which are all
perishable by nature, will be replaced, it is said, by a new order of matter. This fits
in with what we know to have been the doctrine of Philoponus’ de Resurrectione:
God will create new, imperishable bodies, and there will be a new world.249
These examples are sufficient, I think, to show that in the de Aeternitate Mundi
contra Proclum the system of Philoponus 2 was not yet finished. The book was
conceived starting from one idea: anti-eternalism. It is remarkable to see how
this tenet really had to make its way through Philoponus’ Alexandrian
Neoplatonic heritage, turning the whole system upside down. The result shows
that this was not attainable at once. By 529 our author was not yet completely in
line with Christian doctrine,250 nor did he know the Bible as well as he did in the
de Opificio Mundi.251 Apparently it was not so easy for him to Christianise his
original world view. Some Christians must have felt that he was not entirely
successful, even in his de Opificio Mundi. To Cosmas Indicopleustes he remained

245
Philoponus Aet. 131,26–132,28.
246
Philoponus ap. Simplicium in Phys. 1333,27–30: phthoras de ekhonta logon ouk estin apeirodunama.
Pothen gar hoti ho logos tês phthoras houtos . . . oukh hêxei pote kai eis energeian? I think we have to
alter Diels’s punctuation, here as well as in 1334,37–9, in order to avoid ambiguity. In Diels’s reading
(Pothen gar? Hoti ho logos . . .) the hoti-sentence seems to give the answer to the question pothen gar
(‘whence?’), which would give an absurd meaning. If we read Pothen gar hoti . . . the hoti-sentence
is clearly a subject clause. ‘Whence is it that that principle of passing away will never come to
actuality?’, i.e. ‘Why . . . not?’
247
cf. Philoponus ap. Simplicium, in Phys. 1177,22–6.
248
Simplicius in Phys. 1177,38–1178,5. See now also C. Wildberg, ‘Prolegomena’, 198–200.
249
cf. R. Sorabji, ‘John Philoponus’, 32–3.
250
It is noteworthy that in de Aeternitate Mundi contra Proclum Philoponus seems not yet to accept the
Christian view that creation took some time (‘six days’): cf. 367,9–18. In contra Aristotelem, by
contrast, he seems to have understood creation in the sense of the biblical Hexaemeron: cf.
Simplicius in Phys. 1174,22–5.
251
cf. Philoponus Aet. 75,19–20: ‘As the Holy Scriptures say somewhere’ (hôs pou ta hiera phêsin logia).
294 Aristotle Transformed

a pseudo-Christian, someone whose ideas were too ‘Hellenic’.252 In de Aeternitate

Mundi contra Proclum the theological elaboration of anti-eternalism clearly runs
behind cosmology. The cosmological system itself was not complete at once
either. This is shown by such later developments as that concerning the cause of
the movement of the heaven and elemental fire. Moreover, there might be a
connection between Philoponus’ new theory on matter in the de Aeternitate
Mundi contra Proclum (which is related to the idea of a creation ex nihilo) and
his new theory on void in the in Physica. This remains to be investigated.
Finally, let us return briefly to the duality Philoponus 1–Philoponus 2. This
duality does not, of course, exclude there being a lot of continuity between the
two systems, not only in respect of vocabulary and stylistics,253 but also on a
number of minor issues, for instance of natural philosophy. Further, it is
important to realise that the distinction between Philoponus 1 and Philoponus
2 at the level of doctrine is independent of my chronology. In other words: even
if the proposed chronology should prove to be false, the duality of doctrine in
Philoponus’ writings remains what it is. One would then only have to look for a
better explanation. We have to take both systems seriously: up to now students
of Philoponus have not only overlooked the Alexandrian Neoplatonism of
Philoponus 1, but also identified the philosophy of Philoponus 2 too easily with
his dogmatic theology. I think that we first have to study both philosophies just
as they present themselves, irrespective of the fact that Philoponus 1 is more
Ammonius than Philoponus, irrespective also of the motives of Philoponus 2.
For a philosophy is not just a subject’s product; it is a figure of Thought as well.254

252
cf. W. Wolska, La Topographie Chrétienne, 150.
253
I merely draw attention to the close similarity between e.g. Philoponus in DA 116,1–21 and Aet.
196,25–197,15, and in DA 297,1–10 and Aet. 65,11–22.
254
I am particularly indebted to Richard Sorabji for his suggestions for improving my text.
 12

The life and work of Simplicius in Greek

and Arabic sources*
Ilsetraut Hadot

In this attempt to make an inventory of research on the life and works of

Simplicius, I am conscious of the fact (and even rejoice in it) that some fifteen
years hence this inventory will be obsolete, at least in part. For the time has
finally arrived when we are studying Simplicius’ commentaries for themselves,
rather than restricting ourselves to several citations of the Presocratics and to his
Corollary on Time. Several years ago, Philippe Hoffmann undertook a new
critical edition of Simplicius’ commentary on Aristotle’s de Caelo with an
annotated French translation; Leonardo Tarán is undertaking a new critical
edition of the commentary on the Physics with an English translation; and I
myself expect to finish, in the very near future, my own critical edition of the
commentary on Epictetus’ Encheiridion with an annotated French translation.
Henry Blumenthal has devoted himself to an interpretation of Simplicius’
commentary on the de Anima and will publish an English translation of it within
the framework of the great enterprise directed by Richard Sorabji, which has as
its goal the translation into English of a large part of the commentaries on
Aristotle edited in the Berlin Academy’s Commentaria in Aristotelem Graeca.
Concetta Luna, under the direction of Professor F. del Punta, has already finished
a translation of Simplicius’ commentary on the Categories into Italian, and a
French annotated translation of the same commentary has been undertaken by
Philippe Hoffmann, Pierre Hadot, and myself.1 Sylvia Donati (Books 1–2) and
Cecilia Trifolgi (Books 3–8) of the University of Pisa are in the process of
translating Simplicius’ commentary on the Physics into Italian.

* Translated by Victor Caston, in the light of minor revisions by the author.

1
Simplicius, Commentaire sur les Catégories, traduction commentée, fascicles 1 and 3, Brill, Leiden,
forthcoming 1989.

295
296 Aristotle Transformed

I have no doubt that the sum of this research, by making us genuinely familiar
with the thought of Simplicius and the other Neoplatonic commentators on
Aristotle, will considerably modify the accepted image that K. Praechter first
constructed concerning Simplicius’ relations with the so-called Athenian and
Alexandrian schools, and will especially modify the received opinions on the
Alexandrians. Praechter claimed, of course, that the Alexandrian school studied
Aristotle for his own sake, interpreting him in his true sense, whereas the
Athenian school applied a Neoplatonic interpretation to the few works of
Aristotle in which they were interested.2 These ideas have been accepted not only
by philologists of classical languages and historians, but also by Arabists and
Orientalists. Nevertheless, a cursory glance at the commentaries on the
Metaphysics written in part by Syrianus (of the Athenian school) and in part by
Ammonius-Asclepius (of the Alexandrian school), ought to have convinced
K. Praechter of the contrary, since the tendency in the two commentaries is
appreciably the same, and Ammonius-Asclepius’ commentary even borrows a
complete portion of Syrianus’ refutation of Aristotle.3 It is evidence that the
critique of the Pythagorean doctrines on numbers and of Plato’s doctrine on the
Ideas, which Aristotle carries out in several places in the Metaphysics, could not
as such be accepted by any Neoplatonist, whether he taught at Athens or at
Alexandria, whereas there was no insurmountable difficulty concerning
Aristotle’s logical and physical teaching.4 To cite Syrianus himself: ‘I am not one
of those who are hostile to Aristotle, nor one of those who claim the authority of
Aristotle’s teaching on details or trivial matters, but one who thoroughly admires
his lessons on logic and who enthusiastically approves of his ethical and physical
treatises’.5 In fact, the whole of Aristotle’s logical, ethical, and physical treatises
had been definitively incorporated into Neoplatonic teaching6 at the latest with

2
cf. F. Überweg, Grundriss der Geschichte der Philosophie, Part I: K. Praechter, Die Philosophie des
Altertums, 12th edition, Berlin 1926, p. 638: ‘Iamblichus referred to the noera theôria [intellective
theory] everywhere in his commentary on the Categories, and Simplicius, who informs us on this
matter (in Cat. 2,13), did the same as he. In contrast, the Alexandrian commentators on Aristotle’s
logical works restricted themselves to a sober explanation of the text in accordance with its natural
sense.’ Cf. also K. Praechter, ‘Richtungen und Schulen im Neoplatonismus’, in Genethliakon für Carl
Robert, Berlin 1910, pp. 105–56.
3
Asclepius in Metaph. 432,29–436,8. This citation shows that Syrianus must have commented on
Aristotle’s Metaphysics in its entirety, since it corresponds to Book 7, for which Syrianus’ commentary
no longer exists.
4
For a more ample discussion of this matter, cf. I. Hadot, Le problème du néoplatonisme Alexandrin:
Hiéroclès et Simplicius, Paris 1978, pp. 195ff.
5
Syrianus in Metaph. 80,4–7.
6
For Porphyry, cf. F. Romano, Porfirio e la fisica aristotelica (= Symbolon, Studi e testi di filosofia
antica e medievale III ), Catania 1985, the chapters ‘L’interesse di Porfirio verso Aristotele’ and ‘I
commentari aristotelici di Porfirio’.
 The life and work of Simplicius in Greek and Arabic sources 297

Syrianus-Proclus, though probably already with Porphyry, who had himself

commented, at least partially, on Aristotle’s logical works, Physics, de Anima,
Book 12 of the Metaphysics, and on one ethical treatise.7 More than any other
works, the different commentaries on the Categories will probably furnish the
crucial material which will best allow revision of Praechter’s global views, since
we have the opportunity to compare Simplicius’ commentary with six other
Greek commentaries, namely those of Porphyry, Dexippus, Ammonius,
Philoponus, Olympiodorus, and David8 and still further with Boethius’ Latin
commentary. Furthermore, Simplicius’ commentary introduces us to large parts
of the lost commentaries on the Categories, namely, Porphyry’s (addressed to
Gedalius) and Iamblichus’, as well as to several not insignificant remarks from
the lost commentary on the Categories by Syrianus, Proclus’ teacher. Of these
commentaries, four – by Ammonius, Philoponus, Olympiodorus and David –
belong, according to Praechter’s views, to the rational and sober interpretation of
the Alexandrian school, whereas Simplicius’ commentary followed the Athenian
tendency of Syrianus and Proclus, who themselves were inspired by Iamblichus.
Having worked through the first forty pages of Simplicius’ commentary by
comparing them with the four so-called ‘Alexandrian’ commentaries, Philippe
Hoffman, Pierre Hadot and I did not encounter a single piece of evidence which
favoured Praechter’s thesis, that is, we were unable to establish any difference
concerning the interpretative tendency of Simplicius’ commentary, on the one
hand, and the commentaries of Ammonius, Philoponus, Olympiodorus and
David, on the other. Similarly, A. Sheppard,9 in pointing out parallels between
Proclus’ commentary on the Cratylus and Ammonius’ commentary on the de
Interpretatione, concludes that all these parallels between Proclus and Ammonius
show how close the connections between Athenian Neoplatonism and

7
cf. I. Hadot, ‘Les introductions aux commentaires exégètiques chez les auteurs néoplatoniciens et les
auteurs chrétiens’, in M. Tardieu (ed.), Exégèse et philosophie, Paris 1987, pp. 99–122. For the
importance of Porphyry as a commentator of Aristotle, cf. F. Romano, op. cit. in n. 6.
8
A. Busse (Eliae in Porphyrii Isag. . . ., CAG 18.1, pp. viiff ), who was followed by Y. Manandean (L’école
hellénisante . . ., Vienna 1928, in Armenian, pp. 25 and 39), claimed to show that, despite the
unanimous Greek and Armenian manuscript tradition, the author of the commentary on the
Categories could not have been the author of the commentary on the Isagoge, but must have been
the philosopher Elias. This thesis has been refuted point by point by S.S. Arevšatjan in his article,
‘L’héritage de David l’Invincible sous une lumière nouvelle’, in Banber Matenadarani IX , 1969, 7–22
(in Armenian with a French abstract). The expert on Armenian literature, J.-P. Mahé, who had the
kindness to bring these works to my attention, considers this refutation decisive. Cf. J.-P. Mahé,
‘David l’Invincible dans la tradition arménienne’, forthcoming in Simplicius, Commentaire sur les
Catégories, traduction commentée, Appendix II , Brill, Leiden, 1989.
9
cf. A. Sheppard, ‘Proclus’ philosophical method of exegesis: the use of Aristotle and the Stoics in the
Commentary on the Cratylus’, in J. Pépin and H.D. Saffrey (eds), Proclus – lecteur et interprète des
Anciens,Xy colloques internationaux du CNRS , Paris 1987, pp. 137–51.
298 Aristotle Transformed

Alexandrian Neoplatonism are, to the extent that she doubts the justice of the
view that there is a significant difference between the Athenian and Alexandrian
attitude towards Aristotle. In her opinion, the tone of the later Neoplatonic
commentaries is determined more by the text to be commented on than by the
‘school’ to which the commentator belonged. Let us cite Ammonius himself, who
considered his commentary on Aristotle’s de Interpretatione to be merely the
fruit of his teacher Proclus’ instruction on the same subject: ‘In the eyes of the
wise, Aristotle’s de Interpretatione is a work of great importance and has no little
honour thanks both to the density of theories expounded in it and the difficulty
encountered in its literal explanation (lexis). This is why many exegetes have
written many reflections on the subject of this work. If we should be able to
contribute anything to the elucidation of this book ourselves while recalling the
exegeses of our divine teacher Proclus, the Platonic Diadochus, who practised to
the heights that human nature can attain both the ability to explain the views of
the ancients and to distinguish scientifically the nature of things, we would
greatly thank the God of language.’10 Thus, until the contrary is proven, I will
reaffirm the conclusions expressed in my book, Le problème du néoplatonisme
Alexandrin: Hiéroclès et Simplicius, namely, that there was no Neoplatonic school
in Alexandria whose doctrinal tendencies differed from those of the school at
Athens.11 This uniformity of tendency does not imply a uniformity of system in
all its details: there is no uniformity of system between Plotinus and his pupil
Porphyry, between Porphyry and his pupil Theodore of Asine, between
Iamblichus and Proclus, or between Damascius and his pupil Simplicius, but
rather a uniformity of tendencies, and it is this same uniformity of tendencies
which in my opinion ties the schools of Athens and of Alexandria closely
together. The doctrinal differences which may appear do not result from
divergences between schools, but from the divergences between persons.
If I am now speaking about the life and works of Simplicius again after having
devoted a chapter of my book Le problème du néoplatonisme Alexandrin to this
subject eight years ago, it is because in the intervening period new research has
been conducted which seems to me capable of enriching our knowledge
considerably on this subject. In fact, whatever I was then able to say about

10
Ammonius, in Int. 1,3–11. Ammonius also follows Proclus’ interpretations in his commentary on
the Prior Analytics, which is only partially extant (apart from an edition of his lectures made by
Philoponus): cf. CAG 4, 40,16ff, where he cites Proclus who had given him a course on the Prior
Analytics following the commentary of Iamblichus quite strictly, and 43,30f, where he cites the
skholikon hupomnêma on the first book of the Prior Analytics.
11
Paris 1978. In the same book, on pp. 9ff, I likewise explained the sense that must be given to the
phrases ‘the School at Athens’ and ‘the School at Alexandria’, if the use of these names is to continue.
 The life and work of Simplicius in Greek and Arabic sources 299

Simplicius’ life was very slight: according to the statements of Agathias,12

Simplicius, a native of Cilicia in Asia Minor, attended the course of Ammonius,
the son of Hermeias, in Alexandria. This information is given to us by Simplicius
himself.13 Now, Ammonius taught in Alexandria between approximately 475 and
526. Simplicius also speaks of Damascius as his professor,14 but he specifies
neither the date nor the place of his studies under Damascius. Nevertheless, both
teacher and pupil are mentioned together by the historian Agathias,15 who states
that, ‘Damascius the Syrian, Simplicius the Cilician, Eulamius (or Eulalius) the
Phrygian, Priscianus the Lydian, Hermias and Diogenes both from Phoenicia,
Isidore of Gaza, the most noble flower . . . of our time,’ had fled the Byzantine
empire because of the persecution of pagans and had made their way to the
court of the Persian king, Chosroes. The latter, again according to Agathias,
permitted them to return to their homeland and to live out the rest of their lives
there without fear and without being required to abjure their philosophical
convictions, by having a special clause included in the peace treaty about to be
concluded with Justinian in ad 532. A turn of phrase used in the text of the
treaty (as reported by Agathias) raises several difficulties. Here is the clause in
question: ‘. . . dein ekeinous tous andras es ta sphetera êthê kationtas bioteuein
adeôs to loipon eph’ heautois . . .’16 In this clause, eph’ heautois cannot do double
duty with eis ta sphetera êthê kationtes (returning to their homeland) and cannot
therefore mean ‘in their homeland’. Referring to this clause, Westerink17 claims,
‘The ban on teaching activities, it should be noted, was maintained formally: . . .
bioteuein adeôs to loipon eph’ heautois,’ apparently because he understands eph’
heautois in the sense of ‘in undisturbed privacy’ and as indicating that all teaching
activity remained forbidden to philosophers. Formerly, I accepted this view of
Westerink’s.18 Today, in light of research summarised below, I would rather
translate eph’ heautois as ‘according to their wishes’, that is to say, ‘these men, in
returning to their homeland, must be able to live there according to their wishes
without fear for the rest of their lives, without being compelled to think anything
which might be in contradiction with their views or to change the beliefs of their
ancestors’. Neither Agathias nor any other Greek source tells us from which city

12
Agathias, Hist. 2.30.3 (Keydell, p. 80.7).
13
Simplicius, in Cael. 462,20.
14
For references, cf. the Index nominum of Diels’ edition of Simplicius, in Phys. under the name,
‘Damaskios’.
15
Agathias, Hist. 2.30.3ff (Keydell, p. 80.7ff ).
16
Agathias, Hist. 2.31.4ff (Keydell, p. 81.17).
17
L.G. Westerink, Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962, p. xv =
Chapter 14 below, p. 331.
18
In Le problème du néoplatonisme Alexandrin, p. 24 and n. 32.
300 Aristotle Transformed

in the Byzantine empire they came nor if they lived together before their stay in
Persia. There is, however, a strong probability, though no certainty, that they
came from Athens – the reference is probably to the last successor in the private
Platonic school in Athens, Damascius and his students. Agathias paints on the
whole a rather malicious picture of these pagan philosophers by attributing
naïve, utopian, and quickly disappointed ideas about Persia to them.19 In reality,
the decision taken by our pagan philosophers to emigrate to Persia was not at all
a matter of a utopian dream, but on the contrary had a very realistic basis. For
quite some time, they had before them the example of Nisibis, a Nestorian
university formerly established at Edessa in Syria. Already before 532, it enjoyed
considerable freedom of thought under the Persian kings, which contrasted
favourably with Byzantine intolerance, from which the theological school had
had to flee to Persia. This heterodox Christian school was a bastion of Greek
culture mediated by the Aramaic language, which was spoken on both sides of
the border. It may be supposed that Damascius, Hermias, Diogenes and Isidore
knew this language, since they were natives of Syria, and probably also Simplicius
at least, who originated from a country bordering on Syria. Persia thus for a time
combined all the favourable conditions to which our philosophers could aspire:
tolerance in religious matters and provision of a centre of Greek culture.
Agathias tells us nothing further about the place where the philosophers
made their way after their stay with Chosroes. He only informs us that they
undertook the return journey together, at least as far as the Byzantine frontier.20
It must also be noted that Agathias apparently had knowledge of the manner in
which they lived afterwards, since he says in this regard21 – and one senses a bit
of envy piercing through his summary – ‘. . . they took advantage of their exile,
not in a small and negligible way, but so as to spend the rest of their lives in the
most agreeable and pleasant manner.’ Such an expression does not suggest that
our philosophers were forced to live in ideological isolation, merely tolerated by
a hostile environment. I have discussed elsewhere22 the reasons which impelled
me to reject the hypothesis that Simplicius, if not the other philosophers, might
have returned to Athens or Alexandria. Rather, I then envisaged the possibility
that he had retired to a city of Asia Minor situated perhaps in Cilicia, his
homeland. Yet with all these hypotheses, one cannot imagine by what miracle

19
Agathias, Hist. 2.30.3–31.2 (Keydell, pp. 80.5–81.12); cf. I. Hadot, Le problème du néoplatonisme
Alexandrin, p. 24.
20
Agathias, Hist. 2.31.5ff (Keydell, pp. 81.21ff ).
21
ibid. 2.31.3 (Keydell, pp. 81.13ff ).
22
Le problème du néoplatonisme Alexandrin, p. 26ff.
 The life and work of Simplicius in Greek and Arabic sources 301

these philosophers could even have safeguarded their lives far away from the
protection of the Persian king, let alone led the type of joyous life of which
Agathias speaks. For, in reality, what does a clause of a treaty mean against an
omnipotent Christian church which could always organise lynchings, as in the
case of the philosopher Hypatia? Could our philosophers have been reckless and
naïve enough to trust a piece of paper fated to become worthless far from the
armed forces capable of guaranteeing it? An answer to these questions has
recently been provided, not by a philologist of classical languages, nor by a pure
historian, but by a specialist on gnosticism in the person of Michel Tardieu:
working from Arabic texts and interpreting them in light of Greek philosophical
literature as well as the history of Manicheanism, he arrived at the conclusion
that, in all probability, our Neoplatonic philosophers made their way after their
stay in Persia to H.arrân (Carrhae), a Byzantine city of Greco-Aramaic tongue
where ‘Hellenism’ had remained powerful and active, situated quite close to the
Persian border, 30 kilometres SSE of Edessa. Tardieu discussed his arguments in
1984 at the 6th International Congress on Gnosticism, which was held at the
University of Oklahoma, and also in an article entitled ‘S.âbiens coraniques et
“S.âbiens” de H.arrân’, in the Journal Asiatique. In what follows, I will attempt to
give a summary of it.23
The city of H.arrân, or Carrhae, is known to Latinists and Hellenists mainly
for two incidents: the Roman army was defeated by the Parthians in 53 bc in the
vicinity of Carrhae, and its commander, the Proconsul of the province of Syria,
L. Licinius Crassus, retreated into the city with the rest of his troops and was
assassinated. In this same city the emperor Julian in the year 363 ad offered
sacrifices in accordance with local rituals in the famous temple of the Moon,
before being defeated and killed some months later by the Persians. In the eyes
of the Christian Theodoret of Cyrrhus, the visit to the pagan city of H.arrân was
an intentional affront on Julian’s part to the neighbouring Christian city of

23
M. Tardieu’s article has appeared since this chapter was written (Journal Asiatique 274, 1986, 1–44),
but nevertheless I believe it useful to retain the summary of it which I gave at the colloquium on
Simplicius. For scholars and students of classical languages and ancient history or philosophy, the
Journal Asiatique is only rarely to hand. Furthermore, the summary which I offer of Tardieu’s article
concentrates on facts which, in the text and in the notes, relate to Simplicius and to the Neoplatonic
school in H . arrân, leaving aside an important portion bearing on gnosticism and the S.âbians of the
Koran.
If the hypothesis that the Greek philosophers Damascius, Simplicius, and the others mentioned by
Agathias remained in H . arrân after their stay in Persia is already rendered more than probable by
Tardieu’s article, then another discovery by the same author renders it almost certain. See his ‘Les
calendriers en usage à H.arrân d’après les sources arabes et le commentaire de Simplicius à la
Physique d’Aristote’, in I. Hadot (ed.), Simplicius – sa vie, son oeuvre, sa survie, Actes du colloque
international de Paris, Berlin and New York 1987, pp. 40–57.
302 Aristotle Transformed

Edessa. For Orientalists, H.arrân is the city par excellence of the S.âbians (not to
be confused with the Sabaeans) known to the Arabic sources. Who were the
S.âbians of H.arrân? I cannot enter into all the details of Tardieu’s extensive and
judicious argument, but must limit myself to the selection of a few points which
are of particular interest to historians of Neoplatonism. For convenience, I will
begin at the ending, that is, at the destruction of the last pagan temple in H.arrân
in 1081, a date after which there is no longer any mention of S.âbian activity in
this town. Yet in the tenth century S.âbianism is still alive in H.arrân, since the
Caliph cAbd al-Karîm confirms by letter to the S.âbians of H.arrân, Raqqa and
Diyâr Mud.ar their right to worship and education. And it was in the tenth
century, in 943, that the Arab author al-Mascûdî made his way to this town. He
recounted his impressions later on in his book Kitâb murûǧ al-d.ahab wa-macâdin
al ǧauhar. al-Mascûdî’s account contains an essential detail which modern,
contemporary criticism had not raised before Tardieu. The following is the text
in question (from the French translation of C. Pellat): ‘I have seen in H.arrân, on
the door knocker of the S.âbians’ meeting place, an inscription in Syriac
characters, taken from Plato; it was explained to me by Mâlik b. cUqbûn and
other persons of the same sect: “Whoever knows his nature becomes a god.” It is
also Plato who said, “Man is a celestial plant. Indeed, man resembles an upside-
down tree, whose roots are turned towards the heavens and whose branches
[plunge] into the earth”.’24 Chwolsohn, the author of a book entitled Die Ssabier
und der Ssabismus (St. Petersburg 1856, vol. 2, p. 826) had mentioned a correction
by Fleischer of the translation of the inscription found on the door knocker, a
correction whose import had passed totally unnoticed: ‘Who knows his own
essence (who knows himself) becomes divine, resembling God.’ No one before
Tardieu had identified these two citations, of which one can be found in the
Timaeus (90A7–B2) and the other is obviously an allusion to the Alcibiades I
133C. In another book, the Kitâb al-tanbîh wa-l-išrâf, al-Mascûdî returns once
more to this axiom in these words: ‘And we have stated what they [= the S.âbians
of H.arrân] think of Plato’s saying: “Whoever knows himself in truth becomes a
god”, and this other saying of the author of the Logic: “Whoever knows himself
knows all things.” ’ As Tardieu tells us, ‘in Arabic this small text does not include
a main clause. It is part of a long enumeration . . . of the different matters treated
by the author in two of his previous works, no longer extant today, but for which
al-Mascûdî provides the titles. One was entitled Book of Tenets on the Principles
of Religions; the other concerned The Treasures of Religion and the Secret of

24
M. Tardieu, ‘S.âbiens coraniques’, p. 13.
 The life and work of Simplicius in Greek and Arabic sources 303

Worlds.’25 Thus, in sum, al-Mascûdî spoke of the Platonic axiom of the Alcibiades
in at least four of his books. ‘If he returned to it rather frequently, it is because he
had grasped perfectly that this axiom was fundamental to his H.arrânian
interlocutors and just by itself summarised their entire philosophy.’26 According
to the summary of the two lost books that al-Mascûdî gives in the Kitâb al-
tanbîh, ‘it appears that the context of the citation from Plato in these two works
concerned reports relating to the “Greek S.âbians”. The ideas which the author
attributed to them show that this denomination, encompassing that of the
“H.arrânian S.âbians”, is to be taken here in a strictly philosophical sense, as
designating “the Platonists”. Not only is Porphyry’s name cited, but also that of
his correspondent,“the Egyptian priest Anâbû [Anebo]”. The doctrinal differences
between Porphyry and Anâbû have been recorded, al-Mascûdî claims, in epistles
“known to whoever is interested in the ancient sciences”.’27
‘In the immediate context of the citation of the Platonic axiom in the al-
Tanbîh, another citation appears which is attributed by al-Mascûdî to “the author
of the Logic”. This expression, as is so often the case with Arabic authors, does not
designate Aristotle himself, but one of his Alexandrian commentators. In fact,
the axiom: “Whoever knows himself knows all things”, of which al-Mascûdî gives
the Arabic version, is Neoplatonic. It occurs word for word in two Alexandrians,
namely, in Hermeias’ in Phaedrum: “Who understands himself knows everything
(ho heauton gnous ta panta oiden)” and in Olympiodorus’ in Alcibiadem: “Who
knows himself knows all things (ho eidôs heauton ta onta panta oiden)”, which in
both cases are juxtaposed with the Delphic precept. The association of the two
axioms in the al-Tanbîh is in line with the traditional interpretations of “Know
Thyself (gnôthi sauton)” given by the Neoplatonists when commenting on the
Alcibiades. The “S.âbians of H.arrân”, who explained to al-Mascûdî the Syriac
inscription engraved on the knocker of their front door and who considered
themselves to be “Greek S.âbians”, are nothing other than “Platonists” in the strict
sense’,28 or rather, Neoplatonists.
According to the direct testimony of al-Mascûdî, the building whose front
door knocker bore the central axiom of Alcibiades I engraved in Syriac was
called ‘maǧ mac’ in Arabic. This was, therefore, where the philosophers of H.arrân
met. ‘al-Mascûdî grouped the H.arrânians into two categories: the philosophers
. . . “of a low and vulgar level”, partisans of the pagan religion of the city, and the

25
ibid., p. 15.
26
ibid.
27
ibid., pp. 15–16.
28
ibid., p. 16.
304 Aristotle Transformed

“sages in the strict sense”, the heirs of the Greek philosophers. “By associating
them with the philosophers,” the author further specifies, “we have thought not
of their wisdom, but of their shared origin, for they are Greeks”. To give an
example of the upper echelon of these “Greek sages” in H.arrân, al-Mascûdî at this
point recounts what he read on the entrance door of their maǧmac; then, in
conclusion, he launches into a technical exposition of Greek philosophy. In such
a context, maǧ mac cannot mean anything other than a meeting place for
intellectuals.’29
‘al-Mascûdî distinguishes the cult sites or “temples” of the popular religion
perfectly from the maǧ mac where the “Greek philosophers” met. As regards the
former, he recognises at the time’, when he gives his summary of the two lost
books, which we mentioned above, that is, in 947, ‘that there remained only one,
“the temple called Maġlitiyâ; it is located in the city of H.arrân, near the Raqqa
gate.” The second centre, still thriving, of H.arrânian paganism was the institution
of the “Greek S.âbians”, that is, the Platonists.’30 Here, therefore, ‘maǧ mac’ means
their school, a sort of academy. The building whose door knocker was decorated
with the Syriac inscription which al-Mascûdî took care to have explained was the
Platonic school of H.arrân. ‘Those who accompanied al-Mascûdî on the course of
his visit were members of this Academy, and Mâlik b. cUqbûn, cited as the
interpreter of the inscription, was in all probability the head of this Academy.
al-Mascûdî equally claims to have consulted him on technical and historical
questions of Greek philosophy. During this consultation, he learned from Mâlik
b. cUqbûn and the other members of the group, that is, from the other Greek
philosophers of the school, that they categorically rejected the sacrificial and
divinatory practices of the H.arrânians as well as the “secret and mysterious” rites.
al-Mascûdî therefore distinguishes perfectly between the ordinary pagans of
H.arrân and the H.arrânian philosophers. If he made the effort to give an account
of his interview with Mâlik b. cUqbûn, it is because the latter had rank and
authority in the school.31
‘The school of H.arrân, in line with Plato’s ancient Academy (or rather with
reports of it), bore a maxim, not on the pediment of its portal, but on its door
knocker, which invited anyone who crossed the threshhold to the philosophic
life (bios philosophikos). At the time of al-Mascûdi’s visit in 943, H.arrânian
paganism possessed two centres of activity: the temple called Maglitiyâ and the
Platonic school. The critical thoughts he heard from those who attended the

29
ibid., p. 18.
30
ibid.
31
ibid., pp. 18–19.
 The life and work of Simplicius in Greek and Arabic sources 305

school show that the two institutions of H.arrânian paganism answered different
aims and for this reason were rivals.’32
Tardieu has therefore demonstrated that the S.âbians of H.arrân were not
Gnostics and that, contrary to Chwolsohn’s now commonly held opinion –
according to which all intellectual activity ceased after Tâbit b. Qurra left H.arrân
in order to live in Baghdad and to found the famous school of Baghdad there –
the Platonic school of H.arrân still remained a vital organism in the tenth century.
His departure for Baghdad was not the result of a ‘schism’, according to Tardieu.
Rather, it would appear that Tâbit b. Qurra was attracted by the Caliph’s subsidies
and had chosen Baghdad ‘for social and political reasons, as the Abbasid capital
at that time offered greater possibilities for influence than H.arrân, the capital of
Diyâr Mud.ar. Yet his departure for Baghdad meant neither the closure nor the
end of the school in H.arrân, which was to endure, since it still had members and
a head when al-Mascûdî visited it in 943, more than seventy years after Tâbit b.
Qurra’s move to Baghdad. Mâlik b. cUqbûn ensured the continuation in H.arrân
of a prestigious and centuries-old institution.’33
Tardieu also demonstrates convincingly that the argument of Max Meyerhof ’s
famous article, ‘Von Alexandrien nach Bagdad. Ein Beitrag zur Geschichte des
philosophischen und medizinischen Unterrichts bei den Arabern’, written in
1930 and authoritative since then, does not survive critical examination.34
According to Tardieu, Meyerhof was mistaken in accepting to the letter the
testimony of the Arabic sources concerning a relocation of the Neoplatonic
school from Athens to Alexandria, from there to Antioch, from Antioch to
H.arrân, and finally from H.arrân to Baghdad, for which the chronology is in
every way erroneous. The Arabic accounts must be understood not as a
‘chronology of institutional relocations, but as a confirmation of the intellectual
vitality of Antioch (the city and its outskirts) as much as H.arrân (the city and its
outskirts)’ around 700 ad (the end of the first century of the Hegira), ‘each of the
two centres laying claim at that time to the title and the honour of being the
repository of the Alexandrian heritage’.35
‘Under the Ummayyads . . . the two intellectual centres which were the
melting-pot for the transmission of Greek science to the Arab world were Syrian:
Greco-Syriac in Christian Antioch and Greco-Arabic-Syriac in pagan H.arrân,
which was to become the capital of the Empire with the accession of Marwan b.

32
ibid., p. 19.
33
ibid.
34
ibid., pp. 19–22.
35
ibid., p. 21.
306 Aristotle Transformed

Muh.ammad to the Caliphate in 744. The Christians who came from the
Alexandrian school settled quite naturally in Antioch and in the peripheral
monasteries, and from there they launched the intellectual conquest of H.arrân.’36
With the help of another Arabic source, Tardieu shows that intellectual traffic
had always existed between Antioch and H.arrân, but that it was due to the
Christians alone. The testimony of al-Mascûdî on the relocation of Greek science
from Antioch to H.arrân must be taken, according to Tardieu, ‘with great caution,
since it comes from a Christian source anxious to lay claim to Alexandrian
authenticity of its intellectual heritage for its own credit versus the rival school
of the “Greek S.âbians”, that is, of the true Platonists. Thus, there had not, properly
speaking, been a relocation of the old institution of Ammonius, where Simplicius
had taken courses at the beginning of the sixth century, either to Antioch or to
H.arrân . . ., but a double claim to a single heritage, mediated by Antioch in the
case of the Christian H.arrânians, direct and more ancient for those whom the
Arabs would call “Greek S.âbians”.’37
Tardieu has led us through the existence of a Neoplatonic school from the
end of the seventh to the tenth century ad. ‘Now,’ he concludes,38 ‘as the school
of H.arrân did not simply fall from the sky in the first century of the Hegira (thus
between 622 and 722), it is not implausible to suppose that it is there, in Byzantine
territory (but in proximity to the Iranian border), that the exiles from Athens,
the last battalion of Greek resistance to Christianisation, found refuge and
settled, among Platonists who had attended Ammonius’ school. This would
explain the insistence with which the Christian authors, in both Syriac and
Arabic, ridicule the school of the “Hellenes” in H.arrân, a sort of rival to the
neighbouring school of the “Persians” established in Nisibis, the one pagan,
secular, and philosophic, the other Nestorian, ecclesiastical, and theological, but
both under the protection of the King of kings.’ As regards this protection,
Tardieu cites a text of Procopius (Bella 2. 13.7) which specifies that Chosroes in
549 had exempted the inhabitants of H.arrân from paying tribute because they
had preserved the old religion.39 This testimony shows clearly that the Persian
king favoured any ideology opposed to the ideology of the Christian church
adopted by the Byzantine state; he therefore favoured paganism tout court, the
philosophical paganism of the Neoplatonists and the heterodox Christianity of
the Nestorians.

36
ibid., pp. 24–5.
37
ibid., p. 26.
38
ibid., p. 23.
39
ibid.
 The life and work of Simplicius in Greek and Arabic sources 307

‘Olympiodorus compares the Alcibiades to a propylaeum whose adytum is

the Parmenides. It is the same in H.arrân. The Alcibiades is at the entrance, the
Parmenides in the centre. In fact, according to the testimony of al-Kindî, whom
Ibn al-Nadîm cites from al-Sarah.sî, the core of the metaphysical thought of the
“H.arrânian S.âbians” was summarised in the claim that “in the world there is a
cause which has never ceased: a monad, not a plurality, which does not take on
the attributes of anything caused whatsoever.” ’40 D. Chwolsohn, followed by
W. Scott, thought this doctrine to belong to the Hermetica. But Tardieu has had
no trouble in showing that its origin is purely Neoplatonic. Yet al-Kindî’s account
not only confirms that the philosophy of the ‘S.âbians’ of H.arrân belonged to
Neoplatonism, and what is more, to a Neoplatonism of the most refined sort; it
equally reveals, Tardieu tells us,‘the other side of this philosophy. These H.arrânian
intellectuals had their public festivals, their prayers, their fasts and their dietary
customs. Such observances – difficult in the Athens of Proclus and Damascius
and in the Alexandria of Ammonius and Olympiodorus, as well as in the Abbasid
Baghdad where Tâbit b. Qurra went to live – could be performed very freely in
H.arrân, at the heart of a social group where paganism was still alive, in proximity
to a temple (that of the Moon goddess), whose activity had never ceased since
the Hurrian era. Living communally in their school, these Platonists could
without hindrance devote themselves to the theurgic practices which their
philosophy required.’41 I would only add, for my part, that such a life seems to
agree thoroughly with Agathias’ remark, namely, that the Greek philosophers,
after their departure from the court of Chosroes, ‘were able to spend the rest of
their lives in the most agreeable and pleasant manner’.42 ‘The error of perspective
on the part of the Arabic authors – and, consequently, on the part of Western
orientalism’, Tardieu continues43 – ‘is to have placed together in the same category
of “S.âbian” the observances of these philosophers and that of their compatriots
who attended the temples, the doctrine of the one and the religion of the others.
The only two authors who did not make this amalgamation were a philosopher
and a tourist: al-Kindî and al-Mascûdî, both Mesopotamians. The philosopher’s
information is drawn from a good source, since his informant and student, al-
Sarah.sî, was a correspondent of Tabît b. Qurra. The tourist’s information has
incomparable value, since it is the direct echo of what was still the Platonic
school of H.arrân during the tenth century.’

40
ibid., p. 27.
41
ibid., p. 28.
42
cf. p. 280 above.
43
M. Tardieu, ‘S.âbiens coraniques’, pp. 28–9.
308 Aristotle Transformed

According to Tardieu, it is also in all probability in H.arrân that Simplicius

wrote his commentary on Epictetus’ Encheiridion.44 Recall that it was once
believed, thanks to K. Praechter’s theories, that this commentary had been
written in Alexandria before Simplicius’ stay in Persia and that it contained
doctrinal elements characteristic of the school of Alexandria and opposed to the
school of Athens. I have refuted Praechter’s views in my book Le problème du
néoplatonisme Alexandrin, where among other things I arrived at the conclusion
that the commentary on Epictetus’ Encheiridion offered no precise evidence that
would allow for even a relative dating. In the same book,45 I also refuted in detail
Alan Cameron’s hypothesis, namely, that the same commentary had been written
in Athens in the same period as the other great commentaries of Simplicius, after
the stay in Persia. My criticism of Cameron’s two articles must have gone
completely unnoticed, for his theses have been repeated since as if they were
established facts. It is with regard to the date and location of composition of
Simplicius’ commentary on Epictetus’ Encheiridion that Tardieu enters as a
specialist of Manicheanism. In fact, in my analysis of the long discussion in
Simplicius’ commentary devoted to the refutation of Manicheanism, I had drawn
attention to the fact that, according to the text of my new critical edition of this
commentary, Simplicius personally spoke with a Manichean sage.46 I therefore
asked, without being able to answer this question, where Simplicius could have
had the opportunity to speak with a Manichean. According to Tardieu, at this
period, that is, after 533, ‘there were only two cities in the Byzantine empire
where there were certainly Manicheans: Constantinople, where they hid, and
H.arrân, where they had settled since the end of the third century and where,
thanks to the local socio-political conditions, they survived for several centuries.
The zeal shown against them by the Bishop of H.arrân, Theodore Abû Qurra,
around 764–5 according to the Chronicle of Pseudo-Dionysius of Tell-Mahrê,
confirms that they were still significant in the religious landscape of H.arrân in
the eighth century.’47 Tardieu also states that if Simplicius wrote this detailed
refutation of Manicheanism, ‘it is because there was urgency and external
pressure. One fact about anti-Gnostic and anti-Manichean controversies in the

44
ibid., p. 24, n. 105.
45
pp. 33–40.
46
I. Hadot, ‘Die Widerlegung des Manichäismus im Epiktetkommentar des Simplikios’, Archiv der
Geschichte der Philosophie 50, 1969, p. 46. The text of my new edition is actually the following: ‘but
as one of the wise men among them revealed to me (all’ hôs emoi tis tôn par’ autois sophôn exephêne)’
instead of 71.48 Dübner: ‘but as a certain wise man among them said they were (all’ hopoias tis tôn
par’ autois sophôn exephênen)’.
47
M. Tardieu, ‘S.âbiens coraniques’, p. 24, n. 105.
 The life and work of Simplicius in Greek and Arabic sources 309

schools of Platonic philosophy deserves to be repeated unceasingly: Plotinus

refutes the Gnostics because the Gnostics frequented his school. Alexander, the
head of the Platonic school of Lycopolis, refutes the Manicheans because the
Manicheans attended his courses. Why, then, did Simplicius set his heart on
refuting the doctrine of the Manicheans himself, if they were not present where
he worked on the edition of his commentary and if they did not disturb the
serenity of his rationalism through their intrusiveness and inopportune
questions? The only place where the direct contact would have been possible and
lasting, from a perspective which is at once historical, sociological, linguistic and
scholarly, is H.arrân.’48
In favour of the thesis that Simplicius did not write his commentary in either
Athens or Alexandria, I would further add an argument of textual criticism.49 It
has been remarked on by several scholars that the text of the fragments from
Parmenides’ poem which Proclus, Ammonius, Asclepius, Olympiodorus and
Simplicius provide cannot come from the same source.50 The manuscript of the
poem on which Simplicius relies coincides neither with the copy used by Proclus
which must have been in the library of the Neoplatonist school at Athens, nor
with that of Ammonius, Asclepius and Olympiodorus at Alexandria. According
to J. Whittaker and K. Deichgräber,51 Simplicius had at his disposal a manuscript
which had been badly copied from an exemplar which contained scholia.
Concerning Proclus’ text, C. Guérard recently expressed the opinion that Proclus
substituted several Neoplatonic retouchings for the text of the poem with the
intention of making the contents of the poem coincide with those of Plato’s
dialogue the Parmenides.52
In this context, evidence contained in the short entry the Arabic writer Ibn
al-Qiftî devotes to Simplicius might also be interesting. As H. Gätje explains in
his article, ‘Simplikios in der arabischen Überlieferung’,53 Simplicius appears here
in an alphabetical list of scholars as a celebrated mathematician, who had
composed, among other widely read writings, a commentary on Euclid and who
had assembled around himself students and successors who took their name
from their professor. The statement that Simplicius, like all the Neoplatonists,

48
ibid.
49
Yet see above all Tardieu’s article, ‘Les calendriers en usage à H
. arrân d’après les sources arabes et le
commentaire de Simplicius à la Physique d’Aristote.’
50
cf., for example, H. Diels, Parmenides’ Lehrgedicht, Berlin 1897, p. 26; J. Whittaker, ‘God, time, being’,
Symbolae Osloenses, Suppl. 23, 1971, pp. 19ff.
51
K. Deichgräber, ‘Xenophanes’ Peri Phuseôs’, Rheinisches Museum 87, 1938, p. 3.
52
C. Guérard, ‘Parménide d’Elée selon les néoplatoniciens’, in Études sur Parménide, ed. P. Aubenque,
vol. 2, Paris 1987, pp. 294–313.
53
H. Gätje, ‘Simplikios in der arabischen Überlieferung’, Der Islam 59, 1982, p. 16.
310 Aristotle Transformed

was a mathematician, has recently been entirely confirmed, notably by A.I.

Sabra’s article, ‘Simplicius’ proof of Euclid’s parallel postulate’.54 We will return to
this later in the second part of the chapter. The evidence that Simplicius taught
must therefore be taken more seriously as well. Yet where could he have done
this, except in a city like H.arrân?
One might object, however, that Simplicius could not have written his
commentaries without the aid of a very ample library, and that such a library
could not be found except in Athens or Alexandria. I believe this is a prejudice
which will not survive historical criticism. We can leave Alexandria aside, since
Simplicius could not have gone there after his stay in Persia without having
encountered Philoponus. He claims never to have known Philoponus personally.55
As regards the famous libraries of Athens, the wealth accumulated during the
Hellenistic era had been transported by Sulla to Rome following his military
conquest of the city in 86 bc . It is very likely that under the Roman emperors
and above all in the period of the Antonines libraries of value were reconstituted
in Athens, but it is very nearly certain that these were destroyed in the year 267
ad during the sack of the city by the Heruls. This is at any rate the opinion of
H.A. Thompson, who in his article ‘Athenian twilight: ad 267–600’ gives a
summary of the results of excavations dating from before 1959.56 According to
this article, the city of Athens never recovered from its almost total destruction
by the Heruls. The intrusion of the Goths in Attica in 396, even if it spared the
city itself, which at that time was very reduced in extent, nevertheless caused
economic suffering through the devastation of its hinterland. The Neoplatonist
Synesius, when he visited the city around 400, compared it to the skin of a
dismembered victim, the only remaining evidence of what was a living animal
not long before.57 Even though some activity of reconstruction is manifest from
the beginning of the fifth century, this remained modest, and the economic
importance of Athens never redeveloped in any manner comparable to that of a
number of Greek or Greco-Syriac cities in Asia Minor. The least that one can say
in taking account of the collection of facts and evidence that Thompson presents
is that the circumstances were hardly favourable to the reconstitution of libraries
comparable to the ones which had existed before the conquest of Sulla and the

54
Journal of the Warburg and Courtauld Institutes 32, 1969, 1–24. In this context, cf. P. Hoffmann, ‘Sur
quelques aspects de la polémique de Simplicius contre Jean Philopon . . .’, in I. Hadot (ed.), Simplicius
– sa vie, son oeuvre, sa survie, Berlin and New York 1987, p. 192 and n. 42, who speaks about the
public to which Simplicius’ commentary on the de Caelo is addressed.
55
Simplicius, in Cael. 26,17–24.
56
Journal of Roman Studies 49, 1959, pp. 61–72.
57
Synesius, Epist., 136.
 The life and work of Simplicius in Greek and Arabic sources 311

sack of the Heruls. On the contrary, everything suggests that a certain number of
cities on the Asiatic continent had come to surpass Athens as regards the wealth
of their libraries by far.
In considering the group of arguments I have just presented and in taking
account of those I have discussed elsewhere, I am now more than ever persuaded
that it is completely necessary to reject the hypothesis that Simplicius returned
to Athens or Alexandria after his stay in Persia. There is a strong presumption
and, after the publication of Tardieu’s article on calendars,58 even a near certitude
of the correctness of the thesis that, in the company of his teacher Damascius
and his fellow students, Simplicius turned towards the city of H.arrân, where he
was welcomed along with the other philosophers into the heart of a pre-existing
Neoplatonic school where all the philosophers continued to live, work and teach
together.

Let us proceed to Simplicius’ work. As H. Gätje stated in the article cited above,
the Arabic literature preserves characteristics of Simplicius’ scholarly personality
which would have remained unknown to us, had we only taken into consideration
the Greek works which the accidents of manuscript transmission have preserved
for us. Let us begin with the latter. I believe that it is unanimously agreed that all
of Simplicius’ great commentaries on Aristotle which we still read today were
written after the stay in Persia, thus after 532. Simplicius’ cross-references allow
us to establish a relative chronological order for three commentaries whose
authenticity is not contested, namely, the following: the commentary on the de
Caelo, the commentary on the Physics, the commentary on the Categories. In the
commentary on the de Caelo, Simplicius twice mentions Philoponus’ treatise de
Aeternitate Mundi contra Proclum, which was written in 529. On the other hand,
from a report of an observation made close to the river Aboras (Chaboras) in
Mesopotamia, a report which is also to be found in the commentary on the de
Caelo,59 it was concluded that this observation was made during his journey to
Persia and that the commentary on the de Caelo must therefore have been
written either in Persia at the court of Chosroes or somewhat after the stay in
Persia. Today we know that the observation in question could more probably
have been made later, after Simplicius had settled at H.arrân, as this river, a
tributary of the Euphrates, passes to the east of H.arrân at a distance of about

58
M. Tardieu, ‘Les calendriers en usage à H.arrân d’après les sources arabes et le commentaire de
Simplicius à la Physique d’Aristote’.
59
Simplicius, in Cael. 525,13.
312 Aristotle Transformed

40 kilometres.60 As regards the commentary on the Physics, Simplicius speaks

there of his master as someone no longer living. This commentary must,
therefore, have been written at least after 532, the date at which Damascius was
still alive according to Agathias, though probably only after 538. The latter date
is based on an epigram for a deceased slave, contained in the Palatine Anthology
and attributed to Damascius the philosopher.61 In this regard, I must mention a
correction Tardieu62 has made of my book Le problème du néoplatonisme
Alexandrin. There I wrote that the original source of this epigram, that is to say
the funeral stele, had been found in Emesa (H.ims.) in Syria. But in fact, it is
merely preserved in the museum of that city, while its provenance is unknown.
By following P. Waltz, I was once more mistaken concerning the date of the stele.
It is not dated 537, but 538. If we assume the philosopher Damascius of this
epigram to be identical with the Neoplatonist Damascius, the teacher of
Simplicius, the year 538 is obtained as a terminus post quem for the composition
of the commentaries on the Physics and on the Categories.
As regards the fourth Aristotelian commentary preserved under Simplicius’
name, the commentary on the de Anima, its authenticity was already put in
question at the beginning of the seventeenth century by Franciscus Piccolomini63
and more recently by F. Bossier and C. Steel in an article entitled, ‘Priscianus
Lydus en de “In de anima” van Pseudo(?)-Simplicius’.64 Piccolomini already
considered Priscianus Lydus to be the author of this commentary, the same
Priscianus who, according to Agathias, was present with Damascius and
Simplicius at the court of Chosroes and whose Solutiones eorum de quibus
dubitavit Chosroes Persarum rex we possess in a Latin version, a work that was
probably also written in H.arrân. But while Piccolomini took it for granted that
there was no doctrinal difference between Priscian and Simplicius, Bossier and
Steel base part of their argument on the alleged existence of such differences,
differences which H.J. Blumenthal and I deny. In this regard, I continue to
maintain all I said in my review of their article in an appendix to my book Le
problème du néoplatonisme Alexandrin65 and in my article, ‘La doctrine de
Simplicius sur l’âme raisonnable humaine dans le commentaire sur le Manuel

60
cf. M. Tardieu, Coutumes nautiques mésopotamiennes chez Simplicius, in preparation.
61
Anthol. Palat., VII ,553; vol. 5, 90 (Waltz).
62
M. Tardieu, ‘S.âbiens coraniques’, p. 22ff and n. 103.
63
cf. Francisci Piccolominei Senensis commentarii in libros Aristotelis De caelo, ortu et interitu; adjuncta
lucidissima expositione, in tres libros eiusdem de anima, nunc recens in lucem prodeunt, Moguntiae
1608, p. 1001f.
64
In Tijdschrift voor Filosofie 34, 1972, pp. 761–822.
65
pp. 193–202.
 The life and work of Simplicius in Greek and Arabic sources 313

d’Epictète’.66 At the time when I wrote this review, I was nevertheless seduced by
the arguments of Bossier and Steel concerning the textual parallels between the
commentary on the de Anima and Priscian’s Metaphrasis in Theophrastum, so
that I believed it probable that the commentary on the de Anima was Priscian’s.
I must say that today these textual parallels impress me much less, since I have
encountered a large number of such parallels while comparing Simplicius’
commentary on the Categories with those of Ammonius, Olympiodorus et al.,
parallels which are essentially due to a scholarly tradition that had become firmly
settled in its manner of commenting on the Categories and due also to dependence
on the same text for comment. Similarly, I think today that the absence of long
discussions of the doctrines of previous philosophers, which had so astonished
a number of scholars including myself, can be easily explained. Aristotle speaks
in a more or less detailed manner of the opinions of the Presocratics and Plato’s
school at the beginning of the Physics, in several books of the Metaphysics, and
also in the treatise de Anima. His Neoplatonic commentators thus had the
opportunity to discuss this subject in three different commentaries. Yet for a
Neoplatonist like Simplicius, who had already commented on the Physics and
the Metaphysics – since I am convinced that he actually did write a commentary
on the Metaphysics which is lost today67 – it would have been overly fastidious to
repeat himself endlessly. I thus think that the reason Simplicius gives for the
absence of discussions on the Presocratics and the Platonists in his commentary
on the de Anima is completely plausible. It is in effect: ‘I have analysed the thought
of these men in a sufficiently clear manner in the commentaries I have written
on the books of the Metaphysics. For the present, we shall speak of them only to
the degree that is necessary in order to explain the text on which we are
commenting.’ For all for these reasons, therefore, I have today many more
reservations than before concerning the hypothesis of this commentary’s
inauthenticity, although I do not regard discussion of this issue as closed.
In the commentary on the de Anima, Simplicius refers twice to his commentary
on the Metaphysics, three times to his commentary on the Physics, and once to
his epitome on Theophrastus’ Physics, works which are all lost. The commentary
on the de Anima, if it is authentic, is therefore situated chronologically after the
commentary on the Physics, but its chronological relation to the commentary on
the Categories cannot be determined. Of the commentary on the Metaphysics

66
In H.J. Blumenthal and A.C. Lloyd (eds), Soul and the Structure of Being in Late Neoplatonism,
Liverpool 1982, pp. 46–70.
67
cf. my article ‘Recherches sur les fragments du commentaire de Simplicius sur la Métaphysique
d’Aristote’ in Simplicius – sa vie, son oeuvre, sa survie, pp. 224–45.
314 Aristotle Transformed

and the epitome on Theophrastus’ Physics, we only know that they were written
before the commentary on the de Anima.
The Arabic tradition also attributes a commentary on the de Anima to
Simplicius. Here is the English translation of an entry in the Fihrist concerning
the treatise de Anima, in the translation given by F.E. Peters in his book Aristoteles
Arabus:68 ‘Liber animae: It has three books, Hunayn translated it into Syriac in its
entirety; Ishâq translated all but a small part of it, then Ishâq made another
version, emending it in the process. Themistius commented on this work in its
entirety: the first book in two books, the second in two books, and the third in
three books. And there is a (add. Al-Qifti: good) Syriac commentary by
Olympiodorus and I read this in the handwriting of Yah.yâ ibn cAdi. And indeed
there is extant in Syriac a good commentary which is attributed to Simplicius,
and he made this for Athâwâlîs. And indeed there is extant in Arabic an
Alexandrine epitome of this work, about a hundred leaves. And there is Ibn al-
Bitr.îq’s compendium of this work. Ishâq said: I translated this work into Arabic
from a corrupt copy, and after thirty years, I found an excellent copy, and I
collated this with the first version, and this is the commentary of Themistius.’
This text can be interpreted in different ways because it does not contain
punctuation.69 According to H. Gätje in his article, ‘Simplikios in der arabischen
Überlieferung’, this passage was understood by Steinschneider, Dodge, and Ğabr
in the sense that Simplicius had written a commentary on the de Anima that was
dedicated to a person named Athâwâlîs and that this commentary existed in a
Syriac and an Arabic version.70 For Peters, on the contrary – as his commentary
demonstrates71 – this text only signifies that there existed a Syriac translation of
Simplicius’ commentary. H. Gätje himself, when commenting again on the same
passage in his book Studien zur Überlieferung der aristotelischen Pyschologie im
Islam three years after Peters,72 followed the views of Steinschneider, Dodge, and
Ğabr on this point. Unfortunately, orientalists have so far not recovered any trace
of either a Syriac version or an Arabic version. Concerning Simplicius’ dedication
of his commentary on the de Anima to a certain Athâwâlîs, we are in the most
complete ignorance as regards the identity of this person, who was not necessarily
a Greek. For, if we can take the view that the commentary was written in H.arrân,

68
Leyden 1968, p. 40f.
69
cf. H. Gätje, Studien zur Überlieferung der aristotelischen Psychologie im Islam (= Annales
Universitatis Saraviensis, Reihe Philosophische Fakultät, vol. 11), Heidelberg 1971, p. 20.
70
In Der Islam 59, 1982, p. 11.
71
F.E. Peters, op. cit., p. 42f.
72
op. cit. (in n. 6), p. 26.
 The life and work of Simplicius in Greek and Arabic sources 315

it might be that Simplicius wished to dedicate his work to a Syrian,73 just as

Priscian had addressed his Quaestiones et Solutiones to the Persian king Chosroes
(in a Pahlavî or Syriac version?). Al-Qiftî, when rereading this text some three
hundred years later, transformed the sentence ‘he made this for Athâwâlîs’ into
‘Athâwâlîs likewise did this’. I have the impression that Arabists prefer the Fihrist’s
version, probably on account of its antiquity. Whichever version is followed, one
cannot argue from the existence of this dedication, attested by the Fihrist, to the
rejection of the authenticity of the commentary with which we are familiar
today under Simplicius’ name on the pretext that it does not contain the
dedication in question. Nevertheless, in order to refute this argument, it will not
suffice to say, as H. Gätje does74 (following Steinschneider), that all the Greek
manuscripts preserve this commentary without its opening, that is to say, its
preface, and that for this reason we cannot know whether the dedication was
originally there or not. This would not be entirely correct, for we have to
distinguish between preface and dedication. In fact, among the Greek manuscripts
which Hayduck mentions in the preface to his critical edition of this commentary,
there is, besides two manuscripts which contain the preface in the form of a
scholion, one manuscript, Matritensis Biblioth. Nat. 54, dated by Hayduck as
being from the fourteenth century, which possesses the preface and the opening
of the commentary up to page 31,19 of Hayduck’s edition. According to the
editor, this manuscript represents an independent manuscript tradition which is
often better than that of the other manuscripts, all of which appear to belong to
a single branch of the tradition. A new critical edition of this commentary, which
is absolutely necessary in my opinion, will certainly bring further clarifications
concerning the transmission of the text of the preface, though probably nothing
on the subject of the dedication. Dedications, which were often presented in the

73
Tardieu has had the kindness to send me the following note concerning the name ‘Athâwâlîs’: ‘This
patronym,’ he says, ‘is in my opinion a case of a double Greco-Semitic name, associating the name
of the male Arabo-Syrian god ‘Itâc, known from inscriptions CIS V/1, no. 3944 and 2595 (another
spelling: Yatic, in CIS V/1, no. 218, 2288, 2636, a Greek transcription attested by the inscription of
H. Waddington, IGLS no. 2209: Ethaos), with the Latin name Valens (in Greek: Oualês, Arabic and
Syriac: Wâlîs with normal iotacisation of the Greek eta). From this combination we get Etha-Oualês,
‘Etâwâlîs or ‘Itâwâlîs. It is therefore an apotheosising name meaning: “Valens is ‘Itâc”. The occurrences
of double Greek-Semitic names are frequent in the inscriptions of cities with a strong indigenous
population: for Palmyra, see CIS II /3, no. 4566: Šalman Marcellus; no. 4402: Apollodôros Elabêlos;
for Bostra, IGLS XIII /1, no. 9300: Severus Abdisar; no. 9274: Asados Priscus; no. 9415: Onemos
Maximos. The Arabic transcription attested by Ibn al-Nadîm is explained by having been translated
from the Greek, and not the Syriac (in which case the Arabic would have preserved the cayn of ‘Itâc).
Consequently, the commentary on the de Anima attributed to Simplicius, of which Ibn al-Nadîm
speaks, was not composed in Syriac. The expression “suryânî” (“in Syriac”), in the text of the Fihrist
concerns the language of the testimony on which the Arabic version of the in de Anima was made,
and not that of the original itself, which was in Greek.’
74
cf. H. Gätje, Studien zur Überlieferung, p. 26.
316 Aristotle Transformed

form of a detached letter, hardly interested the Byzantine scholars who were
responsible for the transmission of texts, above all if they were lengthy. It is for
this reason that while Simplicius in the sixth century still read75 the dedication
that Arrian, the compiler of Epictetus’ Encheiridion, had addressed to a certain
Messalinus when sending him the Encheiridion, this dedication has not been
preserved for us despite the enormous quantity of manuscripts containing the
Encheiridion that still exist today. Thus, if Simplicius’ dedication to Athâwâlîs, as
attested by Fihrist, is no longer extant today in any of the manuscripts containing
the commentary which we read under Simplicius’ name, this cannot be an
argument in favour of its inauthenticity.
Regarding Simplicius’ commentary on Epictetus’ Encheiridion, K. Praechter
used it in addition to Hierocles’ commentary on the Pythagorean Carmen
Aureum to support his thesis that there existed a fundamental difference between
the doctrines of the Neoplatonic school at Athens and of that at Alexandria. An
essential element of his thesis was the dating of this commentary: according to
Praechter, it must have been composed at a time when Simplicius still adhered to
the Alexandrian school while attending Ammonius’ course and therefore his
departure for Athens, where he would fall under the influence of that school, an
influence with which all his commentaries on Aristotle were marked. I wrote Le
problème du néoplatonisme Alexandrin: Hiéroclès et Simplicius76 as an introduction
to my critical edition of Simplicius’ commentary to Epictetus’ Encheiridion.
There I refuted Praechter’s hypotheses in detail and showed that there were not
such doctrinal differences either between the schools of Athens and Alexandria,
or between Simplicius’ commentary on the Encheiridion and his other
commentaries, and that this commentary reveals, as do all of his commentaries,
Damascius’ influence. In that book, I had not put forward any hypothesis
regarding the date of this commentary, as I was not able to take advantage of
Simplicius’ statement that he had personally spoken about the doctrines of the
Manicheans with an initiate of that sect. Today, I subscribe to Tardieu’s arguments
concerning the dating and the location of the writing of this commentary, which
I have summarised above: it must have been written in H.arrân after 532.
I am profoundly sorry that H. Dörrie, in his article ‘Simplikios’ in the Kleine
Pauly, did not indicate the sources on which he based the following claims:
‘Preserved in the manuscripts, but not yet edited (!) are a commentary on the art
[tekhnê] of Hermogenes and another on Iamblichus’ treatise On the Pythagorean

75
cf. Simplicius, in Ench. Epict., l,13ff (Dübner).
76
Paris 1978.
 The life and work of Simplicius in Greek and Arabic sources 317

Sect [Peri tês Puthagorou haireseôs].’ I have carried out some research on library
catalogues, but I have not found anything of the sort so far. Certainly my
investigations have not been exhaustive, but I none the less believe that H. Dörrie
was the victim of an over-hasty reading of Praechter’s article ‘Simplikios’ in
Pauly-Wissowa. There Praechter in fact states: ‘As for the other unedited works,
supposedly preserved in the manuscripts (Scholia on Hermogenes’ Tekhnê,
commentary on Iamblichus’ three books On the Pythagorean sect, treatise on the
syllogism), see Fabricius-Harles, Bibliotheca Graeca V.770 and IX .567.’77 Is it
possible that H. Dörrie did not see the word ‘supposedly’?
When the text of the Bibliotheca Graeca of Fabricius-Harles that Praechter
mentions is examined, it is apparent that Fabricius refers, regarding Simplicius’
commentary on Iamblichus’ treatise On the Pythagorean Sect, to an entry of
Petrus Ioannes Nunnesius, who in turn had relied on Gesner’s Bibliotheca; and
when Gesner’s text is appealed to, it becomes clear that he reproduces a statement
by Raphael of Volterra, according to whom such a commentary by Simplicius,
comprising three volumes, had existed in the Vatican library. Raffaello Maffei of
Volterra’s book Commentariorum urbanorum octo et triginta libri appeared for
the first time in 1506 in Rome.78 The modern printed catalogues of the Vatican
do not make mention of such a manuscript and, as far as I know, neither does the
manuscript catalogue. For my own part, I do not have a great deal of confidence
that the manuscript which Raphael mentions will ever be found.
Yet, as F. Bossier has very clearly demonstrated,79 a manuscript of this
commentary certainly still existed in the sixteenth century. For in a lending
record of the Bibliotheca Marciana, published by H. Omont,80 we find the
following remark: ‘The above named gentleman Master Sebastiano Erizzo, on
the instructions of the Riformatori, on the 7th day of November 1553, took out
a work, namely, the three books of Iamblichus of Chalcedon on the Pythagorean
Sect, with the commentary of Simplicius; and I took as security a little diamond
set in a gold ring. On the said day, he returned the above named book, which he
said was not what he wanted, and took a second copy, said to be the fourth book
of Iamblichus on the Sect, with a commentary of the same on the letters of
77
‘Simplikios’, Paulys Realencyclopädie, vol. III A.1 (second series), 1927, col. 205.
78
cf. F. Bossier, Filologisch-historische navorsingen over de middeleeuwse en humanistische Latijnse
vertalingen van de Commentaren van Simplicius, part 1 (thesis from the Catholic University in
Louvain, 1975, though unfortunately not yet published), p. 16.011, n. 2: there Bossier cites the
following remark from book 18, f.189 (verso) of the 1511 edition: ‘On his sect, Iamblichus wrote
three books, which are still extant. Simplicius wrote a work on those commentaries, which is to be
found in the Vatican library’.
79
F. Bossier, op. cit., 16.011–16.014.
80
H. Omont, Deux registres de prêts de manuscrits de la Bibliothèque de Saint-Marc à Venise (1545–
1559), Paris 1888, p. 40. [I thank Jill Kraye for help with the interpretation of this entry. – Ed.]
318 Aristotle Transformed

Porphyry, and also Hierocles’ introduction to Pythagoras’ Carmina Aurea, 7. δ

no. 526; and the security remained in my possession’. Bossier notes that the
second codex was identified by Sicherl as the extant Taurinensis gr. 146: it is a
manuscript which belonged to Cardinal Bessarion. Yet what was the first
manuscript which Erizzo had returned and which contained Simplicius’
commentary? Bossier takes the reasonable view that this manuscript must also
have belonged to Bessarion’s library bequeathed to Venice, although we have not
found any trace of it in any of the old and new catalogues of Bessarion’s donation
and of the Bibliotheca Marciana. Yet the two pieces of evidence (one from
Raphael of Volterra, the other from the lending record of the Bibliotheca
Marciana) concerning the existence of a commentary by Simplicius on
Iamblichus’ De secta Pythagorica, one in 1506 in Rome, the other in 1553 in
Venice, exclude the possibility that it was pure invention.81
Bossier’s very rich thesis also contains a discussion82 of a piece of evidence
found in C. Gesner’s enlarged edition of the Bibliotheca universalis,83 which
mentions a Commentariolus de syllogismis by Simplicius. It might be that the fact
that a Tractatus brevis de syllogismis by an anonymous author follows the great
commentary of Simplicius on the Categories in several manuscripts suggested
the notion that this short treatise came from the same author as the commentary
on the Categories. According to Bossier, thorough inquiry into this short treatise,
of which there exist Christian and non-Christian versions, might eventually
bring some clarification to the question of its provenance.
Bossier84 also argues against the claim which H.D. Saffrey and L.G. Westerink
make85 that Simplicius possessed a copy of Proclus’ commentary on the Timaeus,
a copy which contained glosses. Here is the scholion in question, which occurs
in the Paris. Coisl. 322 and which actually presents a distich as Simplicius’ gloss
on Proclus’ commentary on the Timaeus:
By Simplicius on the words, ‘But may the divine Iamblichus be well disposed
to us’:

81
F. Bossier thinks that an exhaustive investigation of the collection of scholia of Iamblichus’ treatise,
On the Pythagorean Sect, which are contained in the Laurent, gr. 86.3, might eventually provide an
answer to this enigma.
82
op. cit. 16.004–16.006.
83
cf. C. Gesner, J. Simler, J. Frisius, Bibliotheca instituta et collecta, primim a C. Gesnero. Deinde in
Epitomen redacta, et novorum librorum accessione locupletata, tertio recognita, et in duplum post
priores editiones aucta, per I. Simlerum. Iam vero postremo aliquot mille, cum priorum tum novorum
authorum opusculis ex instructissima Viennensi Austriae Imperatoria Bibliotheca amplificata per I.J.
Frisium . . . Zurich 1583, p. 759, col. 1.
84
op. cit. 16.006–16.008.
85
In Proclus, Théologie platonicienne, vol. 1, Paris 1968, pp. clii–cliii.
 The life and work of Simplicius in Greek and Arabic sources 319

Here is the work of Simplicius. O Iamblichus, dispenser of all goods,

Forgive me if I have vanquished you, since it is by your own words!

(Simplikiou eis to ‘all’ ho men theios Iamblikhos eumenês estô’:

Simplikiou ponos houtos; Iamblikhe dôter eaôn hilathi nikêtheis, all’hupo sôn epeôn.)

By citing numerous examples of Hellenistic epigrams, Bossier demonstrates

that, to begin with, the phrase ‘Here is the work of Simplicius’ does not mean that
Simplicius designates himself as the author of the distich, as Saffrey-Westerink
thought (p. cliii), but that the distich is an epigram in the etymological sense, that
is, an inscription which, in our case, occurred at the beginning of a book by
Simplicius and indicated its content in a concise and pregnant manner. ‘Here is
the work of Simplicius’ must therefore be understood as follows: ‘The work which
follows the distich is by Simplicius’, or ‘the commentary which follows is by
Simplicius’. Bossier then shows that the distich in question has a greater chance
of referring to Simplicius’ commentary on the Categories than to one of his other
works: it is probably not an accident that the same epigram appears three times
in a manuscript of Simplicius’ commentary on the Categories, specifically in the
cod. Ambros. 306 (E 99 sup.), folios Ir, II r, III r. Moreover, Simplicius, according
to his own words in the preface to his commentary on the Categories, is largely
relying in this commentary on Iamblichus’ commentary. Bossier thinks that a
certain scholiast glossing Proclus’ commentary on the Timaeus, on coming to
the words eumenês estô in the text of Proclus, recalled the similar phrase hilathi
nikêtheis in the distich introducing Simplicius’ commentary on the Categories,
which led him to reproduce this distich in the margin. Another scholiast, when
reading the distich in the margin, made the mistake of thinking that the distich
had been composed by Simplicius himself concerning the words eumenês estô.
Bossier also corrected86 A. Wartelle’s faulty entries in his Inventaire des
manuscrits grecs d’Aristote et de ses commentateurs,87 according to which several
Greek manuscripts contain a commentary by Simplicius on the Sophistici
Elenchi, by showing that there is nothing of the sort.

As regards Simplicius’ commentary on Hermogenes’ Tekhnê, Harles88 refers to a

manuscript in Vienna, cod. 137, that is, Vindob. Phil. gr. 15. The description of
86
op. cit. 16.010–16.011.
87
Paris 1963, pp. 97 and 196.
88
J.A. Fabricius – G.C. Harles, Bibliotheca graeca sive notitia scriptorum veterum graecorum
quorumcumque monumenta integra aut fragmenta edita exstant tum plerorumque e mss. ac
deperditis, editio quarta, Hamburg 1790–1809, vol. 5, p. 770.
320 Aristotle Transformed

this manuscript by Lambeck-Kollar for folios 49r-173v mentions a ‘Compilation

of the Introductions of Various Authors to Hermogenes’ Rhetorical Art, namely
from: Longinus, Iamblichus, Syrianus, and Simplicius, who were pagans; and
John the Sicilian, John the Geometer, Gregory the Metropolitan of Corinth, John
Doxopatres, and George Diairetes, who were Christians’.89 I have not seen this
manuscript and therefore do not know the corresponding Greek title. H. Hunger,
in his study devoted to John Chortasmenos (= Ignatius of Selymbria) which
contains among other things an almost complete edition of those of his works
contained in Vindob. Suppl. gr. 75 (Chortasmenos’ autograph), makes mention
of a ‘Collection of the Most Essential Commentators on Rhetoric, etc.’90 which is
to be found on folios 97r-112v and of which he unfortunately does not give the
complete title. It deserves mention that this very long title names not less than
ten authors, pagans and Christians, as sources for this collection of extracts from
introductions to rhetoric: from the Christians, the philosopher John the Sicilian,
John Didaskalos the geometer, Gregory the Metropolitan of Corinth, John
Doxopatres, George called the Diairetes; and from the Greeks, Simplicius,
Longinus, Syrianus and Iamblichus. In the text itself none of these names
appears again. H. Hunger thinks that this compilation was probably made by
Chortasmenos himself.91
Could the title of this compilation be the same as that of Vindob. Phil. gr. 15
and as that of an anonymous work mentioned by L. Allatius in his Diatriba de
Georgiis, a title which was reproduced by C. Walz in vol. 6 of his Rhetores Graeci92
as follows: ‘Collection of the Most Essential Commentaries on Rhetoric as
Explained by Christian Commentators and by Greeks, of whom the Christians
are John the Great Philosopher and Most Orthodox Sicilian, a different John
Didaskalos the Geometer, Gregory the Metropolitan of Corinth, a different
Metropolitan of Sardes Doxopatres named John, George called the Diairetes; the
Greeks are Simplicius, Longinus, Syrianus, and Iamblichus’?
The same title, with a few slight modifications, is found again in another
manuscript, Londin. Harleianus 5697 on folio 104r: ‘Introduction to Fire
Breathing Rhetoric from the Explanations of the Commentaries on this, by
Christian and Greek Commentators, of whom the Christians are John the Great
Philosopher and Most Orthodox Sicilian, a different John Didaskalos the

89
cf. P. Lambeck, Commentariorum de Augustissima Bibliotheca Caesarea Vindobonensi liber septimus,
editio altera studio et opera A.F. Kollarii, Vienna 1781, cols. 550–3.
90
cf. H. Hunger, Johannes Chortasmenos (c. 1370–c. 1436/37): Briefe, Gedichte und Kleine Schriften,
Einleitung, Regesten, Prosographie, Text (= Wiener Byzantinische Studien, vol. 7), Vienna 1969, p. 29.
91
H. Hunger, op. cit., pp. 29ff.
92
cf. C. Walz, Rhetores Graeci, vol. 6, Stuttgart, Tübingen, London and Paris 1834, p. vi.
 The life and work of Simplicius in Greek and Arabic sources 321

Geometer, Gregory the Metropolitan of Corinth, another Metropolitan of Sardes

the Doxopatres named John, George called the “Diairetes”; the Greeks are
Simplicius, Longinus, Syrianus, Iamblichus’.
This manuscript contains the following works: <John Chortasmenos’>
commentaries (with paraphrases and extracts) on the Organon of Aristotle and
on Hermogenes; Proclus’ Elements of Physics (Stoikheiôsis phusikê). The writing
which occupies the greatest part of our manuscript, including the Prolegomena
to Rhetoric, has been identified by J. Wiesner93 as that of Chortasmenos. The
manuscript belonged to Cardinal Bessarion, a student of Chortasmenos, and
was integrated much later into the collection of Greek manuscripts of the Palazzo
Farnese.94 E. Harley bought it on 18 January 1723/24. H. Hunger, in his study on
Chortasmenos cited above, speaks neither of this manuscript nor of Vindob.
Phil. gr. 15, so that the relationship between the three Prolegomena to Rhetoric
and, if possible, that of Allatius’ manuscript, has yet to be determined. A thorough
investigation will perhaps reveal that this collection of extracts is not the work of
Chortasmenos, but is older.
H. Hunger says the following of the collection from Vindob. Suppl. gr. 75.95 It
consists of two parts. The first, from folio 97r to 104v, essentially comprises
extracts taken from the rhetorical works of John Doxopatres, of Maximus
Planudes, and of Troilus. The second part (folios 104v-111v) contains discussions
on concepts like ‘truth-itself ’ (autoalêtheia), ‘paradoxes’ (paradoxa), and ‘received
opinions’ (endoxa), on the ‘dialectical premise’ (protasis dialektikê), ‘dialectical
and logical problems’ (problêmata dialektika kai logika), on the ‘first principles of
the sciences’ (prôtai arkhai tôn epistêmôn), and on the difference between
‘premise’ (protasis) and ‘problem’ (problêma). Hunger remarks that these
discussions are based in large part on Aristotelian definitions, taken above all
from the Topics, and he adds that he was unable to determine whether
Chortasmenos or his source used one of the already edited commentaries on
Aristotle.96 However that may be, the very word ‘truth-itself ’ (autoalêtheia)
reveals the Neoplatonic tenor of the second part, but nothing indicates that it
came from a commentary of Simplicius. As regards the first part, it may be
assumed that Chortasmenos knew at least the rhetorical work of Doxopatres
and of Maximus Planudes. Now, these two authors themselves used ancient and

93
J. Wiesner, in P. Moraux, Aristoteles Graecus: Die griechischen Manuskripte des Aristoteles I, Peripatoi
8, Berlin and New York 1976, pp. 432ff.
94
cf. L. Pernot, ‘Nouveaux manuscrits grecs Farnésiens’, in Mélanges de l’École Française de Rome 93,
1981, p. 706.
95
H. Hunger, op. cit., pp. 30ff.
96
ibid., p. 30.
322 Aristotle Transformed

above all Neoplatonic sources, sometimes naming them, sometimes not. Thus,
one of the texts used by Chortasmenos which H. Hunger had correctly identified
as belonging to Maximus Planudes, has parallels in a Prolegomenon attested as
belonging to Syrianus in Paris. gr. 1983, a manuscript used moreover by
Planudes.97 As for Doxopatres, he cites for example Porphyry’s Isagoge98 and one
of Porphyry’s two commentaries on the Categories,99 an unnamed work of
David,100 an unnamed work of Simplicius which I believe to be his commentary
on the Categories,101 and Plato’s Gorgias.102 He also cites Dionysius Thrax, the
rhetorical commentaries of four Neoplatonic philosophers, to wit, Marcellinus,
Sopatros, Eustathius, and Syrianus,103 and among Christians a certain George
and a geometer who are probably respectively the George the so-called Diairetes
and the John Didaskalos the geometer mentioned by Chortasmenos in his
Collection. It could very well be that Chortasmenos or his source knew a
commentary by Simplicius on Hermogenes’ Tekhnê, even if he knew it only
through the mediation of other Byzantine commentaries rather than by a direct
reading. This is not improbable, since as we shall see in regard to Simplicius’ lost
commentary on the Metaphysics,104 the Byzantines of the eleventh to thirteenth
centuries clearly possessed more information than we do on Simplicius’ works. If
one is very suspicious, one might clearly imagine that Chortasmenos had taken
Doxopatres’ mere mention of Simplicius’ name as testimony of the existence of a
rhetorical commentary by Simplicius, mention which would probably refer to
the commentary on the Categories.105 For my part, I do not believe this to be the
case, and I base my opinion on a series of three epigrams in honour of Simplicius,
published in E. Cougny’s Anthologiae Graecae appendices, which I reproduced in
the biographical chapter of my book Le problème du néoplatonisme Alexandrin.106
One of these three epigrams is addressed to Simplicius as the author of the
commentary on the Categories, the second makes allusion to his commentary on

97
On the use of Paris. gr. 1983, dated to tenth/eleventh century, and of Maximus Planudes’
Prolegomenon 13, cf. H. Rabe, Prolegomenon Sylloge, Leipzig 1935, pp. xlv and lx. The Prolegomenon
13 was attributed in Paris. gr. 1983 to Syrianus by a later hand: cf. R. Rabe, ibid., p. 183.10 (critical
apparatus). Cf. Prolegomenon 7 (Maximus Planudes), p. 64.11–12 and 15–19 (Rabe) and
Prolegomenon 13, p.190.6–7 and 18–25 (Rabe).
98
cf. C. Walz, Rhetores Graeci, vol. 2, 1832, 95,6ff.
99
ibid., 157.4ff; cf. Porphyry, in Cat. 60,17f; 153,30–154,2.
100
ibid., vol. 2, 154,3 (cf. David, Prol. Philos. 12,23).
101
ibid., vol. 2, 153.30–154.2; cf. Simplicius, in Cat. 55,24–9.
102
cf. C. Walz, Rhetores Graeci, vol. 2, 104,18ff.
103
Concerning the Neoplatonists Marcellinus, Sopatros, Eustathius and Syrianus, cf. most recently
G.A. Kennedy, Greek Rhetoric under Christian Emperors, Princeton 1983.
104
cf. I. Hadot, ‘Recherches sur les fragments du commentaire de Simplicius sur la Métaphysique
d’Aristote’, in Simplicius – sa vie, son oeuvre, sa survie, pp. 225–45.
105
cf. n. 101.
106
op. cit., p. 31.
 The life and work of Simplicius in Greek and Arabic sources 323

the de Caelo, and the third glorifies him in his dual capacity as orator and
philosopher. Here it is:107

To Simplicius:

Is Simplicius an orator or a philosopher?

Both, it seems to me, and by my word, the best.
(Eis Simplikion
Rhetor ho Simplikios, ê philosophos;
Amphô dokei moi, kai, ma tous logous, akros.)
(Utrum rhetor Simplicius, an philosophus?
Utrumque videtur mihi, et per litteras, summum.)

That a Neoplatonic philosopher should preoccupy himself with rhetoric

should not, moreover, be surprising. I described in my book Arts libéraux et
philosophie dans la pensée antique the slow evolution within the Platonic school,
which led, beginning with a hostile attitude towards rhetoric in Plato’s life, to the
inclusion of rhetoric in the plan of studies for the future Platonic philosopher
and to the conviction that the abilities of the orator were indispensable to a
philosopher if he wished to address others for persuasion.108 We know, moreover,
with certainty that Porphyry had written a commentary on Minucianus and
other rhetorical works,109 that Iamblichus had written a rhetorical work cited by
Syrianus,110 and similarly for Iamblichus’ students Sopatros and Eustathius.
Marcellinus and Syrianus, also Neoplatonists, had composed commentaries on
Hermogenes.111 Damascius, Simplicius’ teacher, perhaps did not write a
commentary, but he himself says that he had taught rhetoric for nine years.112
For all these reasons, I believe it probable that Simplicius wrote a commentary
on Hermogenes’ Tekhnê, of which one can perhaps find traces when systematically
exploiting the enormous mass of manuscript material that still exists in the form
of anonymous scholia concerning rhetoric.
To come finally to the Arabic tradition as regards Simplicius, H. Gätje’s article,
‘Simplikios in der arabischen Überlieferung’,113 attempts to provide an inventory.
The same Fihrist of Al-Nadîm, of which we have already spoken in connexion
107
E. Cougny, Anthol. Graecae appendices, Paris 1890, vol. 3, ch. 3, p. 321, no. 180.
108
Études Augustiniennes, Paris 1984, pp. 50, 76–7, 83, 91–4, 107, 228, 279.
109
cf. R. Beutler, ‘Porphyrios’, in Realencyclopädie der klassischen Altertumswissenschaft, vol. 22, 43.
Halbband, 1953, col. 300.
110
Syrianus in Hermogenem I, 9.10ff (Rabe), Leipzig 1892.
111
On all of these Neoplatonists, cf. St. Gloeckner, Quaestiones rhetoricae. Historiae artis rhetoricae
qualis fuerit aevo Imperatorio capita selecta, Breslauer Philol. Abhandlungen, vol. 8, no. 2, Breslau
1901, and most recently G.A. Kennedy, op. cit. in n. 103.
112
cf. Photius Biblioth., cod. 181, vol. 2, p. 192 (Henry).
113
In Der Islam 59, 1982, 6–31.
324 Aristotle Transformed

with the remark on Simplicius’ commentary on the de Anima,114 also attests the
existence of a commentary on the Categories115 as Al-Qiftî does much later, who
as a general rule produces the material found in Al-Nadîm with some additions,
omissions and alterations. Yet regarding Simplicius’ other commentaries on
Aristotle, the Arabic bibliographic sources are silent. Conversely, in these two
Arabic authors, Simplicius is named in the capacity of mathematician and
astronomer, and as having written a commentary on the first book of Euclid’s
Elements. Al-Qiftî adds in this regard, as already mentioned above,116 that
Simplicius had founded a school and that he had students who were named after
him. A.I. Sabra in his article, ‘Simplicius’ proof of Euclid’s parallel postulate’,117
has collected besides the extracts of this commentary cited by al-Nayrîzî118 in
Arabic in his own commentary on Euclid’s Elements, an extract contained in a
letter of cAlam al-Dîn Qays.ar ibn Abi ‘l-Qâsim to Nas.îr al-Dîn al-T.ûsî, and
beyond that a text contained in an Arabic manuscript, Bodleianus Thurston 3,
fol. 148r.119 al-Nayrîzî’s commentary was known in the West through the Latin
translation of Gerard of Cremona. Simplicius is cited there under the name of
‘Sambelichos’. The Greek tradition only allows us to conclude indirectly about
Simplicius’ capacities as a mathematician: his commentary on the Physics actually
contains discussions on the theory of the quadrature of half lunes by Hippocrates
of Chios, a mathematician who lived before Aristotle. Yet there is no need of
indirect testimony in order to be convinced of the identity of Simplicius the
commentator of Aristotle, Epictetus, and Hermogenes with Simplicius the
mathematician. To begin with, the Fihrist clearly connects the philosopher and
the mathematician, and further we know that every Neoplatonist philosopher
was also a mathematician in so far as he was a philosopher. Generally speaking,
as I have shown in Arts libéraux et philosophie dans la pensée antique,120 it was the
schools of philosophy, Platonic or Stoic, which in the Hellenistic or Imperial
period gave instruction in mathematics and in astronomy to youth desiring to
gain knowledge in these matters within the framework of the acquisition of
general culture. The instruction of surveyors and engineers in geometry was
only directed by professionals to future professionals.

114
See above, p. 292.
115
cf. also F.E. Peters, Arisoteles Arabus, Leyden 1968, p. 7.
116
See above, p. 288.
117
The article cited in n. 54.
118
G. Freudenthal has been kind enough to tell me that E. Wart of the Library of the Hebrew University,
Jerusalem, has discovered in an MS from the Yemen some pages containing in Arabic translation
new fragments of the commentary of Simplicius on the first book of Euclid’s Elements.
119
A mathematician living in the ninth century ad.
120
Études Augustiniennes, Paris 1984, 252–61.
 The life and work of Simplicius in Greek and Arabic sources 325

At this point, let us add a further interesting detail. In one of the textual
fragments of Simplicius’ commentary on the first book of Euclid’s Elements
reported by al-Nayrîzî, Simplicius speaks of his ‘s.âh.ib’ named ‘Aghânîs’ or ‘Aġânîs’
and cites a mathematical demonstration from him.121 What might the
corresponding Greek term be? A.I. Sabra translates this by ‘our associate’, which
might possibly make one think of a professional associate in the school which,
according to al-Nadîm, Simplicius directed.122 It might also be a matter of an
Arabic translation of the Greek term hetairos which, in the manner that
Neoplatonists use it, designates a fellow student who has been admitted to the
inner circle of true adherents of Neoplatonic philosophy. As regards his name,
until recently only a Greek or Latin name was sought beneath this Arabic form,
as for example ‘Geminus’ (the mathematician) or ‘Agapius’.123 Yet it might be
more simple to take the name ‘Aġânîs’ as such and recognise in it, with Tardieu, a
patronym attested in Egypt.124 In this case, the ‘Aġânîs’ cited in the fragment of
al-Nayrîzî is not a corrupted form of some Greek or Latin name at all, but the
real name of Simplicius’ ‘s.âh.ib’, who is probably an Egyptian.
Finally, a doctor named Simplicius also appears in the Arabic biographical
and bibliographical tradition. Must this or these doctors named Simplicius (it
seems that al-Nadîm probably spoke of two different doctors named Simplicius)
be identified with our Simplicius? It is a fact that apart from al-Nadîm, the Arabic
doctor ar-Razi (who died in 925) attributed to an author the form of whose
name could be translated by Simplicius a commentary on the Hippocratic
treatise, On Bone Fractures.125 The dating of this Simplicius causes problems.
According to al-Nadîm, he lived after Aristotle and Euclid, but before Galen. Yet
the remark ‘before Galen’ does not necessarily have to be taken strictly, as Gätje
claims. In any case, the latter is certainly wrong in thinking, on the grounds that
there were two different doctors named Simplicius, that the first Simplicius could
have been a contemporary of Hippocrates.126 This hypothesis is impossible

121
cf. A.I. Sabra, op. cit., p. 6 in English translation.
122
cf. above, pp. 288 and 301.
123
cf. the summary of interpretations made until recently in A.I. Sabra’s article, ‘Thâbit Ibn Qurra on
Euclid’s Parallel Postulate’, Journal of the Warburg and Courtauld Institutes 31, 1968, p. 13, n. 6.
124
I am indebted to Michel Tardieu’s kindness for the following points: the patronym ‘Aġânîs’ is formed
from a masculine common noun which means, in Semitic and in Egyptian, ‘cup’, ‘jug’, ‘vase’. This
patronym is attested as such only in Egypt: (a) in the form Agânis, Agenis in the Greek documents
(cf. F. Preisigke, Namenbuch, Heidelberg 1922, p. 6); (b) in the form ‘Agan in the Aramaic documents
of Elephantine (cf. E. Sachau, Aramäische Papyri und Oustraka aus einer Militär-Kolonie zu
Elephantine, Leipzig 1911, Ostracon 75/2: French translation by P. Grelot, Documents araméens
d’Egypte, Paris 1972, no. 12b/15, p. 105; cf. H. Ranke, Die ägyptischen Personennamen, 2 vols,
Glückstadt-Hamburg 1935–52, vol. 2, p. 186).
125
cf. H. Gätje, ‘Simplikios in der arabischen Überlieferung’, Der Islam 59, 1982, pp. 12f and 18f.
126
H. Gätje, op. cit., p. 13.
326 Aristotle Transformed

because of the name ‘Simplicius’ alone, which, for a Greek, can only be late.
Anyway, I believe that, given the present state of research, it is more prudent to
pronounce a non liquet concerning the hypothesis that our Neoplatonist
philosopher Simplicius could have written a commentary on a Hippocratic
treatise.
Here, therefore, are the conclusions to which one might be led as regards
Simplicius’ works. We have extant: the commentaries on Epictetus’ Encheiridion,
on Aristotle’s de Caelo, Physics, Categories, and probably on his de Anima. Lost,
though attested in a more or less certain fashion: a commentary on the first book
of Euclid’s Elements, a commentary on Aristotle’s Metaphysics,127 a commentary
on Iamblichus’ work devoted to the Pythagorean sect, an epitome of Theophrastus’
Physics (if the commentary on the de Anima, where one finds a reference to this
work, is authentic), and perhaps a commentary on Hermogenes’ Tekhnê.

127
cf. I. Hadot, op. cit. in n. 67, pp. 225–45.
 13

Neoplatonic elements in the de Anima

commentaries 1

Henry J. Blumenthal

Most scholars who refer to the Greek commentators for help in the understanding
of difficult Aristotelian texts seem to expect straightforward scholarly treatment
of their problems.2 Not infrequently they are disappointed and complain
about the irrelevance of the commentary they read, or inveigh against the
incompetence of the commentators.3 Only Alexander is generally exempt from
such censure, and that in itself is significant. For he is the only major commentator
whose work survives in any considerable quantity who wrote before
Neoplatonism. Shortly after Alexander the kind of thought that is conveniently
described by this label came to dominate Greek philosophy, and nearly all pagan
philosophy and philosophical scholarship was pursued under its influence,
if not by its active adherents. It is the purpose of this paper to argue that
these facts are not trivial items of background interest, but are fundamental to a
proper assessment of the later commentators’ opinions on points of Aristotelian
scholarship. It is necessary to take account of the ideas and purpose of these
commentators if one is to make any serious critical use of their work, and this
cannot be done if one merely dips into their voluminous works in the hope of
occasional enlightenment.

1
A paper read to the Southern Association for Greek Philosophy at Cambridge, Sept. 1973. A much
abbreviated form of some of this material was included in a communication to the 14th International
Congress of Byzantine Studies. Note: References to the commentators are by page and line of the
Berlin Academy edition. All unspecified references are to their de Anima commentaries.
2
For this attitude cf. K.O. Brink, ‘Peripatos’, RE supp. vol. 7, 1940, 947; W.K.C. Guthrie, History of
Greek Philosophy I, Cambridge 1962, 367, writing on the pre-Socratics: ‘Simplicius . . . was a learned
and careful commentator . . .’; E.R. Dodds, s.v. ‘Simplicius’, Oxford Class. Dictionary2, Oxford 1970,
speaks of his ‘learned and sober commentaries’.
3
cf. W.D. Ross, Aristotle. De Anima, Oxford 1961, 43; W. Theiler, Aristoteles, Über die Seele, Berlin
1959, 82f.

327
328 Aristotle Transformed

That these men were swayed by their own opinions and preconceptions is
perhaps obvious once stated. Even Simplicius, notwithstanding his reputation
for careful scholarship, is no exception. Simplicius may have done us a great
service by preserving fragments of the pre-Socratics, but he was nevertheless a
man who entertained ideas which were not likely to lead to the correct
interpretation of Aristotle, as Hicks for one saw4 – Ross it seems did not.5 In fact
one might go so far as to say that Simplicius was less well fitted than some of the
other commentators to give a good account of his subject. Those whose
immediate reaction to such a statement is that it is grossly unfair to so fine a
scholar might be disturbed by some of the material in the preface to Simplicius’
de Anima commentary – as they would by that in Philoponus’ as well – material
which often escapes notice for the simple reason that one normally refers to
these works for help with specific passages and does not read them as a whole.
This is not to say that there are no obvious signs of what is going on in the body
of the commentaries, for there certainly are. A case in point is Simplicius’ claim
in the de Caelo commentary (640,27–32) that Aristotle’s criticisms of Plato are
directed not against Plato himself, but against those who failed to grasp Plato’s
real meaning. In the preface to the commentary on the Categories Simplicius
goes further and says that in dealing with Aristotle’s attacks on Plato one should
not consider only the philosophers’ language and complain about their discord,
but rather one should concentrate on their thought and seek out their accord on
most matters (in Cat. 7,29–32). Here we have two expressions of the normal
Neoplatonic view that Plato and Aristotle were usually trying to say the same
thing. This view can of course be traced back to the revival of positive teaching
in the New Academy.6 This is not to say that no Neoplatonist was aware of the
differences, and certain Aristotelian doctrines remained unacceptable. In the

4
Aristotle, De Anima, Cambridge 1907, lxv. Cf. also P. Wilpert, ‘Die Ausgestaltung der aristotelischen
Lehre vom Intellectus agens bei den griechischen Kommentaren und in der Scholastik des 13
Jahrhunderts’, in Festschrift M. Grabmann, Beitr. zur Gesch. der Phil. und Theol. des Mittelalters
supp. vol. 3.i, 1935, 451 and n. 20; O. Hamelin, La Théorie de l’intellect d’après Aristote et ses
commentateurs, Paris 1953, 44–7. This long posthumous publication provides the fullest treatment
yet of Simplicius on nous, but is marred by the now outdated view that the Neoplatonists constituted
a more or less unified ‘école d’Alexandrie’, cf. esp. 44 and 56. Though Hamelin makes certain
distinctions between Plotinus, Iamblichus and Simplicius, he takes no account of such differences
as would have obtruded themselves had he considered Ps.-Philop. on 3.5, and in particular he
seems unaware of the typically Athenian elements in Simplicius.
5
cf. his remarks on the interpretation of 3.5, loc. cit.
6
cf. e.g. Cicero Acad. 1.4.17: ‘But on Plato’s authority there was instituted a form of philosophy which
was unitary and harmonious under two names, those of Academics and Peripatetics, and they
agreed materially while differing in designation.’ For the Neoplatonic period cf. K. Praechter, s.v.
‘Simplikios’, RE III a i, 1927, 210f. and the collection of texts in I. Düring, Aristotle in the Ancient
Biographical Tradition, Göteborg 1957, 334–6.
 Neoplatonic elements in the de Anima commentaries 329

passage we have just mentioned Simplicius talks about accord in most matters
(hê en tois pleistois sumphônia, and elsewhere he shows that he is alive to
differences (e.g. in Cael. 454,23ff ), even if he does regard Aristotle as Plato’s truest
pupil (ib. 378,20f)7 or his best interpreter (in DA 245,12). Philoponus, moreover,
actually protested against the view that Aristotle’s attacks on Plato’s ideas were
not directed at Plato himself, a view that seems to have had some currency (cf.
Aet. 3.2 = 29.2–8 Rabe).8 None the less ever since Plotinus, whose adoption of
much Aristotelian thought would be clear enough without Porphyry’s explicit
statement on the point (Vita Plot. 14.4ff ), the new Platonism had been more or
less Aristotelianised: the controversies about whether or not Aristotelian views
could be accepted by Platonists which had been current in the middle Platonic
period were no longer live. By the time Simplicius and Philoponus composed
their commentaries,9 Aristotle’s philosophy had been used as the standard
introduction to Plato for at least two centuries.10 The tendency among certain
modern scholars to see Aristotle simply as a Platonist has a precedent in the
activities of the Neoplatonists: in both cases it depends on a somewhat special
understanding of Plato.
Let us now look at some of the information about his bias that Simplicius
himself provides at the start of the preface to his de Anima commentary (1,3–
21). His purpose is best understood in terms of the word sumphônia, which
appears almost as a slogan. Next to the truth the chief object of study is an
understanding of the views of those who have reached the summit of knowledge,
and therefore we must deal with Aristotle’s work on the soul. Plato has produced
many inspired speculations about it, but these have been adequately and
consistently (sumphônos allêlois) expounded by the commentators on Plato – we
might envy them their success. But in the case of Aristotle there is much
inconsistency among the interpreters, not only as to the meaning of his language,
but also as to the actual content. In other words, one of the usual Neoplatonic
methods of dealing with inconsistency, to argue that it is merely a matter of

7
An accolade elsewhere bestowed on Xenocrates, in Phys. 1165,34.
8
Thus Elias in Cat. 123,1–3, tells us that Iamblichus claimed that Aristotle was not opposing Plato on
the theory of Ideas. In this at least he was not followed by Proclus, cf. Philop. in Cat. 31,7 ff [and for
Philoponus, see Koenraad Verrycken, chapter 11 above – Ed.]
9
In Philoponus’ case we should perhaps rather say published, since his commentary is basically the
publication of Ammonius’ work, as the title proclaims, 1.1 f. On the relation of Philoponus’
commentaries to Ammonius’ courses cf. my paper ‘Did Iamblichus write a commentary on the De
Anima?’, Hermes 102, 1974, 553 n. 56 and the references given there.
10
For one roughly contemporary attitude cf. Marinus Vita Procli 13, who likens Aristotle’s works to
minor preliminary mysteries. A more sober view, perhaps characteristic of the Alexandrian school,
is indicated by Elias in Cat. 123,10f, who merely says they are an introduction.
330 Aristotle Transformed

different terminology applied to the same truth,11 is not available. Therefore, says
Simplicius, he decided to seek and set out Aristotle’s consistency – sumphônia
again – both with himself and the truth. That statement in itself might give rise
to doubts about the objectivity of Simplicius’ interpretation, doubts which are
heavily reinforced by what follows. For Simplicius goes on to say that he will try
as far as possible to stick to the truth of things in accordance with the exposition
of Iamblichus in his own writings on the soul. This, he claims, is the same task as
the exposition of Aristotle in terms of Aristotle’s own philosophy. Here we are
faced with the basic methodological defect of all Neoplatonic commentary, the
conviction that Plato, Aristotle and the Neoplatonists themselves were all
expressing the same truth. Sometimes of course this conviction was shaken, but
as we shall see such doubts often sprang from differences among the Neoplatonists
themselves as to what that truth was. That Iamblichus should be presented
as a witness to it is at least alarming, notwithstanding recent and perhaps
overenthusiastic attempts to rehabilitate Iamblichus, and rescue him from the
reputation of a mere superstitious mystery-monger.12 And it is not simply a
matter of lip service to a figure revered by Athenian Neoplatonists – the
Alexandrian attitude may have been different13 – paid at the beginning of a work
and then forgotten, for Simplicius refers to Iamblichus as an authority on several
occasions in the body of the commentary.14 Had Iamblichus himself written a de
Anima commentary, which he probably did not,15 it is likely that he would have
figured more prominently in that of Simplicius, for in his Categories commentary
Simplicius refers to Iamblichus with great frequency throughout.16 For a similar
attention to Iamblichus’ views we must go to Proclus’ commentary on the
Timaeus,17 and Proclus is not usually reckoned among Aristotelian scholars,
though he did write commentaries on some of the logical treatises.18 If we ask
how close attention to the views of Iamblichus would affect an Aristotle
commentary, the answer is that it would characteristically tend to produce
explanations by the multiplication of entities, for Iamblichus seems to have been

11
See e.g. the passage from Simpl. in Cat. cited above and in Phys. 1336,35ff.
12
So R.T. Wallis, Neoplatonism, London 1972, ch. 4; J.M. Dillon, Iamblichi Chalcidensis in Platonis
Dialogos Commentariorum Fragmenta, Leiden 1973, ch. 2. Cf. also A.C. Lloyd in A.H. Armstrong
(ed.) Cambridge History of Later Greek and Early Medieval Philosophy, Cambridge 1967, 296f.
13
cf. Ammonius in An. Pr. 38,40–39,2 and n. 58 to the article cited in n.9.
14
cf. e.g. 89,33–5, 187,36f and esp. 313,6f.
15
On this see the article cited in n. 9.
16
Roughly once in 2½ pages.
17
The relevant passages are now conveniently assembled by Dillon, op. cit. 106–205.
18
References to these are listed by L.G. Westerink, Anonymous Prolegomena to Platonic Philosophy,
Amsterdam 1962, xii n.22.
 Neoplatonic elements in the de Anima commentaries 331

at least in part responsible for the complex mediations and triadic schemes of
that Neoplatonism which is attested in the work of Proclus, where the relation
between any two terms is likely to be explained in terms of a third which has
some but not all the characteristics of each.19 Hence a tendency to deal with
certain problems about the soul by splitting it into components in a way that
Aristotle did not, and returning to a position in some ways more like Plato’s,
though informed by a somewhat different view of the relations between the
parts. Thus some sort of compromise was possible between Aristotle’s view that
body and soul were inseparable and Plato’s that they were not. The latter was the
basic position of the Neoplatonists, though they thought in terms of degrees of
separability, a matter which we must consider further in relation to their
comments on Aristotle’s definition of the soul.20 For the moment it will suffice to
say that Neoplatonists saw the individual soul as a descending succession of
forms or as a series of reflections from, or images of, its highest part.21 Thus for
instance the sensitive soul was a reflection of the rational soul, the vegetative of
the sensitive and sometimes a further sub-vegetative phase was distinguished
whose function was to inform matter,22 so producing body, which derived life
only from the layer above.23 At times, however, this information of matter was
seen as a function of the world soul:24 in the last resort individual souls were
identical both with the world soul and with each other.25 As one descends down
the series the soul becomes more closely tied to body. Hence it is possible to talk
in terms of information and to see the body: soul relation as one between matter
and form, at least at the lowest level, in spite of the fact that soul was strictly
separate from body, and even at this lowest level was regarded as illuminating or
irradiating matter rather than being its form in any Aristotelian sense.26 This is

19
cf. now Wallis, op. cit. (n.12) 125ff, who may, however, exaggerate Proclus’ dependence on Iamblichus.
That Iamblichus invented at least the henads has now been shown by Dillon, ‘Iamblichus and the
origin of the doctrine of henads’, Phronesis 17, 1972, 102–6. [Dillon now qualifies his view in
response to Saffrey and Westerink in his ‘Iamblichus of Chalcis’ in W. Haase (ed.), Aufstieg and
Niedergang der römischen Welt, 36.2, Berlin 1987, 862–909, at 883–4. – Ed.]
20
See below, pp. 321–2.
21
Terms like image, illumination, appearance (eidôlon, ellampsis, indalma) are frequently found to
express this relation, cf. Plot. 1.1.11, 5.9.6.15–19, 6.5.8., Philop. 195, 2ff, 202,22ff, Simpl. 85,18.
22
For these distinctions in the lower sections of the soul cf. Plot. 4.4.17–21 passim., Simpl. 51,28ff.
23
cf. e.g. Plot. 6.4.15.8ff.
24
Plot. 4.3.7.25–8.
25
Hence Plotinus at least felt no contradiction between statements that the lowest level of soul comes
from the individual soul and others that it comes from the world soul, cf. my Plotinus’ Psychology,
The Hague 1971, 27ff.
26
The Timaeus rather than the middle books of the Metaphysics would probably be seen as the
appropriate authority on the relation of form to matter, though the relevant section of that dialogue,
50Bff, is not, as far as I know, referred to in discussions about body and soul.
332 Aristotle Transformed

probably at least part of the reason for an apparent ambivalence about the
body:soul relation which we find at various points in the commentaries. As
for terms, either the sub-vegetative irradiation, or the vegetative soul, or
both together, could be described as phusis (nature), while psukhê (soul) might
be used not only as a term for soul in general, but also more specifically
to describe the soul above the level of phusis.27 Thus in his preface Simplicius
makes a distinction between phusis, which is the informing cause (eidêtikê
aitia), which makes the thing informed capable of the movement characteristic
of what is alive, and another cause which is responsible for the movement: that
is psukhê (4,14ff ).
Several degrees of separability, as opposed to the two which might legitimately
if arguably be found in the text of Aristotle, appear also in the preface to
Philoponus’ commentary. This preface too provides us with clear indications that
what we are about to read is not necessarily straightforward exposition of
Aristotle. For a start, we are confronted with a series of appeals to Plato (pp. 1–3).
Further on we are told that psychology is important because it affects our whole
life: allôs te ti oikeioteron hêmin tês hêmôn autôn gnôseôs (12,14f).
That was a slogan of the time, and one of the reasons why the First Alcibiades
came at the start of the Plato course, a practice that we learn from Proclus went
back at least as far as Iamblichus.28 Having said that, Philoponus announces his
intention of showing (1) that all soul is incorporeal, (2) that only the rational
soul is separable from all body, (3) that the irrational soul is separable from the
gross body (tou men pakheos toutou sc. sômatos) but inseparable from the
pneuma, particularly if there is a pneumatic body, (4) that the irrational soul
survives for a time after its departure from the body, (5) that the vegetative soul
(phutikê psukhê) has its being in the gross body and perishes with it (12,15–22).
Here we have a number of un-Aristotelian notions which were part of the
general body of Neoplatonic thought. First there is the gradation of degrees of
union between body and soul parallel to that which we have already noticed in
Simplicius, whose description of nature (phusis) as informing cause (eidêtikê

27
This usage is often found in Plotinus, cf. e.g. 4.4.18.1ff.
28
in Alc. 11. Proclus gives us the reasons for it in the first 7 chapters of this commentary. Ch. 11 need
not, as Dillon thinks, op. cit. (n. 12) 264, show that Iamblichus himself established this order of
reading the dialogues: it could simply indicate that Proclus is giving Iamblichus as an authority for
the system. Albinus had already prescribed Alc. I as a start for suitably disposed individuals,
Eisagoge 149.31ff. H, and H.-R. Schwyzer, ‘Plotin und Platons “Philebos” ’, Rev. Int. Philos. 24, 1970,
185, points out that the twelve dialogues in Iamblichus’ canon, with the addition of Rep., Laws and
the Letters, provide nearly all the references to Plato in Plotinus. On the order in which the dialogues
were studied cf. Westerink, op. cit. (n. 18) xxxviiff and A.J. Festugière, ‘L’ordre de lecture des dialogues
de Platon aux Ve/VIe siècles’, Mus. Helv. 26, 1969, 281ff.
 Neoplatonic elements in the de Anima commentaries 333

aitia) implies, if it does not actually state, that the survival of that level of soul is
closely linked with that of the body. And we certainly have the same distinction
between an irrational soul (alogos psukhê) and a lowest level, as opposed to that
simple dichotomy into a rational and an irrational part which may sometimes be
found in Aristotle.29 Thus irrational (alogos) in Philoponus and Simplicius has a
special sense, namely the part between reason on the one side and the vegetative
soul on the other. Most striking is the appearance of the pneuma. The slight
hesitation implied by ‘again if this pneumatic body exists at all (eti ei holôs estin
auto touto to pneumatikon sôma)’ suggests that Philoponus may have had doubts
as to whether that part of the soul was simply closely bound up with that connate
(sumphuton) pneuma which appears in some of Aristotle’s works, but never in
the de Anima, or whether it was actually attached to the full-scale pneumatic
body which the later Neoplatonists used as an intermediate entity to explain how
immaterial soul could be associated with material body, a belief of which there
are already hints in Plotinus,30 but which did not become a fixed and important
part of Neoplatonic thought until some time afterwards.31 Other passages in
Philoponus’ commentary can show that he did in fact accept the usual
formulation.32 Simplicius too believed in such an entity. Discussing why fire
cannot be a living thing in spite of its association with a soul higher than ours
(73,25ff ), he simply assumes the need for a vehicle (okhêma) – the standard
technical term for the pneumatic body – which joins soul to the body. With this
one might compare Proclus’ statement that the beginning of mortal life comes
when the demiurge provides soul with its vehicle (in Tim. III 234,3–5D).33 It may
by Simplicius’ time have become generally accepted that the doctrine was
Aristotelian, for Proclus had already spoken of Aristotle assuming a pneumatic
vehicle (ib. 238,20f).34
What cannot easily be read into Aristotle is a further elaborated version of
this theory which appears in Proclus (e.g. in Tim. III 238,2ff ) and also Philoponus
(18,22–33), namely that there was not one such vehicle for the soul but two, one
immortal and one perishable, the second being that which Proclus finds in
Aristotle. In any case neither Proclus, Philoponus nor Simplicius bothered to
mention that the de Anima says nothing about connate pneuma, and that in the

29
e.g. EN 1102a27–32, Pol. 1260a6f.
30
So at 2.2.2.21f, 4.3.9.3–9: on the significance of these cf. Plotinus’ Psychology 139.
31
cf. Dodds, Proclus. Elements of Theology2, Oxford 1963, app. II .
32
e.g. 17,19–23, 164,11–13.
33
Interestingly both Simplicius and Proclus use the term ‘of mortal type’ (thnêtoeidês), though
Simplicius applies it to the body, Proclus to the moral life.
34
Proclus of course found it in the Timaeus, 41Cff, cf. in Tim. III 233,23ff.
334 Aristotle Transformed

de Generatione Animalium it is part of the body’s mechanisms.35 They were

however being true to the aim of making Aristotle self-consistent through a
knowledge of the whole of his philosophy.36 They were also using the late
Neoplatonic method of explaining the connection between two entities by
interposing a third. Apart from having a semi-material body between body and
soul we have also a part of soul which is separable – from the material body –
and inseparable – from the pneumatic body – between the separable higher soul
and the inseparable lower soul.
Such then are some of the presuppositions with which the commentators
approached Aristotle. They are stated plainly and openly enough: the only
thing that is not stated is the fact that the ideas were not Aristotle’s, and that
is not stated for the simple reason that the commentators were no longer
sufficiently aware of the difference. This unawareness of the differences between
their own views and a legitimate interpretation of Aristotle, which we may call a
first stage of confusion, led, not surprisingly, to a second, where a philosopher’s
view on a subject treated by Aristotle could be seen as commentary on Aristotle’s
treatment. Both can be seen in the discussion of de Anima 3.5 which appears in
what we have as Book 3 of Philoponus’ commentary, but which is clearly the
work of another commentator. This was Stephanus of Alexandria, who was
trained in the school of philosophy there, and then held an official academic post
at Constantinople in the early seventh century.37 The spirit and orientation of
this pseudo-Philoponus commentary are, however, sufficiently similar to the
fifth- and sixth-century commentaries for the time lapse to be of no great
significance.
Stephanus’ discussion of 3.5 gives a doxography presenting the views on the
meaning of active intellect (nous energeiâi) of Alexander, Plotinus, Plutarch of
Athens and Marinus, Proclus’ pupil, biographer and successor. These opinions
are presented as the various solutions of commentators to a problem of
interpretation: ‘the commentators resorted to many different roads (hoi exêgêtai

35
On connate pneuma here cf. A.L. Peck’s Loeb edition of GA, App. B.
36
Stated explicitly by Elias in Cat. 123,7–9.
37
On the authorship of Bk. 3, cf. M. Hayduck on p.v. of his edition, CAG 15, 1897. Though Hayduck
did not claim certainty, Stephanus’ authorship is now generally accepted. Cf., however, M. de Corte,
Le Commentaire de Jean Philopon sur le troisième livre du “Traité de l’Ame” d’Aristote, Bibl. de la Fac.
de Phil. et Lett. de l’Univ. de Liège, Paris 1934, xf. This edition of the partial Latin translation which
is all that survives of Philoponus’ commentary on Bk. 3 has now been superseded by G. Verbeke,
Jean Philopon. Commentaire sur le de Anima d’Aristote. Traduction de Guillaume de Moerbeke,
Louvain/Paris 1966. Though longer, this commentary pays less attention to doxographical detail
than our Greek Bk. 3. For the purpose of making comparisons it has seemed best to stick to the
Greek texts in the first place.
 Neoplatonic elements in the de Anima commentaries 335

epi pollas kai diaphorous hodous etrapêsan).’ They were accepted as such by Ross,
who complained, not unreasonably on that premise, about their inadequacy.38
Fortunately there is no need for the present discussion to lose itself in the
labyrinth of what Aristotle meant by his distinction between the active and
passive intellects – or functions of intellect. No one who has consulted them on
this question would expect the commentators to help him play Theseus. The
point is that an examination of what they do say can show that their remarks are
virtually irrelevant to the solution, and one might go so far as to say that if any of
those whom Stephanus cites – Alexander excepted – had given us Aristotle’s
meaning, whatever it was, that would have been more or less fortuitous. This
assertion clearly requires justification, but such justification is not difficult to
produce.
Let us start with Plotinus, for Alexander’s opinion that the active intellect
was the highest unmoved mover is irrelevant to our present purpose. Stephanus
tells us that Plotinus thought the active intellect was the human intellect,
but held that some part of it was always active: Plôtinus de energeiâi noun
enoêse ton anthrôpinon noun, ton men aei energounta, ton de pote energounta
(535,8–10). This report is at first sight puzzling, and is in fact inconsistent with
what we read a few lines later, where we are told that Plotinus says that by active
intellect Aristotle means the human intellect which is always in act: phêsin oun
ho Plôtinos hoti energeiâi noun legei ton anthrôpinon noun ton aei energounta (ib.
12f). Now we know that, in so far as the distinction between active and passive
intellect can be applied to Plotinus’ views on nous, it is the second of these two
statements that fits rather than the first. The first can, however, be explained in
terms of what Plotinus does say in the Enneads. The second represents what is
simply Plotinus’ view about the human intellect, namely that it is a part, or rather
a faculty or phase, of the human soul which does not descend to be present to the
body – that being how Plotinus will strictly describe the body: soul relation
(4.3.22.1ff ) – but remains in the intelligible world above. Being thus wholly a
part of the intelligible rather than the sensible world, it is necessarily always
active, since eternal activity is a feature of that world. Plotinus explains our
intermittent use of intellect by saying that we think noetically, as opposed to
discursively, when we so to speak switch on to this highest part of our soul; that
is when we focus our attention on it we then become assimilated to it and
participate in the intellection which is always taking place (cf. 1.2.4.25–7,
4.3.30.11–15). As for Aristotle, the only similarity is that Plotinus adopts

38
loc. cit. (n. 3).
336 Aristotle Transformed

Aristotle’s view of intellection as a process of total assimilation parallel to that

involved in vision (cf. 5.1.5.18f).
From the standpoint of the Platonists, Plotinus’ view was unusual, and
Plotinus himself frankly presents it as heterodox. At 4.8.8.1–3 he introduces it
with the words ‘if one must dare to state one’s view more clearly, in opposition to
the opinion of others, not all of our soul descends, but some of it is always, in the
intelligible [realm]’. The status of the human intellect, and in particular Plotinus’
view that it did not descend, remained a controversial issue throughout the
history of Neoplatonism. Most Neoplatonists rejected Plotinus’ view on the
grounds that he offered no adequate explanation of why we are not always
engaged in noêsis, but Damascius, the last head of the Academy, a colleague of
Simplicius and a contemporary of Philoponus, accepted Plotinus’ view, arguing
that an unchangeable substance must have its being and existence in the
intelligible (de Princ. II 254.3–7 (Ruelle)). And he was expounding Plato. It is the
contention of this paper that the views which Stephanus gives us are in the first
place a contribution to this Neoplatonic discussion and were only secondarily, if
at all, produced as Aristotelian exegesis, notwithstanding the way they are here
introduced.
The reservation must of course be made that these two aims were liable to
confusion. But in the case of Plotinus we can say with some confidence that the
view here produced as his opinion on the meaning of Aristotle’s text was not
offered as such. In his case we have, unusually, a complete record of his works
given us by Porphyry (Vita Plot. 24–6), who was Plotinus’ editor as well as
biographer, and the works contained in that list are available to us in Porphyry’s
edition, namely the Enneads. On one subject, the categories, we have a whole
treatise, 6.1–3, part of which, 6.1, is devoted to a discussion and refutation
of the Aristotelian categories. And we do know that some of the Peripatetic
commentaries were studied in Plotinus’ classes.39 But ex professo commentary
on the de Anima did not exist, and Plotinus cannot properly be described
as an exêgêtês (exegete).40 This rules out one of the possible explanations of the
inconsistent reports in Stephanus, namely that Stephanus, or his source, has
simply juxtaposed a piece of Plotinus’ own philosophy and a piece of his
interpretation of Aristotle, an explanation suggested by the way Alexander’s own
de Anima is sometimes confused with his lost commentary on Aristotle’s de

39
cf. Porphyry Vita Plot. 14,10–14.
40
He did of course claim that his three hypostases were merely exegesis of ideas not fully expressed
by Plato, 5.1.8.10–14.
 Neoplatonic elements in the de Anima commentaries 337

Anima even today. There is of course a theoretical possibility that there was a
tradition of Plotinian Aristotle exegesis which was preserved by his pupils but
does not appear in the Enneads,41 but it is probably true that there was no serious
body of unwritten Plotinian philosophy, and it is fairly safe to assume that any
form of commentary on the de Anima would have been recorded, though
certainty is of course impossible.
Another explanation of our text is, however, possible, and in terms of Plotinus’
own works it is a simple one. Plotinus, like Aristotle, uses nous in two different
senses, either, strictly, to refer to the intuitive intellect as opposed to the discursive
reason, or sometimes loosely, to refer to the latter. Confusion rarely arises, and
when necessary, Plotinus will carefully and explicitly distinguish between them
(e.g. 5.9.8.21f). The report that Plotinus thought the intellect consisted of a
permanently and an intermittently active part may be a confused conflation of
what for Plotinus were two distinct levels of soul. Some modern interpreters
have also thought that Plotinus’ intuitive intellect (nous) and discursive reason
(dianoia)42 could be equated with Aristotle’s active and passive intellect, an
equation which, as I have argued elsewhere, is illegitimate.43 The correct
representation of Plotinus’ view about nous in the strict sense is that in Stephanus’
second report. But that it was originally offered as interpretation of Aristotle is
almost certainly not the case. Apart from several occasions where he quotes what
had become a tag about knowledge of immaterial objects being the same as the
objects, which may or may not be in place in 3.5,44 there is no clear reference in
the Enneads to this chapter of the de Anima.45 A sentence at 5.9.2.21f which
distinguishes between an intellect which always thinks and is the true one, and
another which thinks intermittently, is clearly parallel to other passages where

41
A thesis argued by P. Henry, who maintains that a considerable quantity of Plotinian material in
Dexippus’ and Simplicius’ commentaries on the Categories is not based on the Enneads, first in ‘Vers
la reconstruction de l’enseignement oral de Plotin’, Bull. de l’Ac. R. de Belgique ser. 5.23, 1937, 316–
20, and most recently in ‘Trois apories orales de Plotin sur les Catégories d’Aristote’, in Zetesis,
Festschr. E. de Strijcker, Antwerp/Utrecht 1973, 234–65; contra Schwyzer, ‘Nachlese zur indirekten
Überlieferung des Plotin-Textes’, Mus. Helv. 26, 1969, 261ff.
42
Recent work on the so-called Arabic Plotiniana, which had sometimes been taken as dependent on
Plotinus’ lectures, so Henry, ‘Vers la reconstruction . . .’ 320ff, indicates that differences from the
Enneads are probably due to Porphyry. For the status quaestionis see my ‘Plotinus in the light of
twenty years’ scholarship, 1951–71’ in W. Haase and H. Temporini (eds.), Aufstieg und Niedergang
der römischen Welt, vol. II .36.1, Berlin/New York 1987, 536–41.
43
So P. Merlan, Monopsychism, Mysticism, Metaconsciousness, The Hague 1963, 10; cf. Plotinus’
Psychology 104 n. 10.
44
On this see Ross ad loc.
45
References to an intellect that is or may be separate (5.3.3.42, 5.9.3.7), or the statement that the first
intelligible is separate, need not be to this chapter, since that is hardly the only place in Aristotle, or
elsewhere, where immaterial entities are spoken of as separate.
338 Aristotle Transformed

Plotinus distinguishes between intellect and discursive reason (e.g. 6.9.5.7ff.).

Thus it is not discussing two elements in the operation of intuitive thought,
though it is perhaps the text which could most easily have been the ultimate
source for Stephanus’ statement. Be that as it may, we can conclude that when
Stephanus says that Plotinus says that by active intellect Aristotle means the
constantly thinking human nous, he is in fact producing as interpretation of
Aristotle a view that was not so intended, but was merely part of Plotinus’ own
psychology.46
Since then we can show that this was the status of the Plotinus view in our list,
we must at least consider the possibility that the same may be true of the others.
Plutarch of Athens, we are told, thinks that we have a simple rather than a double
nous, and that it thinks intermittently, a view of which Stephanus himself
approves (13–15).47 This view, however, happens to be that of the anti-Plotinian
side of the Neoplatonic controversy about the undescended intellect. The
formulation, ‘Plutarch does not say that it always thinks, but that it sometimes
thinks’, recalls the last proposition, 211, of Proclus’ Elements of Theology, a work
which has never been suspected of being Aristotelian commentary. Proclus there
argues that the undescended intellect would always think, and so according to
him would not be part of a soul at all: uninterrupted intellection would be
characteristic of intellect rather than soul. This is probably the distinction that
Plutarch is making in different words, when he says that active intellect is the
human intellect which sometimes thinks (535,15f). Stephanus himself thought
that in so far as actualisation is caused by another intellect, that is some other
human intellect (539,1ff ).48 At 536,2–5 we are told that Plutarch justified his view
by saying that Aristotle’s work was about soul. This type of argument is
characteristic of Neoplatonic exegesis after Iamblichus, who either devised or
canonised the system of assigning a single purpose (skopos) to each of the works
of Plato, and extended it to Aristotle’s as well.49 Their contents were then
interpreted in such a way as to make them fit with the stated aim. The results can
be imagined if we consider the likely consequences of taking the Sophist as a
theological dialogue, as Iamblichus apparently did.50 That Plutarch is reported as

46
All this is not to deny that Aristotle’s distinction between active and passive intellect may have had
something to do with Plotinus’ distinction between intellect and discursive reason.
47
As did Philoponus himself ad loc., 45,53–9V.
48
So too Philoponus, 55,4–7V.
49
On Iamblichus’ methods cf. Praechter, ‘Richtungen und Schulen im Neuplatonismus’, in
Genethliakon C. Robert, Berlin 1910, 138ff.
50
According to the scholium vetus, printed in the Teubner Plato, VI .249,26f, Iamblichus said it was
about the sublunary demiurge.
 Neoplatonic elements in the de Anima commentaries 339

solving the active intellect problem in terms of the treatise’s overall aim does
suggest commentary, and we do have evidence, apart from the references in
Stephanus’ commentary on Book 3, that Plutarch did produce a commentary on
at least part of the de Anima, if perhaps not the whole.51 But we know that
Plutarch read the de Anima with Proclus who was his pupil,52 and so is likely to
have followed his master’s teaching on many points.53 Now Proclus, as we know
from the Elements, took the same view as Plutarch on the status of the intellect.
Proclus makes no suggestion there, or in other passages when he discusses the
matter,54 that he is himself interpreting Aristotle, and so we may justifiably
suspect that what we have before us is a view on the human intellect shared by
both Plutarch and Proclus which was imposed by Plutarch on his reading of the
de Anima. The usual motivation, seeing Aristotle’s philosophy as a manifestation
of Neoplatonic truth, will have played its part.55 Further, there is still the
possibility that Plutarch’s view as here presented was originally no more a part of
a de Anima commentary than was that of Plotinus which we have just considered,
and against which both Plutarch and Proclus were arguing.56 But since in
Plutarch’s case we know that he did comment on at least some of the de Anima,
whereas in Plotinus’ we know that he did not, we should perhaps accept that
Plutarch did express this view in the course of a commentary.
That it was possible to do so we may see from Simplicius. He states his
adherence to the same side in the controversy when he is talking about his own
views on the soul in the preface to his de Anima commentary, and he proceeds
to introduce it into his discussion of 3.5. In the preface Simplicius writes that he
will not follow Plotinus in maintaining that some part of the soul remains
constantly pure and unchanged and does not proceed: it descends as a whole
though it does not lose its nature in its inclination to the sensible. This position,
he says, will be shown by the whole account to be the opinion of Aristotle, set out
more clearly by Iamblichus (6,12–16). The reference to Iamblichus, apart from
what it tells us about Simplicius’ approach, is puzzling, since the reports we have

51
Commentary is indicated, if not proved, by such references as Simpl. 21,35ff, 186,26ff, and Priscian
Metaphr. in Theophrastum 34,7f.
52
Marinus Vita Procli 12.
53
It might be argued that Plutarch’s influence will not have been great, as he died two years after
Proclus came to him, Marinus ib. 12. But he had also taught Syrianus, to whose care he entrusted
Proclus, ib.
54
cf. in Parm. 948,18f (Cousin)2, in Tim. III 333,28ff.
55
Plutarch should, however, have been aware of the hazards, for Philoponus tells us that he had
accused Alexander of presenting his own philosophy dressed up as commentary (21,20–3).
56
in Tim. 333,29f is directed specifically at Plotinus and Theodore of Asine: ‘preserving in us
something impassive and always thinking’.
340 Aristotle Transformed

of his position in this controversy are not entirely consistent. Proclus agrees in
reporting that he opposed the undescended intellect theory (in Tim. III .334,3f),
but another text from Simplicius, this time from the Categories commentary,
tends to suggest the opposite, for there we read: ‘whether in us too there is some
such (sc. knowing) which always remains above, as Plotinus and Iamblichus
think’ (191,9f). The close connection between Aristotle exegesis and Plato
interpretation which appears in the similarities we have already noted between
Proclus and the other participants in the controversy, is further evidenced by
the fact that Damascius’ approval of Plotinus’ position is expressed in the course
of what is a commentary on the Parmenides,57 and that Hermeias’ Phaedrus
commentary provides another example of the opposite view (160.1–4
(Couvreur)). Its appearance in Hermeias is interesting from the historical point
of view, for it points to Plutarch – if not Iamblichus, who would be the obvious
candidate were it not for the difficulty just mentioned – as the man who may
have been responsible for the adoption of this position by all those we have
mentioned. For Plutarch taught Syrianus58 and Proclus,59 Syrianus taught
Proclus60 and also Hermeias,61 the father and teacher of Ammonius, whose
lectures are represented by Philoponus’ de Anima commentary (cf. 1,1f) and
were attended by Simplicius,62 while the tradition of Ammonius and Philoponus
is probably represented by Ps.-Philoponus on Book 3 of the de Anima.63
Ammonius himself had studied under Proclus.64 Nevertheless it should perhaps
be said that shared opinions do not necessarily follow from common discipleship.
Simplicius and Damascius65 were both pupils of Ammonius but remained
capable of disagreement.
To return to Simplicius. He fulfils his promise to demonstrate his view about
intellect in his discussion of 3.5 (240,1ff ). Though Simplicius’ Aristotelian
scholarship is typical of the Alexandrian milieu where he went to Ammonius’
courses, rather than the Athenian where he was Damascius’ pupil and later
colleague,66 his treatment of 3.5, as of many other topics, is set out in language

57
See p. 313 above.
58
Procl. in Remp. II 64,6f (Kroll).
59
Procl. ib. and Marinus Vita Pr. 12.
60
Marinus ib. 12, 13.
61
Damascius Vita Isid. fr. 120 (Zintzen).
62
cf. in Cael. 271,19, 462,20.
63
The exact relation between Philoponus’ and Stephanus’ commentaries on Bk. 3 is not yet clear. A
possible intermediary might be the lost de Anima commentary by Olympiodorus.
64
Damascius Vita Isid. fr. 127.
65
ib. fr. 119.
66
cf. in Phys. 774,28f, Agathias 2.30.
 Neoplatonic elements in the de Anima commentaries 341

that one would normally associate with the Athenians. That a piece of
commentary should be written in Neoplatonic technical language does not of
course mean that it cannot provide a useful interpretation of Aristotle’s thought,
though Aristotle’s own discussions of pre-Aristotelian ideas should serve as
ample warning of the risks. There is even a passage elsewhere in the de Anima
commentary where Simplicius indicates that he is aware of language differences
between Aristotle and himself (102,12ff ). Yet it is probably fair to say that one
could read a large part of this discussion without realising that it is about
Aristotle’s philosophy at all. The main point at issue in the first part is the level of
soul that is being discussed. And in the vocabulary of the Athenian school the
fully descended human intellect is not nous at all but psukhê (soul). It is, as we
have seen, such a fully descended intellect in which Simplicius himself believed.
Unlike Plutarch and Stephanus he accepted that our nous is double (240,8), a
part proceeding and another not (244,37–9), a view that may be seen as a
symptom of his greater preference for solving problems by multiplying entities.67
But like Plutarch (Philop. in DA 536,2–5), he produces as an argument for the
meaning of the particular text under discussion the overall purpose of the
treatise as he saw it, not methodologically the best possible approach. Simplicius
begins by arguing that Aristotle is talking about the rational soul since soul is the
subject of the treatise. He does however back up this contention by drawing
attention to the words ‘in the soul’ (240,4f). This for him means that the active
intellect is not transcendent, as it could still be on the Plotinian type view which
regarded a transcendent intellect as part of the human soul, a situation well
expressed in Plotinus’ paradoxical remark at the end of Enn. 1.1: ‘for this too is
part of us and we ascend to it.’
In Simplicius’ terms the discussion is not about the intellect participated by
soul, or the one above that (240,2f, 247,41f). The one above that would be what
others called divine intellect.68 Simplicius calls it unparticipated intellect
(244,40f), as Proclus often does.69 Here we have the standard set of three terms,
pure A, pure B, and between them A participated by B, the series in this case
being unparticipated intellect, soul and participated intellect.70 At 244,39ff these

67
See below. It should be noted that when Simplicius talks about our nous being double he is not
simply reproducing Aristotle’s distinction. For Simplicius’ passive intellect is sometimes equated
with imagination (phantasia), cf. 17,2–5, 248,2ff; cf. also Philop. 490,18ff.
68
That is other Neoplatonists, cf. e.g. Syrianus in Metaph. 142,19. Procl. El. Th. 181ff: the meaning is
not of course the same as Alexander’s divine intellect, de An. Mant. 112,19ff which is the supreme
principle; cf. also Philop. 37,28f, 165,8–13.
69
Proclus uses the two terms as equivalent, El. Th. 163.
70
For this series cf. e.g. Proclus, ib. 166.
342 Aristotle Transformed

two levels of intellect are distinguished from the intellect which has its existence
in the highest part of the soul, and which Simplicius himself would not normally
call intellect without qualification. At 245,37 it is called the soul’s intellect: shortly
before we read that intellect without qualification (haplôs) is that above the level
where the highest part of the soul’s knowledge is simultaneous intellection and
knowledge (ib. 13–15). In accordance with such distinctions Simplicius is also
able to find a difference of levels between intellect described as separate
(khôristos) and separated (khôristheis) (DA 430a16,22): the latter is not in
substance activity in the same way as the former (245,21f). There is in our
intellect when separated a less perfect degree of unity. In Neoplatonic thought
higher degrees of unity correspond of course with higher degrees of being, and
the possibility of distinguishing between intellect, its object and its act was for
them sufficient reason to refuse to recognise Aristotle’s unmoved mover as the
supreme principle.71
Faced with this series of Neoplatonic terms and ontological distinctions, one
must exercise some care in differentiating Simplicius’ own views from the piece
of Aristotelian doctrine he is professedly elucidating. But this does not mean that
the two are always inextricably confused: they can be disentangled in places by a
process of translation. Thus for him, as we have seen, rational soul does not mean
discursive reason but intellect, as it did for Philoponus too (165,5–9). Its
attributes, however, tend to be determined by the conceptual structure within
which Simplicius himself operated.
It is in terms of this structure that one can make some sense of the views of
Marinus reported by Stephanus, for Marinus was Proclus’ pupil and successor as
head of the Academy. Marinus, we are told, held that the active intellect was not
that which is principle of all things, but some demonic or angelic intellect
(535,5f). At first sight this might suggest medieval interpretations rather than
the de Anima. Once again what we have is straight Neoplatonism. Marinus too is
working with a series of highly, not to say excessively, differentiated entities: the
angelic intellect is neither ours, nor that of the principle of all things. Though he
has located our intellect at a different level from that chosen by Simplicius,
the layers involved are the same. We can see from Proclus’ Timaeus commentary
(III 165,7–25) that the kind of intellect or being described as angelic is an
intermediary between the divine intellect (theios nous), what Simplicius called
unparticipated intellect, and the level of soul in which our world is located and
which both Simplicius and Proclus call the secondary (ta deutera). In fact it is

71
cf. e.g. Plot. 5.3.13.9ff.
 Neoplatonic elements in the de Anima commentaries 343

angelic in the sense of being a messenger and interpreter: it interprets and

translates the divine intellect for the secondary. It proceeds in accordance with
intellectual life (kata tên noeran zôên) and intellectual intellect (nous noeros) is
the middle participated term below divine intellect (theios nous) in another of
Proclus’ series (cf. El. Th. 175 and 183). The same passage of Proclus’ Timaeus
commentary also tells us why Marinus, having proposed demonic or angelic
intellect as alternatives, chose the latter: it is the angelic intellect which performs
the function of messenger and interpreter in accordance with the intellectual
capacity of nous rather than its demiurgic capacity, in accordance with which the
demonic intellect orders nature and the cosmos (ib. 25–7). The third member of
the triad, the heroic (to hêrôikon) does not concern us here. The point of placing
Marinus’ view in Proclus’ triadic schemes is simply to show that in Marinus’ case
too the opinion about active intellect consists of no more than locating it at some
point in his own intelligible universe. It could be salvaged as a piece of Aristotelian
exegesis if it could be shown that Marinus tried to argue that Aristotle’s active
nous should be identified with some entity between man and the first unmoved
mover, but we have no evidence that he did. Though it can perhaps be shown that
the Arabic commentators’ interpretation of the active intellect as an ‘intelligence’
stems from Marinus’ opinion, that would not in itself show that Marinus was in
fact engaged in the interpretation of Aristotle. Though Marinus will doubtless
have given lectures on Aristotle, for no other reason than that he taught
philosophy, it may be to the point that he is not normally cited as an authority on
Aristotle and that, though Stephanus is quite generous with citations of earlier
authorities, Marinus’ name is not mentioned outside the page on 3.5. His status
as an Aristotelian is therefore not very different from that of Plotinus.
Much more could be said about the details of these Neoplatonic theories
on nous and the upper soul: as we have already seen there are important
differences. But such details belong to the study of Neoplatonism as such. For the
present purpose the type of thought involved is more important, and for the
moment it should be sufficiently clear just what the ‘interpretations’ we have
been considering in fact are, that is personal positions in a live philosophical
debate.
Simplicius’ commentary is perhaps more thoroughly pervaded by
Neoplatonism than were others. This is a factor that should be taken into account
in looking at the commentators’ views on other questions, though it will be safer
to make one initial reservation. That is that the status of our intellect was clearly
a more important question to these men than were, at the other extreme, details
of the soul’s operations. In this area Neoplatonic doctrine was fairly thoroughly
344 Aristotle Transformed

Aristotelianised in any case. But on questions falling between these extremes the
extent to which the commentators read their own views into Aristotle could, in
principle, be less marked.
A totally different approach may be seen in Themistius, who was not involved
in this debate and was, moreover, still trying to be an Aristotelian. The aims he
states at the start of his de Anima paraphrasis are to follow Aristotle, to clarify
and if necessary expand him (1,2–5): there is no mention of the truth, no appeal
to Plato or to figures in the Platonic tradition. Themistius was of course writing
some two centuries before Simplicius and three before Stephanus but
Neoplatonism was already the dominant philosophy: Plotinus, Porphyry, and
Iamblichus were all dead before Themistius’ career had begun.72 That does not
mean that he excluded Platonic or Platonist ideas.73 That was no longer possible
for an Aristotelian, any more than it was possible for a Platonist to refuse to
accept any of Aristotle’s notions. But his orientation was certainly different from
that of the Neoplatonists. His attitude to Plato was independent: he followed him
occasionally if he thought fit. Simplicius remarks on this in his de Caelo
commentary: ‘although Themistius offers the views of the Peripatos in most
things, in this he seemed to favour those of Plato’ (69,9f). A fairly cursory
examination of Themistius’ comments on de Anima 3.5 will make the point
clearly enough. That Themistius appears to be puzzled, and offers little help, is
not to the present point.74 But what he does is, basically, to say what Aristotle
says, if at much greater length, with some attention to the opinions of
Theophrastus (98,12–109,3). The vocabulary is Peripatetic, and the thought
shows no signs of the Neoplatonic multiplication of entities: there are no
participated and unparticipated transcendent intellects, no distinctions between
intelligible and intellectual life, no members of triads. Whether or not it would
still have been possible to write this kind of paraphrase in the sixth century is
another matter.75
Having now established that the late commentators’ own views were likely to
have been the determining factor in their interpretation of 3.5 let us look at their
comments on some other sections of the treatise. It should not now be surprising

72
Iamblichus died in 325 or 6, cf. J. Bidez, ‘Le philosophe Jamblique et son école’, REG 32, 1919, 32.
Themistius was born about 320.
73
Thus he accepted the Platonist concept of philosophy as assimilation to God as far as possible
(homoiôsis tô theôi kata to dunaton) and followed Plato on divine benevolence, cf. Or. 2,32d–33a =
39,4–20 (Dindorf). That he expresses more enthusiasm for Plato in the speeches than in the
commentaries may be a function of the genre, cf. Düring, op. cit. (n. 6) 333.
74
On the difficulties in this account cf. Merlan, op. cit. (n. 43) 50 and n. 3.
75
To do so might well have been seen as a sign of perversity, cf. Damascius on Isidore’s conversion of
Dorus, Suda s.v. ‘Doros’ = Vita Isid. fr. 338.
 Neoplatonic elements in the de Anima commentaries 345

if they turn out to be somewhat over-Platonic. Perhaps the most obvious test
case is Aristotle’s definition of the soul as the form and first actuality of a natural
body that has life potentially and has organs (organikou). A simple scholarly
explanation should perhaps concern itself with explaining how soul could be the
form of the body and why it is first actuality. On this basis Simplicius’ relatively
brief comments (90,29–91,4) would be puzzling. His main concern is with
organikou which might seem to be the part of the definition which requires the
least comment. For him organikou (with organs) is not a description of the
matter in the compound of matter and form which is the living being. It is rather
a term implying that the soul uses the body as an organon (instrument), as well
as causing it to be one in the first place. The idea of soul using body as an
instrument is of course straight Platonism. Simplicius manages to make it
compatible with what Aristotle says by using another piece of Neoplatonic
doctrine, namely the notion that the soul is a series of layers.76 Thus he exploits
the idea of a double entelechy by explaining it in terms of levels, and finds two
separate components in what Aristotle presents as a single definition: having life
potentially and being equipped with organs (organikou). Though, he argues, the
two come together in so far as what potentially possesses life is an organon
(instrument) of the soul, and the instrument of the soul potentially possesses
life, they are not the same. One refers only to the lowest form of life by which the
body is informed and made a living body, the other to that which uses and moves
it. What we have here is that Neoplatonic distinction to which we have already
referred, between a layer of soul which imparts form to the body, and a higher
layer or series of layers which control the already informed body, a distinction
Simplicius had initially made in his preface. One might compare the way Plotinus
treats the concept of entelechy (actuality) as if it deals with how soul informs an
already ensouled body (4.7.8), and Philoponus’ remarks about body and each
section of soul being related to that above it as matter to form (198,22–9). Lest
there be any doubt about Simplicius’ intentions, the final sentence of the present
section shows us what he thinks a definition of the soul should tell us: ‘and that
is why you need both for the complete definition of the soul, so that it can be
made clear that it uses the body as an instrument’. All Aristotle’s care to show that
body and soul form a single entity has gone for nothing.
That could be described as blatant misrepresentation. With Philoponus the
case is different. Perhaps it would be as well to recall that Philoponus’ commentary

76
On this see above, p. 309f and n. 22.
346 Aristotle Transformed

on Book 2 is not the same work as Ps.-Philoponus’ commentary on Book 3, and

therefore we should be prepared for a different approach. As it is, Philoponus
does not comment directly on Aristotle’s revised definition of soul in its full
form, but he gives us a few lines to elucidate the words ‘and any body with organs
is of that sort’ (217,9–15). From these it is immediately clear that he does not
take ‘with organs’ (organikon) as an indication that the body is subordinate to the
soul. He explains that organs are that in virtue of which activities involving life
take place, and therefore that in virtue of which a body has life as its first actuality,
as opposed to being merely a natural body like fire, whose first actuality is the
power of heating and not soul. Here Philoponus has avoided importing
Neoplatonic views of the kind which his preface shows that he held,77 and his
preceding comments on 412a16ff (215,4–216,26) show that he is quite prepared
to explain Aristotle’s concept of soul as actuality in the sense in which it was
intended.
Philoponus’ account of ‘with organs’ seems to be on the same lines as
Themistius’ rather less concise exposition (42,15ff ). Interestingly Themistius says
quite specifically that ‘with organs’ should be taken as an explanation of ‘having
life potentially’ and not as signifying something different (42,35–7). It looks as if
he is objecting to the interpretation later offered by Simplicius, or something
very like it, which both shows that such an interpretation had long been current,
and helps to confirm that Themistius’ position was a matter of personal
conviction rather than mere chronology.
Similar differences appear when the commentators tackle the crux at the
end of the same chapter where Aristotle, having argued that the form is the
inseparable form of the body, ends the discussion by saying that it is not clear
whether it is separable (413a8f).78 Once again Simplicius gives a Platonist
solution (96,3–15) – which might have some merit if Aristotle’s remark were a
piece of Platonism which somehow survived revision, an explanation that was
not of course available in antiquity. Simplicius is disturbed by ‘it is not clear’
because for him it not only makes sense, but is actually correct, to think of soul
as a separable entelechy. He solves the problem in terms of two layers or aspects
which he finds within that phase of soul which we have already met as the one
which uses the body as an instrument: this, he says, in inseparable in the sense
that it always uses the body, and separable in the sense that it uses it in a

77
See above, p. 310f.
78
The reading psukhê <ê> does not help, since the second of the alternatives it produces is no easier
to explain than the sentence as it stands.
 Neoplatonic elements in the de Anima commentaries 347

transcendent fashion (exêirêmenôs), as a subordinate instrument. In accordance

with this distinction he takes the comparison with a sailor in a ship as an
indication that the actuality both uses the body and does not use it at all: in the
latter case it is separable.79 According to Simplicius ‘not clear’ does not apply to
the intellect theôrêtikê because that does not even use the body as an instrument,
that is, it is separable beyond any possibility of doubt. What is unclear, he thinks,
is the status of the irrational faculties, and this will be clarified by investigation
of each of these. Once again we have a set of three terms with two extremes being
bridged by a middle term sharing characteristics of each: soul altogether separate,
elsewhere called rational soul, the inseparable life which in discussing the
definition he speaks of as informing body, and the actuality which uses the body
which is both separate and inseparable.
Simplicius may be given credit for an acute piece of misrepresentation.
Philoponus’ comments (224,10–225,31) are less incisive, but once again show
him trying to understand Aristotle in terms of the de Anima itself, in spite of his
own convictions. The result, not surprisingly, is that he is unable to offer a clear-
cut solution. His attempts to find an explanation show the same kind of
vacillation which appears to be present in Aristotle’s text, and he is even prepared
to entertain the idea that the rational soul itself is inseparable in some sense
(224,28ff ). But in the end he refused to follow Alexander’s suggestion that
Aristotle himself was in doubt, and refers the point to the as yet unresolved
status of the lower faculties.80 To do this, however, he has to say that one can have
an actuality like the steersman in a ship (225,20–31), which of course makes
nonsense of Aristotle’s previous discussion. In contrast to Simplicius, Philoponus’
discussion is complicated by his attempts to find an explanation which does not
simply jettison Aristotle’s concept of the soul as the form of the body.
Themistius, on the other hand, is not distracted by hesitations about the
relation between the soul below intellect and the body. He is forced to explain
‘not clear’ as referring to the status of intellect only and like Philoponus, who
may of course have taken the idea from him, has to deal with actuality by
suggesting that the helmsman is a separate actuality. He may have been aware
that this is unsatisfactory, for he ends by exploiting Aristotle’s ‘let this be a sketch’
(413a9f): he suggests that some of what has been said is not accurate and remains

79
At. 4,18–20 Simplicius uses the comparison for the soul which moves rather than informs the body:
the former is the same layer as that which uses it as an instrument.
80
cf. his comments on 408b29, ‘Perhaps intellect is something impassive and more divine’, where he
takes ‘perhaps’ (isôs) to refer not to doubt but merely to the as yet incomplete state of the discussion,
165,4ff.
348 Aristotle Transformed

to be cleared up when all the soul’s faculties have been discussed. But he does
make it clear that as far as he is concerned the uncertainty affects only the
possibility that some part of the soul is separable (43,25–34).
The passages we have inspected should suffice to show that the personal
standpoint of the commentators was liable to be a major factor in determining
their interpretations. Clearly this will be most apparent when they were
confronted with texts that related, or could be taken to relate, to crucial points in
their own philosophy. But it is not unfair to say that, notwithstanding differences
between commentators, the late Greek commentaries are pervaded by
Neoplatonism. Perhaps enough has been said to show how this affected details of
their interpretation and not just the broad outlines: the effects of this influence
may be found on almost every page of these long and painstaking commentaries.

Addendum: March 1988

Though the de Anima commentary ascribed to Simplicius is not written in the

same way as Simplicius’ others, it was accepted as his by everyone, with the
exception of F. Piccolomini, Commentarii in libros Aristotelis de Coelo, orto et
interitu . . . de anima, Mainz 1608, 1001, until 1972 when F. Bossier and C. Steel
wrote a long and important article in which they maintained that both the
language and doctrine of the de Anima commentary were different from the
other commentaries and closer to that of Priscian, to whom it should be attributed:
‘Priscianus Lydus en de “In De anima” van Pseudo(?) Simplicius’, Tijdsch. voor
Filosofie 34, 1972, 761–822. Their thesis was critically discussed by I. Hadot in an
appendix to her Le problème du néoplatonisme Alexandrin: Hiéroclès et Simplicius,
Paris 1978, 193–202, where she argues that the doctrines are the same, but thinks
that language and authorship are an open question. In Chapter 12 she expresses
further doubts about the Bossier-Steel view, and in particular questions the
significance of verbal parallels with Priscian, but regards the matter as not yet
settled. I have tried to show that there are no doctrinal differences in ‘The
psychology of (?) Simplicius’ commentary on the de Anima’, in H.J. Blumenthal
and A.C. Lloyd (eds), Soul and the Structure of Being in Late Neoplatonism,
Liverpool 1982, 73–93. During 1985 I had the opportunity to discuss the question
with Bossier and Steel on separate occasions: both assure me that they are if
anything more strongly convinced that Simplicius could not have been the author
of this commentary.
 14

The Alexandrian commentators and the

introductions to their commentaries
L.G. Westerink

I shall pass in review the known representatives of the Alexandrian school since
Proclus: Hermeias, Ammonius, Eutocius, Olympiodorus, Elias, David, Stephanus.1
Biographical details are scarce; it is possible, however, in some cases to gain more
exact data than those given in the handbooks. Besides the question which of
them taught Platonic Philosophy, their attitude towards Christianity is also a
relevant point.

1. The Alexandrian School from Hermeias

Hermeias, the first Alexandrian professor to be considered, was also the first to
introduce the doctrine of the Athenian School at Alexandria. At least, we may
suppose so, since he was a pupil of Syrianus, but in fact nothing is known of his
own teaching, for the commentary on the Phaedrus that we possess is only a set
of unrevised notes of Syrianus’ lectures on the dialogue.2 Damascius describes
him as industrious and kind-hearted, but limited.3 His wife, Aedesia, was a
relative of Syrianus, who had originally destined her for Proclus, until a divine
warning prevented the match. Hermeias and Aedesia had at least three sons, but
the eldest died at the age of seven; then came Ammonius and Heliodorus.4

1
cf. my introduction to Elias On the Prior Analytics, Mnemosyne Series 4, vol. 14, 1961, 126–39.
(= Texts and Studies in Neoplatonian and Byzantine Literature, Amsterdam 1980, 59–79).
2
K. Praechter, Byzant. Zeitschr. 18, 1909, 524, n. 4. The suggestion of A. Bielmeier (Die neuplatonische
Phaidrosinterpretation, Paderborn 1930, 33–5) that Hermeias inserted extracts from Iamblichus,
requires further proof.
3
Vita Isidori, Zintzen (ed.), Hildesheim 42, 1907, p. 100,9–12.
4
106,1–108,8; 109,7–9.

349
350 Aristotle Transformed

As Proclus came to Athens only in 432, the earliest possible date for the birth
of Ammonius is about 435; several years later seems much more probable, of
course, but on the other hand Syrianus may well have been in a hurry to see his
beautiful and accomplished charge settled, for Aedesia was then no longer a
young woman, certainly not by the standard of those days. She was an ‘old lady’
somewhere between 470 and 475, when precocious young Damascius made her
acquaintance at Alexandria and soon afterwards recited some of his own poetry
at her funeral. This makes the year 445 the terminus ante quem for Ammonius’
birth. Tannery5 has advanced arguments to show that he was probably alive as
late as 517, when Philoponus published the course on Aristotle’s Physics. They
are confirmed by the fact that Olympiodorus, whose own commentary on the
Meteorologica is later than 565, cannot very well have attended the lectures on
the Gorgias much before 515. If Asmus6 is right in surmising that Ammonius was
no longer alive when Damascius wrote his Life of Isidorus, we must place his
death between 517 and 526.
The only book from Ammonius’ own hand that has survived is his commentary
on the de Interpretatione: it seems to be also the only large work he ever wrote,
though he occasionally used the convenient form of the monobiblos to set down
his ideas on some special question, as Proclus had done before him and
Damascius did after him. To this kind belong the treatise on Phaedo 69D,
defending Plato against the charge of scepticism,7 another on hypothetical
syllogisms,8 and, probably, a work on the Aristotelian Idea of God, cited by
Simplicius.9 On the whole, however, M. Richard10 is right in saying that Ammonius
did not like writing and preferred to leave the publication of his lectures to his
students. Three of these editions bear Ammonius’ own name (Isagoge, Categories,
Prior Analytics), one that of Asclepius (Metaphysics), eight more that of
Philoponus (Categories, Prior and Posterior Analytics, Physics, de Generatione et
Corruptione, Meteorologica, de Anima, de Generatione Animalium.) Philoponus
wrote up his (or perhaps a fellow-student’s) notes and added ‘some observations
of his own’, as the titles of his editions usually state; Asclepius, having nothing of
his own to add, mixed his lecture notes with extracts from Alexander, but only as
far as the fourth book, 5, 6 and 7 being apparently pure Ammonius.

5
‘Sur la période finale de la philosophie grecque’, Revue philosophique 42, 1896, 274–5.
6
Das leben des Philosophen Isidoros von Damaskios aus Damaskos, Leipzig 1911, 186 (note on
110,22.)
7
Ol. in Phaedo. 8 §17, 6–7.
8
Amm. in An. Pr. 67,32–69,28.
9
in Cael. 271,13–21; in Phys. 1363,8–12.
10
‘Apo Phônês’, Byzantion 20, 1950, 192.
 The Alexandrian commentators and the introductions to their commentaries 351

The reputation of Ammonius as a commentator rests entirely on his

Aristotelian studies. Damascius already says that Aristotle was his speciality;11
besides personal preference, caution also may have played a part in this, but
there is no evidence to show that it went so far as to cause a break in the Platonic
tradition at Alexandria. At some time between 475 and 485 Damascius had
attended a course on Platonic philosophy by Ammonius,12 and about forty years
later Olympiodorus heard the old man lecture on the Gorgias.13 Asclepius refers
to lectures on Platonic philosophy in general14 and to a discussion of the
Theaetetus in particular.15 This leaves little room for the hypothesis16 that the
pact he made with the patriarch Athanasius II (c. 489–496)17 obliged him to
confine himself to Aristotle. If nothing has survived of his interpretation of
Plato, the reason is no doubt that the work of the Athenian school was felt to be
more important; just as, on the other hand, Proclus’ commentaries on Aristotle
have left hardly any traces.18
Whatever the concessions Ammonius made, they must have been considerable
enough to make him a renegade in the eyes of Damascius, who refers to the
episode in terms of bitter contempt. The most likely explanation is that
Ammonius pledged himself to silence on certain essential doctrines contrary to
the Christian faith (especially the eternity and divinity of the world).19 A casual
remark in Philoponus’ de Anima commentary from the voice of Ammonius
points the same way: ‘Though it (the soul) may be forced by tyrants to profess an
impious doctrine, it can never be forced to inner assent and to belief,’20 an
unheroic view, which seems to be rather a personal utterance than a stock phrase.
As long, of course, as no systematic effort has been made to detach Philoponus’

11
op. cit. 110,1–2.
12
ibid. 111,10–11.
13
Ol. in Gorg. 199,8–10.
14
77,4.
15
70,31.
16
H.D. Saffrey, ‘Le chrétien Jean Philopon et la survivance de l’école d’Alexandrie’, in Revue des ét. gr.
67, 1954, 400–1.
17
Damascius Vita Isidori, 250,2 and 251,12–14, two extracts from Photius, one of which names the
patriarch. Since the context seems to be the persecution of the philosophers during the rebellion of
Illus (484–488), Asmus (Byz. Zeitschr. 18, 1909, 469–70) concluded that this must be an error for
Peter III Mongus (482–489); accordingly, Zintzen deletes the name of Athanasius. However, the
name must come from Damascius, and the digressive character of his work makes inferences based
on supposed context extremely precarious; it is to be noted especially that the alleged motive of
Ammonius was not fear, but greed.
18
Proclus on the Isagoge: Asclep. 142,36–7; introduction to the Cat.: Elias 107,24–6; lectures on the de
Int.: Amm. in Int. 1,6–11; cf. 181,30–2; Steph. in Int. 46,25–47,12; on the Prior Analytics: Philop. in
An. Pr. 40,30–1, cf. 31,24; 29,2; on the Posterior Analytics: Philop. in An. Post. 111,31–112,36.
19
For other views see Koenraad Verrycken, Chapter 10 above, and also Chapter 1.
20
104,21–3.
352 Aristotle Transformed

‘own observations’ from the strictly reporting parts of the commentaries, it

remains risky to use Philoponus as a source for Ammonius’ teaching. There is, no
doubt, an occasional phrase that has a Christian ring,21 in one case belief in a
God who works miracles is expressed,22 and more importantly, in the commentary
on the Meteorologica there is a consistent attempt to eliminate the fifth element
and with it, by implication, the eternity of the heavens. As a rule, however,
Philoponus repeats objectively and without any comment Ammonius’ hard line
position on such matters as the eternity of the world, the world soul, the pre-
existence and reincarnation of the human soul. All these points, then the main
issues between Christianity and Hellenism, are mentioned in the lectures on the
Metaphysics also, but Asclepius, either from conviction or from caution, always
makes it clear that he is merely reporting Ammonius.23 As regards Ammonius
himself, it is evident that his treaty with the patriarch, whatever its terms, cannot
have had any lasting influence on his teaching; twenty years later he continued
to lecture on Plato and continued to express his unorthodox opinions even in his
courses on Aristotle, which must have been given to a larger and more mixed
audience than those on Plato. Zacharias of Mytilene’s dialogue Ammonius
reflects the reaction of at least one Christian student to his persistent belief in the
eternity of the world. This seems to indicate that the position of the Hellenes had
become easier towards 520 than it was a quarter of a century before.
Olympiodorus calls Ammonius ancestor (progonos) and, though it is perhaps
worthwhile calling attention to the possibility that he may mean simply
grandfather or great-grandfather, the usual explanation that there was another
professor between them is the more obvious one.24 It seems likely now that
Ammonius’ direct successor was the mathematician Eutocius, who gave a course
on the Organon according to Elias On the Prior Analytics p. 134.25 No further
particulars are known. Philoponus never held the chair, for he is always given the

21
in Cat. 126,29–31.
22
ibid. 169,19 (cf. Elias 242,11); not in Ammon. in Cat.
23
89,4–5: if, as they say, the world is eternal. 90,27–8: as our philosophy professor said, the reference is
to the world soul. 171,9–11: there is also an art of medicine on the intelligible level, the reason-
principle in the demiurge, he says, by virtue of which principle the world is free from disease and
ageing. 186,1–2: thus the sky is perishable in its essence, but it continues to exist, he says, owing to
an inflow from the one principle of all things, as long as this inflow lasts. 194,23–6: . . . through the
heavenly bodies, which, he says, are also eternal, but their eternity is no longer of the same kind. As
regards their substance, he says, the heavenly bodies are eternal, but not as regards their action.
226,12–15: for if, he says, the demiurge produces all things by his generative power and needs no
time, but does so without change, why should we say, he says, that he began to produce at a certain
point in time? Their contention is wrong, as (he says) we have shown elsewhere.
24
Ol. in Meteor. 153,7 etc.; cf. propatôr, Proclus in Parm. 1058,22; Marinus Vita Procli 29.
25
Article cited above, n. 1; cf. ibid. 129. His commentary on the Isagoge (20,10–11 Busse) is cited in
Arethas’ scholia on the Isagoge, Vat. Urb. Gr. 35, f. 4r, line 9.
 The Alexandrian commentators and the introductions to their commentaries 353

title of grammatikos (professor of philology), never that of philosophos. H.D.

Saffrey,26 who pointed out the probable connection between the closing of the
Athenian school in 529 and the publication of Philoponus’ book Against Proclus
on the Eternity of the World in the same year, thinks that he may have had a semi-
official position in the school as editor of Ammonius’ lectures. In that case the
object of his book would have been to reassure the authorities as regards the
religious views of the school. It is equally possible, however, that he wrote it on
his own behalf only, to renounce his former allegiance to the school, which
subsequently, under Olympiodorus, confessed its paganism more openly than it
had ventured to do under Ammonius.
The birth of Olympiodorus must be dated before 505, or he could not have
heard Ammonius, and after (or at least not much before) 495, since he was still
teaching in 565. Of his three commentaries on Plato (Alcibiades, Gorgias, Phaedo)
and two on Aristotle (Categories and Meteorologica) that on Meteorologica can
be dated with certainty after 565;27 an anonymous commentary on Paul of
Alexandria (Ps.-Heliodorus), written down in 564, very probably also represents
a course taught by Olympiodorus.28 The evidence in the lectures on the Alcibiades
is contradictory. Creuzer29 thought that they were given before 529 on the ground
of the famous passage on the estate of the Academy: ‘Perhaps Plato made a point
of refusing fees because he was a well-to-do man; that is why the endowment
funds have lasted until the present time, in spite of the many confiscations that
are taking place.’30 I can make no sense of Zeller’s objection31 that diadokhika
(the successor’s, or head of the school’s, endowment) can mean the salary of a
professor of philosophy in any town. The remark is pointless unless applied to
the wealth of the Athenian school. Olympiodorus’ argument had been put
forward before and had been answered by Damascius,32 who explains that,

26
op. cit.
27
in Meteor. 52,30–53,2.
28
J. Warn, ‘Le commentaire attribué à Héliodore sur les eisagôgika de Paul d’Alexandrie’, Recherches de
Philol. et de Ling., 1967, 197–217; L.G. Westerink, ‘Ein astrologisches Kolleg aus dem Jahre 564’, Byz.
Zeitschr. 64, 1971, 6–21 (= Texts and Studies in Neoplatonism and Byzantine Literature, Amsterdam
1980, 279–94).
29
In his ed. of Ol. in Alc., Frankfurt 1821 xiv. On the question in general there are three papers by
A. Cameron: ‘The end of the ancient universities’, Cahiers d’histoire mondiale 10, 1967, 653–73, ‘The last
days of the Academy at Athens’, Proceedings of the Cambridge Philological Society, 195, 1969, 7–29; ‘Le
fin de l’Académie’, in Le néoplatonisme, Actes du colloque de Royaumont, CNRS, Paris 1971, 281–90.
30
141,1–3.
31
Philos. d. Gr. III 24, 917–19.
32
Vita Isidori 212,1–5 = 213,8–14. It is to be noted that Damascius identifies the Academy with the
Athenian School of the sixth century, as does Cameron. I. Hadot (Le problème du néoplatonisme
Alexandrin, Paris 1978, 33) has pointed out that this is an error archaeologically, i.e. as regards the
site. Nevertheless, the institution was considered as being the same, and this is how Damascius
understood his title of Platonic successor (diadokhos Platônikos).
354 Aristotle Transformed

contrary to current opinion, the estate of the Academy did not come from Plato’s
private fortune, but from the legacies of its friends and benefactors in the course
of the centuries. The discussion throws a curious sidelight on the relations
between the school of Athens, proud of its financial independence, and the
Alexandrian university, whose professors were dependent on the salary of the
city and the fees of students;33 it also provides a background to the unkind
comments of Damascius on Ammonius’ alleged avarice.
Zeller’s motive in advocating a later date was apparently the space of nearly
forty years that would have to be assumed between the commentaries on the
Alcibiades and the Meteorologica. This is not a valid objection, for Olympiodorus
may well have come into the professorate before 530, but there are other
chronological difficulties. In the biography of Plato that forms the beginning of
the commentary on the Alcibiades, Olympiodorus tells a story about the welcome
given to Hephaestus as governor of Alexandria by a grammarian, one Anatolius.
He is clearly alluding to an event that had happened several years before: ‘a
certain Anatolius, a grammarian, once won considerable success by quoting this
line in honour of Hephaestus, when he made his entry as governor of the city.’34
Both Hephaestus and Anatolius are mentioned by other contemporaries.
Hephaestus is depicted by Procopius in the Anecdota35 as a hateful creature of
Justinian, who had started his career as a barrister (rhêtôr) and then, as augustalis
of Egypt, was responsible for the corn monopoly and for the confiscation of the
poor fund started by Diocletian. John Lydus36 knew him as praefectus praetorii in
Constantinople and was addressed by him when he left the service in 551. He
also provides the interesting detail that Hephaestus was from a noble Egyptian
family and derived his descent from the first of the Pharaohs, the Hephaestus
mentioned by Diodorus 1,13,3. An Alexandrian grammarian Anatolius occurs
in the Christian Cosmography of Cosmas Indicopleustes: it was at his request
that Cosmas added a seventh book on the eternity of the heavens not long after
547.37 But for the possibility of a close relationship his identity with the one in
Olympiodorus’ story is practically certain. If Olympiodorus gave his lectures
before 529, the episode of the welcoming speech can hardly have taken place
later than the early twenties. This is impossible if Hephaestus owed his

33
cf. Ol. in Gorg. 224,20–4; 225,19–21; 226,24–6.
34
2,80–2.
35
26,35–44. Cf. L. Cantarelli, La Serie dei Prefetti di Egitto III , Atti della R. Accademia dei Lincei, classe
di sc. mor., stor. e filol., vol. 14, Roma 1909, 413–14. The terminus a quo is c. 520 (Licinius), the
terminus ad quem 535 (Dioscorus).
36
de Magistr. III ,30.
37
VII 385 C.
 The Alexandrian commentators and the introductions to their commentaries 355

appointment to Justinian (as Procopius seems to imply though he does not say
so expressly); it is highly improbable in any case, as it would make both the new
governor and the official orator of the city very young men.
The use of the present participle ‘that are taking place (ginomenôn)’ by
Olympiodorus shows that the expropriation of the Academy’s estate took place
only gradually. This leaves room for two explanations: either confiscations had
been going on for some time and the edict of 529 was only the finishing stroke,
or the action was started in 529 and completed in slow stages; all the slower, one
imagines, because the local authorities cannot have co-operated whole-heartedly.
If we take it that they succeeded in dragging the matter out till 532, it becomes
possible to fit in Tannery’s hypothesis that Chosroes in the peace-treaty of that
year obtained the restitution of (a part of) the revenues for his protégés from the
Academy.38 The treaty itself, as reported by Agathias II 31, only stipulated that
‘those men were to be free to return to their country and live quietly by themselves
(eph ’heautois) without being compelled to accept any belief against their
conviction or to renounce the creed of their fathers, but if we decide for the late
dating of the commentary on the Alcibiades some arrangement of the kind must
have been made. The ban on teaching activities, it should be noted, was
maintained formally,39 and the remainder of the endowment can have served
only to afford Simplicius and his group the leisure for literary work.
The case for the later dating is strengthened by some characteristic features of
the in Gorgiam which point to a period at the beginning of Olympiodorus’
career. It abounds in reminiscences and anecdotes about Ammonius, while the
actual substance is of the poorest. It looks as if, besides Proclus’ commentary on
the myth, Olympiodorus had only his own notes of the lectures of Ammonius to
draw upon. In the two other commentaries on Plato he shows himself better
equipped: the most significant difference is the attention paid to Damascius.
The erroneous identification of Olympiodorus with his namesake the deacon
and exegete by Anastasius Sinaita (PG 89,936 C 9–11; 1189 A12–13), together
with the questionable testimony of an alchemical treatise, have given rise to the
persistent belief that he was a Christian,40 in spite of his own unequivocal
statements to the contrary, some of which have already been brought forward by

38
op. cit. 286. On the hypothesis of a return to Athens and the meaning of eph’ heautois, cf. I. Hadot,
Chapter 12 above.
39
. . . live henceforth in undisturbed privacy.
40
Tannery, op. cit. 277; K. Praechter, ‘Richtungen und Schulen im Neuplatonismus’, in Genethliakon
Carl Robert, Berlin 1910. On Olympiodorus the alchemist, see Ol. in Phaed., introd. 22–3.
356 Aristotle Transformed

Skowronski and Norvin.41 R. Vancourt, in his study Les derniers commentateurs

Alexandrins d’Aristote (p. 24) still wrote: ‘At all events, it is certain that at the time
of Olympiodorus instruction in opposition to (Christian) dogma would not
have been tolerated at Alexandria.’ As the point affects the present question and
is of great interest for the religious situation in the reign of Justinian, I shall
quote the principal passages to show what was tolerated at Alexandria in the
time of Olympiodorus.
A most illuminating one is in Gorgiam pp. 32–3, where Olympiodorus
declares himself a monotheist, but at the same time distinguishes himself
expressly from his (Christian) audience. After explaining Hera allegorically as
air or rational soul he continues: ‘You should not, therefore, put so superficial an
interpretation on doctrines expressed in the form of a myth: we, too, know that
there is only one First Cause, God, because there cannot be a plurality of first
causes. This First Cause has not even a name, for names denote certain attributes,
and if God transcends all attributes there can be no name either to call Him by.’
In support of this negative theology Olympiodorus quotes the great hymn ‘O
thou that art beyond all things (Ô pantôn epekeina)’,42 then concludes that there
must be certain intermediate entities between the First Cause and its lowest
creatures.
The passage is supplemented by a parallel one towards the end of the same
commentary:43

You must know that the philosophers hold that there is one principle of all things
and one primary supramundane cause from which all things have sprung, and
to which they have not even given a name. [Here the same hymn is quoted.]
They say, then, that the principle of all things is one, but that it does not produce
the world here below directly; for that would be contrary to the natural order of
things, if our world were created immediately by the First Cause . . . It was
therefore necessary that other, greater powers should be created by the First
Cause and then, by these, ourselves, who are the dregs of the universe, since we
had to be, or the world would be imperfect. There are, then, other, greater powers,
which the poets call the Golden Cord because of their unbroken continuity. First

41
L. Skowronski, De auctoris Heerenii et Olympiodori Alexandrini scholis cum universis tum iis singulis
quae ad vitam Platonis spectant capita selecta, Diss. Breslau 1884. W. Norvin, Olympiodoros fra
Alexandria, Copenhagen 1915, 319.
42
Printed in the editions of Gregory of Nazianzus, but claimed for Proclus by A. Jahn, Eclogae e Proclo
de philosophia Chaldaica, Halle 1981, 62ff. Recently M. Sicherl, on the basis of an exhaustive survey
of the MS tradition, has presented a strong case for the authorship of Ps.-Dionysius, to whom the
hymn is explicitly attributed in Vat. Pal. gr. 39 (‘Ein neuplatonischer Hymnus unter den Gedichten
Gregors von Nazianz’, Gonimos, Buffalo 1988, 61–83).
43
243,16–244,17.
 The Alexandrian commentators and the introductions to their commentaries 357

there is the intellectual power, then the vivifying power and the healing power,
and so forth; when they want to express them in words, they give them symbolic
names. So do not be confused by mere words, when you hear of Kronian and
Zeusian powers, and so on, but think of the reality for which they stand, think
that we try to express something else by those words. If you like, you may take it
that these powers have no being of their own and no separate existence, but that
they are implicit in the First Cause, to which you can ascribe intellectual and
vivifying faculties. And therefore, when we say Kronos, do not be shocked by the
word, but ask yourself what I mean: Kronos is koros nous, i.e. pure intelligence.

These texts are open to one explanation only: the ‘philosophers’ are the pagan
Neoplatonists, and Olympiodorus reckons himself among them, as the use of the
first person proves. It should be stressed that the context does not permit one to
understand the ‘we’ as ‘we Christians’.
At first sight there is a certain likeness between Olympiodorus’ reasoning and
the theology of Hierocles, who also interpreted the Gods as created (in the sense
of caused) beings which we must postulate to bridge the infinite chasm between
the Supreme God and His lowest creatures. But whereas Hierocles claimed
divinity in the strict sense of the word only for the Supreme God, Olympiodorus
does not make any real concession and is only trying to make his Pantheon
acceptable for his Christian students. It is merely on their behalf, to soothe their
Christian consciences, that he suggests they regard the lower orders of deities as
functions or attributes of God. It is, in fact, a deliberate invitation to talk at cross-
purposes, professor and students each attaching a meaning of their own to the
terms used.
Olympiodorus even ventures to end with a few words in defence of the cult of
images: ‘And please do not think that the philosophers pay divine honours to
stones and idols (eidôla). It is rather like this: since we live in a world of sense-
perception and are therefore unable to attain the level of the incorporeal and
immaterial powers, idols have been invented to remind us of them in order that
by beholding and worshipping these, we may be put in mind of the incorporeal
and immaterial powers.’44 One wonders whether the student whose notes we
possess translated agalmata (statues) by the Christian term eidôla (idols) or
whether Olympiodorus did it for him.
Afterwards, Olympiodorus may have grown more cautious, but his opinions
underwent no change. On the cardinal issue of the eternity of the world, on
which Aeneas and Zacharias had attacked Ammonius and which Philoponus

44
246,7–12.
358 Aristotle Transformed

had chosen to emphasise his adherence to Christian dogma, there is no trace of

a compromise even in the commentary on the Meteorologica.45 The inevitable
complement of this doctrine is the belief in reincarnation; for otherwise, if the
world has existed from eternity, either there must be an infinite number of
surviving souls (an inference contrary to the fundamental truth that form is
finite) or the soul must be mortal, and for this very reason Aristotle is supposed
to have taken the consequence of denying individual immortality.46
Olympiodorus is no less emphatic on the point of eternal punishment, which
had been rejected by Neoplatonists ever since Plotinus: ‘But we are not punished
forever; for it is better to deny the immortality of the soul than to believe this.’
What Plato calls aiônion (eternal) is a cosmic year.47 (There is an indignant note by
the scholiast: ‘Note how this man explains away the unquestionable truth of
eternal punishment.’) The commentary on the Phaedo48 takes the same view: the
punishment for souls that have sinned consists in returning to life, but even the
souls of the theurgists, i.e. of those who have attained highest perfection, cannot
always stay in the intelligible world. The fullest discussion is found in in
Meteorologica pp. 141–50, a digression on the myth of the Phaedo. What Plato calls
Tartarus is simply genesis, either viewed from the physical side as the world of
perpetual change, or from the moral side as the place where souls are purified. For
punishment is not inflicted because of the wrath of God; when every sin has been
expiated the soul is set free. The Phaedo commentary pictures a hereafter widely
different from the Christian conception: whether we shall meet our loved ones is
doubtful, because each soul returns to the sphere to which it naturally belongs.49
Olympiodorus’ attitude towards suicide also conflicts with Christian ethics.
After citing the arguments contra and pro at great length he concludes that
suicide can sometimes be justified, when a greater spiritual good can be attained
by it, an opinion that must have been very shocking to Christian sentiment.50
A few chance phrases show that he shared with Proclus and Damascius their
contempt of the Christian establishment: ‘They [the poets of old] did not know
that there would be a corrupted society which values only appearances and does

45
115,11–13; 118,5–119,8; 120,13–14; 20–1; 153,19–21; in Gorg. 65,24–66,4.
46
in Phaedo 10, §1,2–10; 12 §2, 14–18. (The passage 9 §6 is misreported or corrupted; its object, as 10
§5,1–6 shows, was to prove that irrational soul too is reincarnated). in Gorg. 97,5–9; 97,24–98,2;
109,18–22; in Alc. 27,10–16; 213,20–3; in Meteor. 147,24–148,6.
47
in Gorg. 263,17–264,26.
48
10 §14.
49
7 §4, 1–12.
50
in Phaed. l§2–9; cf. 3 §11,7–8.
 The Alexandrian commentators and the introductions to their commentaries 359

not search at all for the hidden depth of the myth.’51 Elsewhere he says: ‘Formerly
free men were distinguished from slaves by their names and by the way they
wore their hair, slaves having such names as Geta, Davus or Phryx; but in our
days, confusion prevails in these things too.’52 Otherwise the tone in this work is
discreet and even, in a way, conciliatory. A curious example is the comment on
Alc. 111A-C, where Socrates makes mutual agreement a criterion of true
knowledge. Proclus had observed that agreement is not necessarily evidence of
truth, citing the Christians as an instance, who are agreed in denying the existence
of the Gods, but only through ignorance.53 Olympiodorus repeats this remark,
but substitutes the Democriteans and their doctrine of the vacuum for the
Christians.54 He makes no secret of the motive for this attitude. After a lengthy
discussion of traditional opinions on the subject of demonology, he concludes:55
‘And this is what the commentators have to say about genii and guardian spirits;
we, however, shall try to give an interpretation adapted to the prevailing
circumstances.’56 For Socrates, too, was condemned to the poison-cup for
propagating new deities among young people and believing in Gods that the
state did not recognise. Our view must be, then, that the guardian spirit (eilêkhôs
daimôn) is conscience, which is the purest essence of the soul and the only part
of us that is free from sin and a merciless judge and an unimpeachable witness of
what has happened here below when we appear before Minos and Rhadamanthys.’
The word daimôn (spirit, demon) was of course a dangerous one because of its
pejorative development in Jewish and Christian usage, and Olympiodorus clearly
did not want to be decried as a devil-worshipper. In the rest of the commentary
he therefore avoids the word pretty consistently; twice he actually substitutes the
word conscience where Proclus had compared Socrates to the Guiding Spirit or
Good Spirit (ephestôs, or agathos daimôn).57 Before mentioning this solution,
however, he had already drawn the more obvious parallel between guardian
spirits and guardian angels:58 ‘But since we have mentioned guardian spirits, it
should be observed that they are known also in common usage, though not by
the same name. For instead of “demon” they speak of the “angel” of each of us;

51
in Gorg. 238,16–19.
52
149,3. The phrase ‘in our days confusion prevails in these things too’, is Olympiodorus’ own. Cf. in
Gorg. 34,18. cf. Palladas, Anth. Gr. X 90,7: ‘For today everything is turned upside down’ (Paton).
53
in Alc. 294,3–6.
54
ibid. 92,6–7.
55
ibid. 22,14–23,4.
56
Damascius, too, uses ta paronta (‘the present circumstances’) in the meaning of ‘the supremacy of
Christianity’, Vita Isidori 64,1.
57
Proclus 199,9–14/Ol. 68,24. Proclus 229,20–2/Ol. 87,19.
58
21,15–18.
360 Aristotle Transformed

thus you can hear them say “by your angel”, a phrase used of those who lead the
life that is most pleasing to God.’ The last clause is not quite clear; Olympiodorus
may mean that the phrase was usually addressed to monks.59 He goes on to say
that whereas Plato speaks of demons only, the Chaldean Oracles distinguish
between angels, heroes and demons. Angels had, of course, been incorporated
into the Neoplatonic system long before.60
More interesting is the particular way in which sunêtheia (usage, custom) is
used here. If there were only this passage to go by, it could be given its normal
linguistic meaning; but it can only mean ‘popular religion’ in the preceding
paragraph, where Olympiodorus, following Proclus, discusses the peculiarities
of the daimonion (genius) of Socrates. It impressed him as a voice, Proclus
explains,61 not because the genius actually spoke, but because illumination came
to Socrates in a form that affected the acoustic organs. Olympiodorus62 adds an
observation of his own: ‘In the same way we can see it happen now in the
common religious practice (sunêtheia) that those who follow the hieratic way of
life suddenly become aware of a sweet smell and say that it is the presence of an
angel, illumination affecting, in their case, the organs of smell.’ Here there can be
no doubt that monks are meant and the addition ‘in the common religious
practice’ mainly serves to distinguish them from those who practise the native
Egyptian religion, for whom the same expression ‘who follow the hieratic way of
life (hieratikôs zôntes)’ is used.63 But this very fact proves a more positive attitude
towards the religious experiences and practices of the Christians; Olympiodorus
seems to have accepted Christianity at least as a creed for the uneducated, a
course already suggested, more than two centuries before, by his fellow-
countryman Alexander of Lycopolis.64 Two passages in which Olympiodorus
alludes to confession and charity may be mentioned for completeness’ sake.65
The fact that nothing of Ammonius’ lectures and writings on Plato survives
makes it difficult to draw a comparison between him and Olympiodorus as
regards their relations with the Church. Platonic philosophy raised more
controversial points, and the lecturer could count on a more select and more
loyal audience. Yet the impression remains that Olympiodorus enjoyed a little

59
The ninth-century scholion paraphrases: ‘May your angel be disposed in such and such a way
(towards you), if you do this.’
60
Porphyry Letter to Anebo: Iamblichus de Myst. 2,3–4.
61
in Alc. 79,16–80,18.
62
ibid. 21,9–14.
63
ibid. 18,14.
64
Contra Manichaeos 3,1–18 Brinkmann, Leipzig 1895.
65
in Gorg. 139,11–18 and 125,25–6.
 The Alexandrian commentators and the introductions to their commentaries 361

more freedom in religious matters. One reason may have been the Monophysite
schism, which partly paralysed even the political power of Constantinople in
Egypt. Then there was the failure to build up a continuous tradition in Christian
philosophy; to those classes of society which used to provide the students for the
philosophical schools the word philosopher must have meant what Olympiodorus
implies it does, a pagan thinker. Even the Christians who continued the work of
Olympiodorus did little to change this essentially pagan outlook.
In spite of the scarcity of biographical facts, it is likely for reasons to be
mentioned below, that Elias66 was the first of these. The heading of the incomplete
commentary on the Prior Analytics identifies him as philosopher and apo
eparkhôn. While the first of these two titles may safely be taken to designate a
professor of philosophy, the second is more problematic. In his article on the
subject,67 R. Guilland says of it: ‘The title apo eparkhôn – former eparch – lends
itself to confusion, when it is not qualified by the expression tês poleôs (eparch of
the city), or tôn praitôriôn (eparch of the praetorians), or less commonly tou
praitôriou (of the praetorium). For it can designate either a former eparch of the
city, or a former prefect-eparch, or an honorific title apoeparch.’ The practice of
bestowing the title on (e.g.) men of letters, comparable to the knighting of
authors, actors and musicians in Great Britain, and formerly in Germany, is at
least as old as the sixth century. Guilland cites as examples the historians Evagrius
Scholasticus, John of Epiphania and Theophylact (this last of the seventh
century), as well as Tiberius II ’s court physician Zacharias. It is much more
probable that Elias the philosopher belonged to this category than that he is
identical with Elias prefect of Illyricum to whom Justinian addressed his Novella
CLIII in Dec. 541.68 What has remained of his lectures relates to the Aristotelian
Organon only: prolegomena to philosophy and commentary on the Isagoge;
prolegomena to Aristotle and commentary on the Categories; some scholia on
the de Interpretatione;69 and the beginning of the commentary on the Prior
Analytics. Yet there is no reason to doubt his express statement that he taught
Aristotelian philosophy as a preliminary to Plato.70 A reference in the commentary
on the Isagoge proves that he also lectured on Galen’s de Sectis.71

66
cf. my paper cited above, n. 1; its biographical conclusions are open to doubt.
67
‘Études sur l’histoire administrative de l’Empire Byzantin III , L’apoéparque’, Byzantinoslavica 43,
1982, 30–44 (30).
68
ibid. 38; Westerink, op. cit. in n. 1, 127–8 (60–1).
69
In CAG 4.5 xxvi–xxviii.
70
123,9–11. Cf. 110,28–30. – At Constantinople, Platonic philosophy was taught at least as late as 511,
by Agapius, a pupil of Proclus (Lydus, de Magistr. III 26).
71
in Isag. 6,7–9.
362 Aristotle Transformed

The usual view that Elias was a pupil of Olympiodorus is merely a deduction
from the numerous parallels in the work of the two. But they are striking enough
to justify the supposition. I shall cite the more important ones in extenso.

Elias 3,17–20: Or, again as Aristotle says in Ol. in Alc. 144,15–17; And Aristotle said
the work entitled Protreptic . . . ‘If we must in the Protreptic; ‘Whether we must
philosophise, we must philosophise, and if philosophise, we must philosophise, or
we must not philosophise, we must whether we must not philosophise, we
philosophise; therefore we must must philosophise, but in any case we
philosophise in any case’ (cf. David 9, 2–5). must philosophise.’
El. 16,20–5: The poets characterise the Ol. in Gorg. 65,20–4; In brief, you must
divine in three ways, by goodness, power know that the Demiurge possesses these
and knowledge, as shown in ‘the Gods three: goodness, as witness the poets, who
givers of good gifts’, ‘but the Gods can do say: ‘the Gods givers of good gifts’,
anything’, ‘the Gods know everything’ (cf. knowledge, for which reason it is said, ‘but
Ammon. Isag. 3,9–15; David 17,2–9). the Gods know everything’, and power,
because ‘the Gods can do anything’.
El. 22,19–23: Thus a man wounded in Ol. in Gorg. 140,28–141,2; Let us then
war while protecting a friend said: ‘I have think of saving our souls, knowing that
saved myself, what does that shield possessions and the body make no
matter to me? Away with it!’ calling his difference; let us do as that man said: ‘I
soul himself and his body a shield (cf. have saved myself, what does that shield
Ps.-Elias 12,19). matter to me? Away with it!’
El. 22,25–6: And this: ‘Pound, pound the Ol. in Gorg. 185,20–2; This is what
bag of Anaxarchus, for you will never Anaxarchus said: ‘Pound, pound the bag of
pound Anaxarchus himself ’ (cf. Ps.-Elias Anaxarchus, for you will never pound
12,18). Anaxarchus himself (cf. Alc. 105,5).
El. 27,16:. . . that from childhood on we Ol. in Gorg. 237,22–3; since then, from our
love stories (cf. Ps.-Elias 18,14). earliest childhood on we have been
brought up with stories . . .
El. 31,11–13: To show that music heals Ol. in Gorg.41,7–13: Thus Pythagoras, when
the passions of the soul: they say that at he found a young man in the company of a
one time Pythagoras, when he saw a cither-playing girl, knew as a
young boy following a flute-playing girl, physiognomist that the young man was
ordered her to turn the flute around, and naturally gifted and could be helped; so, out
by spoiling the tune he put an end to his of pity, he ordered the girl to turn her flute
infatuation (cf. Ammon. Isag. 13, 24–7). or cither around and play it in that way,
thus producing a tune that immediately
cured the young man of his infatuation.
El. 39,12–14: . . . when Porphyry himself Ol. in Alc. 2,94–6: And since the
had travelled to Sicily to see the craters of philosopher should be curious to
fire of Mt. Etna, since the philosopher contemplate the works of nature, he (Plato)
should be curious to contemplate the also travelled to Sicily to see the craters of
works of nature. fire on Mt. Etna (cf. Philop. de Opif. 182,10:
the craters of fire on Mt. Etna).
 The Alexandrian commentators and the introductions to their commentaries 363

El. 109,10–14: He (Zeno) was called Ol. in Alc. 140,12–16: For such was Zeno, a
double-tongued, not because he was a great dissembler, and this was why he was
dialectician, as Zeno of Citium, and called double-tongued, not because he
argued against and for the same points, would defend both of two contrary
but because he was a dialectician in his propositions, but because he dissembled.
way of life, saying one thing and thinking Thus when a certain tyrant asked him who
another. Once, when asked by a tyrant were the ones that had plotted with Zeno
who were the most dangerous plotters against his rule, he denounced the
against his rule, he denounced the bodyguards. The tyrant killed them and
bodyguards. The tyrant, believing him, was promptly murdered.
killed them and was murdered.
El. 110,18–20: And in the Phaedo, that we Ol. in Phaed. 6 § 14: He himself says that
must go to Hades with this adamantine we must go down to Hades with an
faith in the immortality of the soul (Pl. adamantine faith in this (cf. 8 § 17,7–10,
Rep. 10 619A: with an adamantine faith Gorg. 237,26–7).
in this we must go to Hades).
El. 130,28: For philosophers are lovers of Ol. in Alc. 160,16: because of the
the universal. philosopher’s love of the universal.
(Meteor. 2,11–12; 55,9).
El. 236,6: For the a- is privative. Ol. in Alc. 171,8: For the ma- is privative.

These examples are practically all from the introductions to the Isagoge
(pp. 1–39) and to the Categories (pp. 107–134), not from the commentaries
proper. This is a common phenomenon: in the introductory part the lecturer
follows a narrow traditional scheme, in dealing with the text he is more apt to
extemporise and to arrange the subject-matter according to circumstances. Still,
if the correspondences were the result of a general familiarity with Olympiodorus’
phraseology and ideas, one would expect them also in the commentary. A more
probable explanation is therefore direct borrowing. If that is the cause, Elias
cannot have used the extant version of the prolegomena to the Categories, in
which there is nothing that answers to the extracts quoted from Elias (I have
purposely omitted all direct correspondences found in the same context).
Olympiodorus’ commentary on the Isagoge is lost, but the most natural way to
account for a word-for-word correspondence such as that between El. 39,12–14
and Ol. in Alc. 2,94–6 is that Olympiodorus expressed himself in almost identical
words in the introduction to Porphyry. This literary dependence does not
exclude a personal relation, of course; if Elias followed Olympiodorus rather
than e.g. Philoponus (whom he also knew), the most obvious reason is that he
was his pupil.
Both from his name and from his social status it appears that Elias must have
been a Christian. As a rule, however, he does not allow his Christianity to
364 Aristotle Transformed

interfere with his philosophy, though he does admit the possibility of miracles
understood as direct acts of Providence, a point on which he follows Philoponus.72
Such phrases as Kata tas Hellênôn phêmi pseudeis doxas (that is, according to the
false opinions of the Greeks),73 confidently deleted by Busse, are certainly a little
suspect because they are in flat contradiction with the importance attached in
the context to these ‘false doctrines’ and ‘lying myths’. They may therefore be
regarded as insertions by the reportator, but it is at least as likely that Elias himself
used these convenient formulas to shield himself against possible attacks. In the
crucial matter of the eternity of the world he is clearly unwilling to abandon the
old belief.74 This personal freedom of thought coupled with the outward
conformist attitude must have been common enough among the intelligentsia of
Justinian’s reign. Its most outstanding representative is of course Procopius of
Caesarea.75
David is even less known than Elias. The detailed biographical data preserved
in Armenian sources76 relate to a theologian of the fifth century, whereas the
‘most pious and devout philosopher’ whose commentary on the Isagoge we
possess, was admittedly a late Olympiodorean of the second half of the sixth
century or even the beginning of the seventh; the Armenian text77 is a translation
from the Greek, not conversely. The David of Armenian tradition is therefore
either the result of a confusion between two, or more, different persons or a
fictitious character named for the Greek commentator, whose translated works
soon became classics of Armenian literature.
It is unlikely that David lectured on Plato as well as on Aristotle. Quotations
are pretty frequent, but they are hardly at first hand. Of his own lectures he only
mentions those on the Organon and on (Aristotelian) physics.78
Closely related to David’s commentary, though not derived from it, is the text
published as Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,
Amsterdam 1967. Since the beginning is lost, the ascription to David has no

72
p. 327, n. 22 above.
73
7,3; 12,1; 69,22.
74
120,16–17; 187,6–7.
75
de Bellis V 3,5–9.
76
cf. C. Neumann, ‘Mémoire sur la vie et les ouvrages de David, philosophe arménien du Ve siècle de
notre ère’, Nouveau Journal Asiatique 3, 1829; M. Khostikian, David der Philosoph, Bern 1907. A.K.
Sanjian (ed.), David Anhaghtc, the ‘Invincible’ Philosopher, Atlanta, GA , 1986 (9 papers and a full
bibliography).
77
David the Invincible, The Definitions of Philosophy, Armenian text and Russian translation by S.S.
Arefshatian, Erevan 1960 (= introductory part of David in Isagogen); Commentary on Aristotle’s
Analytics, Armenian text and Russian translation, by S.S. Arefshatian, Erevan 1967; Commentary on
Porphyry’s Isagoge, id., Erevan 1976.
78
See Busse’s index nominum s.v. ‘Dabid’.
 The Alexandrian commentators and the introductions to their commentaries 365

authority; long extracts under the name of Elias may come from a complete
copy, but the work is altogether too different from the genuine Elias to pass for
an alternative version. The lecturer shows himself familiar with medical literature,
and he may have taught medicine also, or even in the first place. As far as can be
inferred from a few incidental utterances, the religious views of David and Ps.-
Elias are much the same as those of Elias. Without strictly committing themselves
one way or the other, they refer to the doctrines of the eternity of matter79 and of
the divinity of the celestial bodies,80 to the irrational avenging spirits81 and the
long-living nymphs of Neoplatonic demonology.82 The remark that God first
made the elements, then man,83 sounds like a citation from the Bible,84 but it may
also be a free adaptation of the Timaeus. On the whole, their main concern seems
to be to keep philosophy and dogma as far apart as possible.
It should be repeated, however, that it is difficult to judge of these matters
merely from a logical treatise, in which fundamental questions are rarely touched
upon. In the case of Stephanus we are in a better position, because we have his
commentary on the third book of the de Anima,85 in which a number of crucial
points could not possibly be avoided. His way of dealing with them is confused.
On the one hand he accepts unquestioningly the authority of Christian dogma86
and of the Bible.87 On the other hand there is no attempt at a wholesale revision
of the traditional material from the Christian point of view. All the old tenets
continue to reappear: the eternity of the world (‘according to Aristotle’)88 and of
the fifth substance (‘according to some’),89 the pre-existence of the human soul90
and the rationality of the heavenly bodies91 (both without any attempt at
refutation or reconciliation). We miss the clearness and consistency with which

79
Ps.-Elias 42,20.
80
ibid. 34,27: ‘This question about the sky, whether it is animate or inanimate, is controversial to the
present day’. Curiously enough, the sentence returns almost word for word in Theophilus, in
Hippocr. Aphor. 247,35–248,2 Dietz (ed.), Scholia in Hippocr. et Gal. vol. 2, Königsberg 1834): ‘For
even now philosophers are engaged in controversy regarding the sky, whether it is a body or
incorporeal, and whether limited or unlimited.’ The link may be Stephanus, although the sentence
is not found in the (slightly abridged) text of this author (CMG XI 40,6–9).
81
Ps.-Elias 41,29–31.
82
CAG 18.2 15,21–6; 24,11–12; 14–15. cf. Ol. in Meteor. 301,18–19.
83
CAG 129,8–10.
84
Some biblical phrases: Ps.-Elias 15,4, ‘King of Kings’ (in Tim. 6, 19; also in Elias 21,4); 13,26, ‘is
crowned by the Creator with the prize of victory.’ (I Cor. 9:24)
85
Published in the third book of Philoponus’ in de Anima, CAG 15, 446–607. See Hayduck’s preface
and R. Vancourt, Les derniers commentateurs Alexandrins d’Aristote, Lille 1941, 43–8.
86
527,29–32.
87
547,11–14.
88
540,27.
89
448,6–7.
90
541,20–542,5.
91
595,33–598,7. Cf. above, n. 80, and Philop. de Opif. VI ,2.
366 Aristotle Transformed

Philoponus handles controversies between philosophy and dogma in his work

against Proclus and, in the exegetical sphere, in the de Opificio Mundi. Philoponus,
however, had the advantage of being an outsider, who could in turn play the part
of an opponent or of an impartial judge, while Stephanus was handicapped by
being the official representative of classical philosophy, long before the purely
historical point of view had developed.
That he had this function already at Alexandria before he was called to
Constantinople is not certain, but probable. Soon after the accession of Heraclius
(in 610) he was nominated oikoumenikos didaskalos, i.e. professor at the newly
founded or restored Imperial academy.92 According to the author of an astrological
forgery written under his name about 170 years later, he taught Platonic and
Aristotelian philosophy, the quadrivium, alchemy and astrology.93 As far as we can
check him, the falsifier was not badly informed; commentaries on the de
Interpretatione and de Anima Book 3 are extant, a course in arithmetic is mentioned
by Stephanus himself,94 an astronomical work has been published (partly) by
Usener,95 and treatises on alchemy also circulated, rightly or wrongly, under his
name. This, together with a note of not much more authority in an alchemical
opusculum,96 is the only formal confirmation of the presumption (in itself not an
unlikely one) that Platonic philosophy was taught as late as the seventh century.97

92
H. Usener, De Stephano Alexandrino, Bonn 1879, 3–5. On the life and works of Stephanus in general
see Vancourt, op. cit. 26–42; Lemerle, Le premier humanisme byzantin, Paris 1971, 79–81. The
question of his identity with the professor of medicine Stephanus of Athens needs further
investigation. His commentaries on Hippocrates and Galen are to be found partly in Dietz, Scholia
in Hippocratem et Galenum, Königsberg 1834; new editions are in the course of publication in the
Corpus Medicorum Graecorum: Prognastic, J.M. Duffy (ed.), Berlin 1983; Aphorisms vol. 1, L.G.
Westerink (ed.), 1985, vols. 2 and 3 in preparation; Ad Glauconem, K.M. Dickson (ed.), in
preparation. The author of the medical texts taught at Alexandria (I 335,5 Dietz); he is called
philosophos (I 51; 233; II 237) and shows himself more familiar with current philosophical ideas
than most of his colleagues; his lectures are edited on the pattern usual in the school of Ol. The fact
that in the title of the Aphorisms commentary he is called Stephanus of Athens is no obstacle: cf.
H. Rabe, Prolegomenon sylloge, Leipzig 1931, lviii n. 1: ‘How many philologists were classified as
“Alexandrian” by Suidas! No one believes they were all born in Alexandria.’ On the other hand, the
philosophical and astronomical information is often superficial or confused (though, of course, it is
always possible to hold the notetaker, a medical student, responsible for this); further, if the author
knew the iatrosophist Gessius personally, as the anecdote at Aphor. I 256,3–8 suggests, he cannot
possibly have survived until the reign of Heraclius. The early sixth-century Latin commentary on
the Aphorisms which P. Courcelle (Les lettres grecques en occident, Paris 1943, 387–8) attributes to
Stephanus, cannot serve as proof, because the grounds for the attribution are too weak.
93
H. Usener, Stephani Alexandrini opusculum apotelesmaticum, Bonn 1879, 17.
94
in de Anima 457, 24–5.
95
De Stephano Alexandrino commentatio altera, Bonn 1880.
96
Vindob. Med. Gr. 14, f. 53r.
97
The anonymous sequel to Proclus’ commentary on the Parmenides (Procli. Opera, Cousin (ed.),
1864, 1237–1312), is the work of George Pachymeres (thirteenth century), and can therefore be left
out of consideration. [For a different chronology, which puts pseudo-Elias and David back in
Alexandria after Stephanus’ move to Constantinople, see Mossman Roueché’s suggestion recorded
in Chapter 1 – Ed.]
 The Alexandrian commentators and the introductions to their commentaries 367

2. Prolegomena to Aristotle and Porphyry

Most of the philosophers discussed above are represented in the Commentaria in

Aristotelem Graeca by an introduction to philosophy in general and to Porphyry’s
Isagoge or by one to Aristotle and the Categories, or by both. These parallel texts
offer valuable material for comparison.98 Besides some individual features a survey
of them will bring out, above all, the common characteristic of extreme conservatism.
The family likeness is so strong that a comparison is most easily given in synoptic
form. This is more especially true of the introductions to Aristotle.
The five commentaries on the Categories, of which they form the introductory
part, are the following: Ammonius (CAG 4.4). Philoponus (13.1), Olympiodorus
(12.1), Elias (18.1), and Simplicius (8). With the sole exception of Simplicius,
they are all in the form of lecture notes. The contents are these:

I. Introduction to Aristotle in ten sections

The order of the sections varies slightly: Amm. 1–6,8–10,7; Ol. and El.1–
10; Philop. 1–5, 8–9, 6–7, 10; Simpl. 1–5, 8–9, 7, 6, 10.

(1) Origin of the names of the various philosophical schools. They may be derived
from: (a) the founder’s name; (b) his native city or country; (c) the locality where
he taught; (d) a particular style of life (Cynics); (e) its belief (Sceptics: Plato’s
arguments against them); (f) accidental circumstances (Peripatos); (g) its view as
to the aim to philosophy (Hedonists).

Arrangement as above in Amm. and Ol.; Philop. and El. have (a)(b)(c)(e)
(d)(g)(f), Simpl. (a)(b)(c)(f)(e)(g)(d). Both Ol. and El. add a diaeresis to
show that this list exhausts the possibilities. Simpl. is very brief, omitting
the digressions, El. is more explicit than even Ol.

(2) Classification of Aristotle’s writings: (i) particular (merika), e.g. Letters;

(ii) intermediate, e.g. the Constitutions; (iii) universal (katholou), subdivided into
(a) notebooks (hupomnêmatika) (notes for personal use, either specialised
(monoeidê) or various (poikila)) and (b) treatises (suntagmatika) consisting of
(a) dialogues (exôterika) and (b) non-dialogical works (akroamatika). The last
group includes (1) theoretical philosophy (true or false): theology, mathematics,

98
The whole subject was treated as long ago as 1949 in a wider context (grammar, rhetoric and
philosophy) by Marian Plezia, De commentariis isagogicis, Kraków 1949 (Archiwum Filologiczne
23). Unfortunately this book did not come to my attention until quite recently.
368 Aristotle Transformed

physics; (2) practical philosophy (good or evil): ethics, economics, politics; (3)
the Organon: its elements (Cat., Int., An. Pr.), the method itself (An. Post.),
accessories (Top., Soph. El.).

Amm. omits the subdivision of (iii)(a). Ol. and El. discuss the difference
between (iii)(b)(a) and (b) and Alexander’s opinion on it. Only Elias gives
a further classification of the physical works. In (iii)(b)(b)(3) Ol.
distinguishes the method, contributing additions, clarifying additions, El.
and Simpl. preliminaries, the method itself, substitutes for the method.

(3) The starting point: courses should begin with logic, ethics coming first only
in the sense of practical morals.

Philop. and especially El. give a fuller doxography; Boethus (physics),

Andronicus (logic), Platonists (ethics or mathematics), Amm. and Ol. also
outline the rest of the course, as in section 5.

(4) The final goal: knowledge of God, the First Principle.

Ol., El. and Simpl. cite Hom. B204 (Arist. Metaph. 12.10 end).

(5) The way to the end: ethics, physics, mathematics, theology.

(6) Qualifications for the student.

Wording varies very much. Ol. and El. cite Pl., Alc. 114E. El. and Simpl.
Arist. Top. 1.11.

(7) Qualifications for the exegete: critical attitude required.

El. and Simpl. mention the necessity of familiarity with the whole corpus
Aristotelicum as well as with Platonic philosophy, in order to show their
essential agreement.

(8) Aristotle’s style: varies with the kind of writing.

Ol. and El. both characterise the epistolary style as koinos hama kai idios,
destined for the general reader and yet personal (Hermogenes). Simplicius
deals with the nature of Aristotle’s works in general.

(9) The purpose of his obscurity: to exclude the unworthy (comparable to the
function of curtains in temples).

Elaborate series of analogies in Ol. and especially El.

(10) Preliminaries to each separate work: (a) subject (b) usefulness (c) order of
treatment (d) title (e) authenticity (f) disposition
 The Alexandrian commentators and the introductions to their commentaries 369

Philop. lists these points in the order (a)(b)(d)(c)(f)(e), next (in Part II )
discusses them in the order (a)(b)(d)(c)(e)(f); in Ol. this is (a)(b)(c)(d)(e)
(f), resp. (a)(b)(d)(e)(c)(f); in El. (a)(b)(d)(e)(c)(f), resp. (a)(b)(c)(d)(e)(f);
in Simpl. (a)(b)(d)(c)(e)(f), resp. (a)(b)(c)(d)(e)(f). Philop., Ol., El. and
Simpl. dwell on the question of authenticity in general, listing the possible
motives for forgery in slightly different ways.
Between I and II Ol. inserts, as an introduction to logic, a treatise on the
question whether logic is a part of philosophy or merely its instrument.
The subject is usually reserved for the introduction to the Prior Analytics.

II. Introduction to the Categories

(a) The subject: neither ‘words’ nor ‘things’ nor ‘thoughts’, but ‘words denoting things
by way of thoughts’ (Iamblichus), with the further restriction ‘simple’ and ‘as assigned
in the first instance’ (in contradistinction from propositions and syllogisms).

Doxographical details vary: Philop. ascribes the rejected definitions

respectively to Alexander, Eustathius, Porphyry; Ol. to Porphyry, Herminus,
Alexander; El. (combining Ol. and Philop.) to Alexander and Eustathius.
Herminus, Porphyry. Only Simpl., who has consulted the sources, gives
reliable details. Philop. adds a digression on the principle ‘the beginning of
action is the end of consideration’, which reappears under (b) in Simpl. and
under (c) in Ol.

(b) Usefulness: way to proof.

(c) Order of treatment; beginning of logic.
(d) Title.

Amm. and Philop. explain only the word Categories, Ol. discusses two
alternative titles, El. and Simpl. several more.

(e) Authenticity: the old libraries possessed two versions of the Cat., of which
one has stood the test of criticism. Criteria.

Amm., Philop. and Ol. compare the 40 (El. 42) books of the Analytics, of
which four are left. Only Ol. mentions specific arguments against the
authenticity of the Cat.

(f) Disposition: (i) preliminary notions, ch. 1; (ii) categories, chs. 2–9: (iii)
explanation of some terms, chs. 10–15. Part i deals with entirely new notions,
part iii gives a technical meaning to familiar words.
370 Aristotle Transformed

In the introductions to Porphyry, and especially in the general prolegomena to

philosophy that precede them, there is considerably more variation. As time goes
on, they tend to become more prolix and more schematic. Ammonius (CAG 4.3),
who may have introduced the Isagoge at Alexandria, started on a moderate scale
with a 24 page treatise on the definition and subdivision of philosophy.
Olympiodorus’ introduction is not extant, but that of Elias (CAG 18.1) ought to
give a reasonably correct idea of it. Elias devotes 15 lectures (39 pages) to this
introductory matter; two and a half lectures are needed to discuss definition in
general, and five more are taken up by lengthy commentaries on the six definitions
of philosophy chosen by Ammonius. In David and Ps.-Elias, the length of the
prolegomena is again nearly doubled: David’s version (CAG 18.2) consists of 28
lectures (94 pages), its counterpart in Ps.-Elias of 27 lectures, of which 7 are lost.
Because of these differences it is more convenient to analyse at least
Ammonius’ prolegomena separately.

Ammonius’ prolegomena

I. Introduction to philosophy
(A) Definition of philosophy. – What is a definition? – Any art can be defined
either from its matter or from its aim. Out of the numerous definitions of
philosophy Amm. chooses ‘five or a few more’.

(1) ‘Knowledge of reality as such’;

(2) ‘Knowledge of things human and divine’;
(3) ‘Assimilation to God as far as attainable for man, both in theory and in
practice; the restriction ‘as far as attainable’, because God and man are
essentially different;
(4) ‘Training for death’ (but not suicide: Phaedo 62B); ‘natural’ and ‘voluntary’
death;
(5) ‘Art of arts and science of sciences’ (Arist. Metaph. 1.2); difference between
art and science; all derive their principles from philosophy;
(6) ‘Love of wisdom’ (Pythagoras); loose usage of sophos in older times
(Archilochus, Hom. Il. 23.712; 15.412), limited to God by Pythagoras.

(B) Subdivision of philosophy. – What is division, alternative division, subdivision?

(I) Theoretical philosophy: (1) theology (2) mathematics (3) physics (objects:
the immaterial, the intermediate, the material; mathematics as a transitional
 The Alexandrian commentators and the introductions to their commentaries 371

phase). Mathematics is subdivided into (a) geometry (b) astronomy (c) music
(d) arithmetic (ethical influence of music), i.e. continuous quantity, immovable
(a) and eternally moving (b); discrete quantity, related (c) and absolute (d).
(II ) Practical philosophy: (1) ethics (2) economics (3) politics. Amm. rejects
the view that the three are identical. Each can be subdivided into (a) legislation
(b) jurisdiction.

II. Introduction to the Isagoge

The notion of a category; development of the category of essence and, very
summarily, of the others. – The ‘five words’ are used by Arist. in Cat., Porphyry
explains them. Subject and title of the book. The usual prolegomena or technical
preliminaries:

(1) The subject of the book: what is genus, species, differentia, property,
accident?
(2) Its usefulness.
(3) Authenticity.
(4) Order of treatment: before Cat.
(5) Title.
(6) Disposition: preface, then (i) the five words separately, (ii) in pairs, (iii)
together.
(7) Section of philosophy to which it belongs (logic).

Busse observes that the ‘usual prolegomena’ partly repeat what has already
been said and thinks that they were interpolated in accordance with the later
practice.99 It is true that the text of Amm. has been treated with scant respect by
scholars and scribes and the repetition is undeniable. On the other hand, there is
no question of a ‘later practice’, for the same points are discussed in Ammonius’
introductions to the Cat. and the Prior Analytics, and the fact that Boethius, too,
deals with them in his commentary on Porphyry100 is independent confirmation
that Amm. himself must have done so already. Whether the passage in our text
was added as an afterthought by Amm., or by some one else, makes no difference.
The most elaborate form of these prolegomena, David’s version, can serve as
a basis for a synopsis of the remaining three.

99
Note on p. 21,5 of his edition, and preface p. v n. 4.
100
4,17–5,10 Brandt. Another passage deleted by Busse (20,25–21,4) is proved genuine by Boethius
10,19–22.
372 Aristotle Transformed

David’s prolegomena

I. Introduction to philosophy
Planned on Arist. An. Post. 2.1: (i) if it is, (ii) what it is, (iii) of what kind it is, (iv)
why it is.

Dav. only.

(i) Does philosophy exist? Four arguments of sceptics are answered: (1) to on (being)
is homonymous, therefore undefinable, therefore unknowable; (2) all things are in
perpetual flux; (3) all knowledge derives from the senses; (4) philosophy is general
knowledge, which cannot exist in a particular subject. Positive proof of its existence
is derived from the belief in providence. The argumentation of Pl. Theaet.
170C–171C, and Arist. Protr. fr. 51.

El. disposes of the whole point in a few lines, with only the last argument.

(ii) What is philosophy? As a unity, it is known through definition (A), as a

complex, through subdivision (B).

(A) Definition of philosophy. Nine preliminary questions:

(1) What is definition?

(2) In what way does it differ from description, horos (term, definition), and
descriptive definition?
(3) Origin of the word.
(4) From what are the definitions of philosophy derived? From its matter
(= reality) or its aim (= becoming like God).
(5) Perfect and deficient definitions.
(6) The six definitions of philosophy (as given by Amm.).
(7) Why are these the only possible ones? Proof by diaeresis and from
arithmetic (six is the first perfect number).
(8) (a) The order of the six definitions. – (b) Two additional ones are proved
superfluous.
(9) Their origin: 1,2 and 6 come from Pythagoras, 3 and 4 from Plato, 5 from
Aristotle.

El. deals with much the same matter, though a little less schematically, in
the order (l)–(4), (8a), (6),(9),(7),(8b) – Ps.-El. begins at (7), to which he
adds a long treatise on the qualities of the numbers from one to ten; he
leaves out (8b).
 The Alexandrian commentators and the introductions to their commentaries 373

The definitions explained separately:

(1) ‘Knowledge of reality as such’.

(2) ‘Knowledge of things human and divine’.
(3) ‘Training for death’; four arguments against suicide. ‘Natural’ and
‘voluntary’ life and death. The five possible reasons for suicide according
to the Stoa.

El. gives essentially the same matter in a somewhat different disposition.

(4) ‘Assimilation to God as far as attainable for man’. Four meanings of

homoios. Essential difference between God and man. The virtues of Theaet.
176A–B.

El. does not discuss the notion of homoios.

(5) ‘Art of arts and science of sciences’; in the same way a king is called ruler of
rulers and God King of kings. The other arts derive their principles from
philosophy and are corrected by it. The definition does not imply that
philosophy is an art. Art and science defined; the difference between the two.

The digression on art and science not in El.

(6) ‘Love of wisdom’; the old general sense of wisdom (Hom. Il. 23,712)
restricted by Pythagoras. The degrees of knowledge: sense-perception,
imagination, opinion (irrational and rational), discursive thought,
intellection. – The four groups of definitions (1–2 from matter, 3–4 from
aim, 5 from rank, 6 from etymology); significance of the number four.

Dav. adds the treatise on the properties of the numbers from one to ten
which precedes the discussion of the separate definitions in Ps.-El. El.
leaves out the digression on degrees of knowledge but adds one on perfect
numbers.

(B) Subdivision of philosophy. – What is division, alternative division,

subdivision? (I) Theoretical philosophy, (II ) practical philosophy. Grounds for
this division. While (I) has a wider range, (II ) has a wider aim (the Good).

(I) Plato: (1) physics (2) theology (with mathematics as propaedeutics);

Aristotle: (1) physics (2) mathematics (3) theology (1 material, 2 immaterial,
3 material in substance, immaterial in idea). Function of mathematics: Epinomis
992A; Plot. 1.3.3; meaning of the name. – Mathematics is subdivided into:
(a) arithmetic (b) music (c) geometry (d) astronomy (in the order named), i.e.
374 Aristotle Transformed

discrete quantity, absolute (a) and related (b), continuous quantity, immovable
(c) or moving (d). Invented by Phoenicians (a), Thracians (b), Egyptians
(c), Chaldeans (d). Practical counterparts of these theoretical sciences. Ethical
influence of music.

Ps.-El. omits the paragraphs on music, but adds one on the individual
‘inventors’ of the quadrivium.

A long disquisition follows on the modes of divisions: what is the principle of

division in the case of philosophy, of theoretical philosophy, of mathematics?

Only in Dav. and Ps.-El.

(II ) Aristotle subdivides practical philosophy into (1) ethics, (2) economics
(3) politics (his books on these subjects). The Platonists object that politics
includes the other two, and divide into (1) legislation and (2) jurisdiction, for
which the other three provide the material conditions.

Both El. and Ps.-El. try to prove that Pl. and Arist. really agree, but in entirely
different ways.

(iii) The nature of philosophy: on the strength of the definitions it can be described
as contemplative, directive, purificatory, social.

(iv) The aim of philosophy: Tim. 47B. Both the cognitive and the vital faculties of
the soul are guided by it.

Headings (iii) and (iv) are absent from El.

II. Introduction to the Isagoge

(1) The subject: genus, species, differentia, property, accident. Diaeresis
showing that all words can be reduced to these.
(2) Usefulness: the phônai contain the elements of (a) Aristotle’s Categories, (b)
all philosophy, (c) the dialectical methods (division, definition,
demonstration, analysis).
(3) Title.
(4) Authenticity.
(5) Order of treatment: before Cat.
(6) Disposition; (i) the five words separately, (ii) correspondences and
differences.

Dav. and Ps.-El. reject the division into three parts as given by Amm. and El.
 The Alexandrian commentators and the introductions to their commentaries 375

(7) The method of instruction used (as under (2c)).

(8) Section of philosophy to which it belongs.

The order of these points is, in El., (1)(2)(5)(3)(4)(6)–(8); in Ps.-El. (1)(2)

(5)(3)(4)(6)(8)(7) with a final paragraph on Porphyry’s style.

In spite of successive additions, the main scheme has remained the same for a
century; but not only the outline is traditional, a great many details, arguments,
illustrative citations and anecdotes, too, have been passed on from generation to
generation. For example the same lines from Homer are cited in proof of the
omnipotence and omniscience of the Gods in Amm. (3,9–15), El. (16,20–5, and
Dav. (17,2–9); another line (Il. 5,442) serves to emphasise the difference between
Gods and men (Amm. 3,22–3, El. 17,7–8, Dav. 34,21–2, Ps.-El. 14,27.) In the
discussion of voluntary death the epigram on Cleombrotus’ suicide is obligatory
(Amm. 4,22–5, El.14,4–7, Dav. 32,30–3, Ps.-El. 12,5). The same rare variant of
Homer Il. 23,712 illustrates the use of sophos before Pythagoras (Amm. 9,8–15, El.
23,34–24,3, Dav. 45,7–9, Ps.-El. 17,5). Plotinus 1.3,3, on the function of mathematics
in education, is referred to by Amm. 12,25–7, Dav. 59,17–19, Ps.-El. 18,29. The
story of Pythagoras converting a young man by music is told by Amm. 13,24–7, El.
31,11–13 (cf. Ol. in Gorg. 41,7–13). Philosophy is characterised as theia kai ephetê
with reference to Tim. 47B by Amm. 16,17–19, El. 1,19–2,2, Dav. 78,28–79,1. In
Amm. 15,17–16,3, El. 34,10–21, Ps.-El. 22,20–1, the Carmen aureum is quoted for
examples of the dikastikon (jurisdiction) in individual morals. This list could be
made considerably longer, and a similar one can be drawn up for the introductions
to Aristotle. In each of them only a fraction of the contents is the lecturer’s own
contribution, and the same may safely be inferred with regard to the prolegomena
to Plato, lost or extant.
376
 15

Boethius’ commentaries on Aristotle

James Shiel

It is recognised that Aristotelian logic, which was to become an integral part of

medieval scholasticism, was first transmitted to Latin Christianity through the
work of Boethius. But the way in which he set about his self-imposed task has
never been determined in detail.1 We know that he promised to translate and
comment upon ‘every single work of Aristotle I can lay hands on (omne Aristotelis
opus quodcumque in manus venerit)’.2 ‘To form the idea was a silent judgment on
the learning of his day; to realise it was more than one man could accomplish;
but Boethius accomplished much. He translated the Isagoge (Introduction to
logic) of Porphyry and the whole of Aristotle’s Organon. He wrote a double
commentary on the Isagoge and commentaries on the Categories and de
Interpretatione of Aristotle and on the Topica of Cicero. He also composed
original3 treatises on the categorical and hypothetical syllogism, on Division and
on Topical Differences.’4
The genuine texts of Boethius’ versions of Aristotle (except that of the Posterior
Analytics) have now been identified among the manuscripts and his distinctive
method of translation firmly identified.5 The present article therefore proposes

1
M. Cappuyns, ‘Boèce’, in Dict. d’hist. et géog. ecclés. 9, Paris 1939, 367: ‘The exact role of Boethius in
the transmission of Aristotle’s works is hard to disentangle at present.’ This statement prompted the
present enquiry. Dom Cappuyns’ article is the best introduction to the subject. [Now however see
the prefaces of Aristoteles Latinus, vols 1–6, and the supporting essays in L. Minio-Paluello,
Opuscula: the Latin Aristotle, Amsterdam 1972.]
2
in Int. II 79,16 Meiser.
3
An adjective that is here disputed.
4
H.F. Stewart and E.K. Rand, Boethius: Tractates and Consolation, Loeb ed. 1926, ix.
5
L. Minio-Paluello, ‘The genuine text of Boethius’ translation of Aristotle’s Categories’, in Mediaeval
and Renaissance Studies (MRS), 1942, 151–77 (=Opuscula, 1–27) and ‘The text of the Categoriae:
the Latin tradition’ in Class. Quart. 39, 1945, 63–74 (= Opuscula, 28–39).

377
378 Aristotle Transformed

to examine the other extant results of Boethius’ promise, the commentaries and
treatises. Are they really original or are they too translated from Greek?

1. In Categorias6
This work invites our first probings, for here we have an explicit indication of
source from Boethius himself (160a):‘This comment is meant for the introductory
stage, the stage of simple explanation. In composing it I have followed Porphyry
because he seemed to be easier and clearer than others (haec est tempori
introductionis et simplicis expositionis apta sententia quam nos nunc Porphyrium
sequentes quod videbatur expeditior esse planiorque digessimus)’. And in fact
detailed comparison shows a marked similarity between Boethius’ work and the
extant short commentary on the Categories by Porphyry by question and answer
(kata peusin kai apokrisin).7 We immediately notice the very same kind of phrase,
detail, or concrete example being used by both authors: ‘. . .as in geometry the
terms are explained first. . .’; ‘. . .there emerge then four relations between these
terms’; ‘. . .for example, “I embrace – I am embraced” ’; ‘. . .if a bushel is full of
grain. . .’; ‘. . .Herminus says that “above” and “below” represent not place but
position’; ‘. . .this kind of definition is considered Platonic’. And so on.
Practically every section of Boethius’ commentary exhibits this sort of
dependence on the corresponding part of the Kata peusin. Often the very
questions of Porphyry’s catechism are reproduced:

If the treatise is about words that signify objects why does he here discuss objects
rather than words?. . . Why if he here divides speech into ten categories does he
in On Interpretation divide it into just two parts?. . . Why does he here treat the
category of quantity before that of quality?. . . Why does he a little further on
describe virtue and vice as qualities?

The final part of Porphyry’s commentary (on Cat. 11b17–15b32, a sort of

appendix, the so-called postpraedicamenta) is missing from the single extant
Greek manuscript. But numerous parallels in other Greek commentaries show
that the corresponding part of Boethius, Book 4, is still following some Greek
work. There are dozens of parallel expressions in the Greek of Simplicius: ‘This
question has been raised by many, including Andronicus’; ‘. . .the single mean

6
PL 64, 159–294.
7
Porphyrii in Aristotelis Categorias Expositio per interrogationem et responsum, A. Busse (ed.), CAG
4.1, Berlin 1887. [The Latin and Greek parallel passages are set out with their references in the
earlier printings of this paper.]
 Boethius’ commentaries on Aristotle 379

between heat and cold’; ‘. . .for example, puppies and children’; ‘. . .Aristotle here
uses induction to strengthen his case’; ‘We call Pythagoras “older” with reference
to Socrates’. And in Ammonius: ‘This will be explained in the book On
Interpretation.’ And in Philoponus: ‘Aristotle here urges us on to deeper inquiry.’
And since Busse makes it clear that Porphyry did go on to comment on the
postpraedicamenta8 it is reasonable to assume that Boethius’ fourth book still
follows him. Some details support this assumption. At 264c Boethius mentions
the disparate views of Stoics and Peripatetics. Such discussion of Stoic views is a
common thing with Porphyry, as Simplicius remarks.9 At 264d ‘those of better
judgement’ prove to be the same as ‘certain of the Peripatetics’ mentioned by
Simplicius at 381,2, where he is certainly following Porphyry because a little later
(381,20) he presents Iamblichus with a different view. At 283c Boethius says,
‘these matters come under the heading of quality’ and Simplicius says (414,34)
that this point was discussed by ‘the followers of Porphyry’. As the main pillars of
Simplicius’ commentary are Iamblichus and Porphyry, and Boethius’ fourth
book offers no parallels for the Iamblichus passages, it appears that for this as for
the previous three books his main dependence is on Porphyry.
Thus far we have merely been independently testing and corroborating a
suggestion already made by Bidez – that Boethius’ commentary was taken almost
completely from the Kata peusin.10 But Bidez did not notice that there are
important differences between Boethius and Porphyry. First, Boethius omits
many of Porphyry’s explanations, even some useful ones: Porph. 56,14:
explanation of the word katêgoriai; 59,10: reference to Athenodorus, Cornutus,
Herminus, Boethus; 64,30: explanation of logos; 78,35: explanation of ‘chiasmic
arrangement’, and others.11 Had these been in his Greek source he would surely
have included some of them. Again, Boethius’ topics are not always in the same
order as Porphyry’s, or even as Aristotle’s.12 Also, many of his explanations are
simpler than Porphyry’s: some of them add hardly anything to Aristotle’s own
words. The comment on dispositiones at 216d, a long explanation, not found in
Porphyry, adds little or nothing to Cat. 6a32ff. And at 228b, the ‘father-son’
exemplification is from Porph. 118,8, but Boethius enlarges it with an extra note
about the ‘double-half ’ of Cat., 6b1.

8
cf. Porph. in Cat. Busse, li, n. 2.
9
Simpl. in Cat. 2,8 Kalbfleisch.
10
J. Bidez, ‘Boèce et Porphyre’, Revue belge de philol. et d’hist. 2, 1923, 189ff. This close adherence to
Bidez seemed important to me before the appearance of the Aristoteles Latinus editions of Boethius.
Had these editions been available I could have chosen rather to emphasise from the start the
disparity of word-usage which rules out any idea of direct translation from Porphyry.
11
Porph. 81,7; 81,16; 82,23; 98,34; 101,33; 132,4.
12
Contrast Boeth. 184a, 190a, 192c with Porph. 88,14; 93,31; 94,20.
380 Aristotle Transformed

Even where dependence on Porphyry is apparent, the verbal similarities are

seldom exact. For instance, in the passage quoted above (p. 350) where Porphyry
has ‘four relations’ Boethius has ‘four differences’. And there is a telling difference
between the two explanations of the Aristotelian phrase ‘of a subject’ (Cat. 1b10):
Porphyry says it means ‘univocally’; Boethius says it means ‘in essence’, ‘in very
substance’. In such cases Boethius’ Greek is certainly not Porphyry’s.
As a matter of fact, Bidez13 seems to have overlooked at least seventy passages
or phrases from all points of Boethius’ work where non-Porphyrian Greek
parallels come to hand. Here are a few: ‘. . .the Categories comes first in the
order of school reading’; ‘. . .the work is also useful for the study of
moral philosophy’; ‘. . .without it Aristotle’s work would look truncated and
unfinished’; ‘. . .there is another work by Aristotle about categories, much the
same in content but different in style’; ‘Archytas too wrote a work in two books
called On Universals. He was a Pythagorean from Tarentum and lived for a
while with Plato’; ‘As I said before, this work is not about the genera of things,
nor about the things themselves, but about words’; ‘Aristotle mentions only three
of these descriptions and two of them he unites into one’; ‘. . .where there is
a single concept and a single word to express it the expression is called by
him “incomposite” ’; ‘. . .as milk, for example, hardens into cheese’. For all those
expressions, and a host of others, there are Greek parallels in Simplicius. ‘A strong
indication that Aristotle is here thinking of words and not of things is given in
his phrase, “among forms of speech” ’ (Cat. 1a16): that comment is matched
in Olympiodorus. ‘ “Skindapsus” is a meaningless utterance’ is found also in Elias
(skindapsos).
Even where we find no extant parallel ad locum we can still at times sense the
presence of the Greek source. Even the more casual and personal phrases of
Boethius and his very manner of introducing Porphyry have a Greek touch:
‘But someone may say. . .’: this is a common opening in Greek Aristotelian
texts; ‘. . .the prolixity of the discussion must be curtailed’: Iamblichus in Greek
similarly curtails the prolixity of Porphyry’s comments; ‘many thinkers, including
Andronicus, say. . . but Porphyry’s view is’: this antithesis is found in Greek; ‘I
follow Porphyry because his comment is simpler’: so does Simplicius; ‘As I have
already said (ut docuimus)’: this looks at first like a personal statement by
Boethius, but it turns out to be merely the Greek scholiast’s eirêtai kai proteron.

13
Bidez, op. cit., 194: ‘Apart from the references to Themistius and to Iamblichus (224c) cited above, I
have discovered nothing in Boethius’ commentary which is not a reproduction of, or which could
not be considered as the development of, the parallel passage of the Kata peusin’.
 Boethius’ commentaries on Aristotle 381

All this indicates that Boethius was not simply following an exact or complete
or first-hand copy of the Kata peusin.14 But his Greek document obviously
reproduced a good deal of it, though not verbatim.
To proceed let us briefly compare his work with all the other known
commentaries on the Categories.
(1) Pre-Porphyrian commentaries (Herminus, Alexander, Andronicus, the
Stoics). Boethius knew these through the medium of Porphyry, for all his
quotations from them are found in the Kata peusin.
(2) Porphyry (died 309 ad ) Pros Gedaleion. It seems that Boethius cannot
have followed this larger and more profound commentary on the Categories,
now lost. For in Boethius’ first three books we find hardly any parallels for the
numerous fragments of the Pros Gedaleion preserved by Simplicius, and usually
the elementary Kata peusin accounts for all the subject-matter. And in Boethius’
fourth book several whole sections are merely a spun-out repetition of Aristotle’s
text with no external material:15 in these sections the Greek which Boethius
followed cannot have been very voluminous or profound.
But there are in Boethius two explicit quotations from Porphyry not found in
the Kata peusin. These must therefore, as Bidez admitted, come from the larger
commentary.16 But we must observe that in both cases the Greek manner of
ushering them in and out shows that Boethius is not quoting at first hand:
‘Certain commentators, among them Porphyry himself, argue as follows. . .:
that is what Porphyry says’; the very same kind of post-Porphyrian phrasing
occurs in Simplicius (379,8). Boethius shows no acquaintance with the numerous
other passages of the Pros Gedaleion preserved in Simplicius. These two pieces
are obviously detached fragments which he found in his Greek document:
immediately after they end, the Kata peusin reappears.
(3) Iamblichus (died c. 325 ad ). Boethius mentions Iamblichus by name twice.
And there are traces of his lost commentary at other points,17 for Simplicius in

14
This argument, written before the appearance of the Aristoteles Latinus, receives a totally new
emphasis from the work of Minio-Paluello, dramatically portraying the psychology of Boethius as
a translator. For instance, it is now abundantly clear that in the examples here quoted, to go no
further, Boethius could never have translated skheseis as diversitates, andrapodisamenos as
complector, topon as loca, hupographê as definitio, prothesis as intentio. He was simply not translating
Porphyry’s Greek, but that of a later work, ‘a Greek Neoplatonist commentary which had used both
commentaries by Porphyry on the Categories, but which was not that of Ammonius’ (I. Hadot, 105).
As for Ebbesen’s suggestion (1987, 292ff ) that ‘Boethius turned to Porphyry . . . in . . . concern for
pedagogy’, both the preference expressed for Porphyry and the concern for pedagogy are already
plain to be seen in the Greek commentators.
15
e.g. 264b, 266abc, 270ab, 277d–278d, 280bcd, 288d.
16
Attention should also be paid to the extract edited by P. Hadot.
17
180c, 181d, 184ab, 206d, 208b, 224c, 230c.
382 Aristotle Transformed

the parallel passages is expressly following Iamblichus. But Boethius’ two explicit
references to Iamblichus are obviously secondhand; at 162a he knows him
through Themistius’ criticism, and at 224c–225b he seems to be drawing upon a
similar Greek criticism: ‘in this matter we yield to the authority of Iamblichus’
(paralleled in Simplicius). This latter passage is again a detached fragment, for it
is awkwardly wedged into another comment on the art of composing names,
which begins at 224c and is resumed at 225b.18
(4) Dexippus (died c. 330 ad ). This commentator deals extensively with the
Plotinian ‘problems’ (aporiai) of Iamblichus. The extant part of his commentary
has hardly anything closely similar to Boethius.
(5) Themistius (died 384 or 385 ad ). The few fragments of Themistius’
commentary one can discover19 have no parallels in Boethius.
(6) Syrianus (died c. 437 ad ). The fragment of Syrianus preserved by
Simplicius in Cat. 164,4 is not in Boethius.20 But there are some affinities between
the two commentators. Boethius at 168c notes that Aristotle decided not to
discuss multivocal and diversivocal words in his opening chapter; this, according
to Simplicius (in Cat. 23,13), is a view of Syrianus. Syrianus was interested in the
pythagorica scientia mentioned by Boethius in his prooemium (160b). The
Athenian school of Syrianus was interested in rhetoric, and Syrianus himself
wrote a commentary on Hermogenes; Boethius wrote two short rhetorical
treatises.
(7) Proclus (died 485 ad ). Though no written commentary by Proclus is
mentioned by Simplicius in his list of commentators (in Cat. 1), his oral teaching
on the Categories continued the tradition of Syrianus, and his school was
flourishing in Boethius’ time.
Here it is opportune to examine Boethius’ prooemium (159a–163c). In it he
goes through the didascalica or kephalaia (lecturer’s headings), the traditional
items dealt with in the prooemia of philosophical and rhetorical treatises. E.A.
Quain has tabulated the number and order of these headings as they appear in
various ancient treatises.21 He shows that the early Aristotelian commentators

18
Such awkward combination of detached scholia is noticeable elsewhere in Boethius’ works. See
Wiesner, Aristoteles II , 328.
19
Themist. in Phys. 4,24; Olympiod. in Cat. 35,4; 48,14 [cf. AL I 5: Anonymi Paraphrasis Themistiana].
20
The addition of the cognomen ‘Philoxenus’, at Boeth. in Int. II 87 Meiser, indicates quotation in
Greek by someone later than Syrianus, not Syrianus himself (Pattin).
21
E.A. Quain, ‘The Mediaeval Accessus ad Auctores’ in Traditio 3, 1945, 243–56. Now there is the
important work of I. Hadot, ‘Les introductions aux commentaires exégétiques chez les auteurs
néoplatoniciens et les auteurs chrétiens’, in M. Tardieu (ed.), Les règles de l’interprétation, Paris 1987,
as well as of L.G. Westerink in Chapter 14 above.
 Boethius’ commentaries on Aristotle 383

have only two or three such headings (Porphyry’s Kata peusin has only two);22
but the commentators later than Proclus have an elaborate arrangement of
from six to eight headings. Now we find Boethius with a complete set of six as
follows:23

160a: intentio = ho skopos (the aim of the work).

161b: utilitas = to khrêsimon or hê khreia (its utility).
161c: ordo = hê taxis (its order of topics).
161c: ad quam partem philosophiae = hupo poion meros (which part of
philosophy it belongs to).
161d: Aristotelis liber = to gnêsion (its authenticity).
162a: inscriptio = hê epigraphê (its correct title).

The late commentator Elias remarks that the organiser of these didascalica
was Proclus (Elias in Cat. 107,24): ‘Proclus said that students beginning to read
in class the Aristotelian treatises should first go through all these headings, but
he did not give a reason; so let us now add the reason as we distinguish them’. A
briefer gloss intruded into Elias’ text says, ‘this is a Proclan arrangement’. Hence
Boethius’ fully-fledged system of headings seems to have come from some
source contemporary with or later than Proclus.24
(8) Ammonius and his school (c. 490 onwards ad ). It is immediately clear
from Quain’s table that Boethius’ prooemium is not modelled on Ammonius,
who omits the heading ad quam partem but gives another instead, ‘its division

22
Only title (epigraphê) and purpose (prothesis) receive formal and full treatment, but en passant there
is informal reference to order and to utility (only of preliminaries and appendices to the main text,
not of the whole treatise). For the references see I. Hadot, 127, nn. 45–8. Whether Porphyry, ‘the
cause of all good things for us’ (Simpl. 2,6), had a longer and more formal list in his larger
commentary we have no way of telling for certain. Certainly Proclus did not invent these headings;
he simply ‘arranged’ them, as Elias says.
23
Quain has overlooked two of Boethius’ headings in his table: Inscriptio and Aristotelis neque ullius
alterius liber est. The difference in number, vocabulary, and order, and the lack of technical terms,
clearly rule out Porphyry as Boethius’ immediate source.
24
I. Hadot (103) rightly points out that Elias, and therefore presumably Proclus in his oral classwork
(sunanagnôsis), gives a wider list of ten general questions and that these six kephalaia would come
only as answer to the tenth question: ‘For each treatise of Aristotle how many introductory points
should be considered, and what are their names?’ Boethius, working from fragmentary material,
knows only about the details under this tenth heading which, dealing specifically with the treatise
under consideration and not like the other nine about Aristotle’s philosophy generally, easily got
copied as a separate scholion and is found so detached in the margins of some Greek manuscripts.
The six Boethian kephalaia therefore – if Elias is to be trusted – may still derive from the teaching
of Proclus: they are found fully, though in varying order, only in the commentaries posterior to
Proclus, that is, the ‘Athenian’ Simplicius and the four ‘Alexandrians’. The fact that Boethius is
inconsistent about the total number in his other commentaries is a good sign that he was working
from fragmentary and discrepant Greek materials.
384 Aristotle Transformed

into chapters’ (hê eis ta kephalaia diairesis) (in Cat. 8,6). This division into
chapters displays in fact the whole arrangement of Ammonius’ commentary –
which never influences Boethius at all. Courcelle is not justified in concluding
that because Boethius has some of his didascalica corresponding to those of
Ammonius he is therefore following him.25 These stereotyped headings were
traditional; Ammonius is certainly following the scheme arranged by his master
Proclus.26 Dependence on the same source would explain the similarities in
Boethius.
Courcelle argues that the rest of Boethius’ commentary, too, is dependent on
Ammonius as well as on Porphyry. But the only parallels he adduces between
them are a pair of similar diagrams (Boeth. 175b; 291bc: Ammon. 25,12; 106).
This, allowing for the possibility of a common source, is not enough to prove
dependence. I find that detailed comparison produces about half a dozen other
parallel passages in Ammonius, but all these turn up in Simplicius as well –
which is in favour of an Athenian source for Boethius. Such comparison also
reveals several very definite discrepancies between Boethius and Ammonius.
For instance, the passages in Boethius which show the ultimate influence of
Iamblichus have no parallels in Ammonius. Bidez has observed that Ammonius
shows hardly any acquaintance with Iamblichus, for he often reproduces
comments of Porphyry that had been improved upon by Iamblichus.27 Neither
he nor his pupils, Philoponus and Olympiodorus, in commenting on the
Categories, give a single quotation from Iamblichus or Syrianus. Obviously their
mentality differs from that of the Athenian School.28
(9) Simplicius (after 532 ad ). Nearly all the Greek parallels we have found
for Boethius occur in the Athenian commentator Simplicius. He preserves
the school tradition of his predecessors Proclus and Syrianus, and, unlike
Ammonius, he makes full use of Iamblichus. His commentary is too profound,
too long, and apparently too late to have been Boethius’ source. But the
numerous parallels between them indicate a common origin – the oral teaching
of Proclus.

25
P. Courcelle, Les lettres grecques en Occident de Macrobe à Cassiodore, Paris 1944, 273.
26
Amm. in Int. 1,6; 181,30 Busse.
27
J. Bidez, ‘Unpublished papers on Porphyry’ (use of which I owe to the kindness of E.R. Dodds and
B.S. Page), cahier 16, 71.
28
Minio-Paluello from an early stage of his research into Boethian word-usage rejected with finality
the idea that Boethius depended on Ammonius. Yet it is Courcelle’s view which is still being retained
by many scholars, in spite of the fact that Courcelle himself eventually published in a gracious
review (see Bieler, xix, n. 24) his final palinode deferring to Minio-Paluello’s unrivalled insight into
the habits of the Aristotelian translators.
 Boethius’ commentaries on Aristotle 385

That, then, seems to be Boethius’ point of contact in the line of Greek

commentaries on the Categories.
To sum up: (a) Boethius does not give us an integral copy of Porphyry’s Kata
peusin. He seems rather to depend on a later Greek work in which the order of
Porphyry’s commentary had been changed, the catechetical form dropped, and
in which headings had been added. (b) The non-Porphyrian elements in his
work are not substantial enough to prove that he had a second complete
commentary in Greek. They are brief, and some of them appear to be detached
insertions. (c) Boethius does not distinguish the Porphyrian material from the
non-Porphyrian. For him it must all have been in the same hand. Apparently he
can name a source only where his Greek material has done so. (d) In several
passages he does not use the Kata peusin at all, but constructs his commentary
out of brief non-Porphyrian glosses or repetition of Aristotle’s text. (e) Some of
the non-Porphyrian elements are Proclan.
These threads will tie together as follows: Boethius’ Greek copy of the
Categories contained brief marginal explanations. Most of these were passages of
the Kata peusin copied verbatim;29 the rest were additions from later sources,
including Iamblichus. They all seem to derive from the school of Proclus.

2. In librum de Interpretatione
(two commentaries, editio prima and editio secunda)30

Here the only useful Greek material for comparison is the commentary of
Ammonius.31 But the parallels between them are sufficiently numerous to
indicate that at all points in his work Boethius is translating from Greek. I have
notes of ninety instances of Greek parallels.
These parallels, however, prove no more than that Boethius is everywhere
using Greek. They do not prove Courcelle’s thesis that he is following Ammonius.32
In the case of the Aristotelian commentaries textual parallels must be treated
carefully. One will often find the same words repeated through a whole series of

29
cf. Simpl. in Cat. 2,10. This, I now fear, was overstated out of youthful reverence for Bidez’s
Porphyrian scholarship. The similarities are only sporadically ‘verbatim’. Nor were the comments
always ‘brief ’. Some of them were evidently substantial, and all of them called for concentration and
skill in their translator: it is far from being a case of ‘one source, no thinking’ (Ebbesen), as any
modern Aristotelian translator (e.g. Barnes, 77) will agree.
30
Boethii commentarii in librum Aristotelis peri Hermeneias, Meiser, Teubner, 1877–1880.
31
The later commentary of Stephanus offers hardly any additional parallels.
32
Courcelle, op. cit., 274–8.
386 Aristotle Transformed

commentaries. What one finds, for instance, in Ammonius will almost certainly
reappear in Philoponus, Elias, or Olympiodorus. Therefore the parallel passages
given by Courcelle do not prove Boethius’ dependence on Ammonius. In one
such passage (361,9) Courcelle’s emendation, sicut Ammonius docet (Meiser:
sicut Eudemus docet; MSS : sicut audivimus docet), is unconvincing because
Boethius here does not in fact say the same thing as Ammonius: the latter does
not distinguish two kinds of ‘investigative question (erôtêsis pusmatikê)’; also,
there is no palaeographical justification for the change.33
In general structure the two commentaries are completely different. For
instance, Ammonius’ comment on Aristotle’s first chapter begins with the
standard treatment of five didascalica or headings. The last of these, the division
(7.15), marks off Aristotle’s work, excluding ch. 14, into four sections. There
follows a series of paragraphs, each dealing with a special problem arising out
of the text. All these points are missing from Boethius, whose division is into
six books. Vice versa, all the main points made by Boethius about the first
chapter are missing from Ammonius.34 It is the same through the rest of the
two works. Ammonius often quotes Plato and insists on the harmony of Plato
and Aristotle; Boethius merely quotes in turn Alexander, Aspasius, Herminus
and Porphyry, and sometimes adds a snippet from the Stoics or Syrianus.
In ch. 14 the difference is very marked: Ammonius goes through Syrianus’
six ‘contentions (epikheirêmata)’ against the authenticity of this chapter; of these
Boethius says not a word, but gives instead a very simple and thin commentary
on it. Finally, certain of Boethius’ phrases and examples that are obviously
taken from Greek, because we can find Greek parallels, are not found at all in
Ammonius. To take just a few:

. . . the intentio of the Categories is to deal with words that signify things through
the medium of thoughts (noêmata). [The Greek parallel for this is found in
Philoponus; Ammonius (like Porphyry) omits the mention of nous/noêmata as
the medium.]

Interjections, pronouns and adverbs come under Aristotle’s heading of ‘noun’

(onoma), for they can be treated as indeclinable one-case nouns (monoptota).
[Porph. in Cat. 62,1]

. . . ‘sword’, ‘blade’, ‘dagger’, can all be referred to a single substrate. [Simpl. in Cat.
38,26]

33
See my further note in Vivarium 12, 1974, 14–17: Meiser’s emendation is brilliantly correct.
34
This rejection of Ammonius, and of Courcelle’s emendation, is supported to the letter by
Tarán.
 Boethius’ commentaries on Aristotle 387

. . . having dealt with the preliminaries let us go on to the comments on Aristotle

himself [praedicenda is the literal Boethian translation of Greek prolegomena].

The concept and nature of the heavenly bodies is such that they are not receptive
of opposites [paralleled in Simplicius].

Here the philosopher wanted to exercise our intelligence [paralleled in

Philoponus].

A proof that these expressions are ‘by convention (secundum placitum)’ is that
their parts, verbs and nouns, are also ‘by convention’.

If just one syllable of the word ‘human’ is added to it that will not produce a
proposition. [For these last two instances parallel is found in the Syriac
commentary by Probus (which too is evidently translated from Greek), but not
in Ammonius.]

It therefore seems clear that Boethius’ commentary on the de Interpretatione

was not derived from Ammonius. Courcelle, in dwelling upon certain
resemblances, did not allow for the much greater divergences. These seem too
great to allow even his plea that Boethius may have followed a redaction of
Ammonius’ lectures differing from that now known to us. The various redactions
would not differ any more widely, say, than the commentary of Philoponus on
the Categories differs from that of his master Ammonius. But in the case of
Boethius and Ammonius the whole structure and nature of the commentaries
differ radically. Nothing indicates that one copied the other.35
There is one basic certainty. As in the commentary on the Categories, most
of Boethius’ material comes from Porphyry: ed. sec. 7,4: ‘I have translated
into Latin this elucidation, mainly from Porphyry but also from the others
(expositionem. . .quam maxime a Porphyrio, quamquam etiam a ceteris
transferentes, Latina oratione digessimus)’. Did Boethius know these ‘others
(ceteri)’ – obviously Alexander, Aspasius, Herminus and the Stoics – as well
as Porphyry? Bidez thought that he had direct access to the commentary of
Alexander,36 but the passages he adduces (87,15–16; 143–4) do not adequately
prove this. Courcelle rightly argues from 293,27 that Boethius took from
Porphyry all his information about the earlier commentators; this argument is
supported by the fact that when Boethius quotes them he habitually adds
Porphyry’s criticism of them. Several passages in Boethius which do not thus
35
As noted earlier regarding Porphyry on the Categories this argument could now be made more
succinct and decisive by use of the Aristoteles Latinus editions to illustrate the divergent vocabularies
of Boethius and Ammonius.
36
Bidez, ‘Unpublished papers’, cahier 16, 535–6.
388 Aristotle Transformed

mention Porphyry have parallels in Ammonius which do.37 It is therefore safe to

conclude that the main bulk of Boethius’ work comes ultimately from Porphyry.
But Porphyry does not account for everything in Boethius. There are four
quotations from Syrianus. They are obviously adventitious fragments: ‘At this
point there is a question by Syrianus’; ‘. . .let us add to this an extract from
Syrianus’ (18,26, Syrianus hoc loco quaerit; 321,20, nos quoque a Syriano
transferamus). Boethius shows no acquaintance with the other Syrianus extracts
preserved by Ammonius. But the four insertions mentioned are enough to show
his contact with a post-Porphyrian source. And at times his very manner of
quoting Porphyry betrays a later Greek medium: 160,5, ‘Porphyry has given a
fine explanation of when signified objects can be contraries’; the same
compliment is paid to Porphyry in the Greek of Ammonius. Besides, Boethius’
sixth book is not from Porphyry because, according to Ammonius (259,9),
Porphyry did not comment on this section of Aristotle (ch. 14); and nowhere in
this last book does Boethius give us his usual Porphyrian criticism of Alexander,
Aspasius and Herminus.
In general, it seems evident that Boethius was not following an exact or
complete copy of Porphyry’s work. The latter was lengthy (Stephanus in lib. de
Int. 63,9). But Boethius’ commentary is built from scanty Porphyrian material,
which he first translates and then applies to Aristotle’s text. In the case of some
lemmata the Porphyrian comment at his disposal was very small, and so he
expanded it considerably by comparing it with each phrase of the text.
It is the same in the non-Porphyrian parts. For instance, the long Syrianus
calculation at 321,20ff could all be constructed from a simple Greek diagram. In
fact, such a diagram is extant in the Syriac commentary (fifth century) of Probus
of Antioch,38 whose work, just like that of Boethius, has all the signs of being a
translation from Greek. Boethius’ method is particularly evident in his last book
where he had very little Greek and no Porphyry. The Greek comment in each
case was a brief ‘meaning’ of the lemma, a sententia. This he first translated and
then applied. For example:

428,12: That is the meaning; but the sequence of Aristotle’s words is as follows
(sensus ergo huiusmodi est; talis vero est ordo sermonum)
430,24: The comment fits the text as follows. . . (quare quod dicitur [i.e. the
sententia] hoc modo est [as applied to the text])

37
Compare: Boethius: 65,20; 66,18; 66,31; 68,18; 75,10; 105,15; 143,1
Ammonius: 44,11; 32,4; 47,26; 50,10; 56,16; 70,4; 94,29
38
J.G.E. Hoffmann, De Hermeneuticis apud Syros Aristotelis, Leipzig 1869, 111.
 Boethius’ commentaries on Aristotle 389

438,10: That is the entire comment; but the order of the text is like this (omnis
quidem sententia talis est, ordo autem sermonum huiusmodi est)
446,11: And if one goes on to connect the order of my comment to Aristotle’s
words. . . (quod si quis Aristotelis verbis seriem nostrae expositionis annectat)
449,14: Now I must turn to explain the order of the text (nunc quis sermonum
ordo sit explicetur)

Boethius’ material, then, consisted of brief notes in Greek, most of them, but not
all, based on Porphyry.
Examination reveals that both editio prima and editio secunda come from the
same Greek. The smaller is merely a shortened form of the larger: it displays the
same language and is divided under the same headings. In the first work Boethius
extracted simple and lucid points from among his Greek sententiae and in the
second he translated them entirely: 251,1, ‘What I omitted in the first edition
because of its depth or subtlety I must fill in as clearly as possible in the second
(. . .quam pote planissime quod in prima editione altitudinis et subtilitatis
omiseram secunda commentatione complerem)’. Where his Greek materials are
scanty this means repetition: 479,6, ‘Although I explained all this as fully as
possible in the second book of my first commentary, nevertheless to ensure that
the explanation of the work should not seem curtailed I shall go over it here
again (licet haec omnia in primae editionis secundo commentario diligentissime
explicuerimus ne tamen curta expositio huius libri esse videatur hic quoque
repetentes explicabimus)’. In the two prooemia we get a very clear proof of how
both commentaries use the same Greek material. The first editio gives a few
points about the kephalaion of intentio: 32,7, ‘A brief clarification of its intentio
must be given (quae sit intentio breviter demonstrandum est)’. To these the
second adds other details, without repeating anything previously said – not even
the word intentio – and then goes on to give the remaining three kephalaia: ‘title’,
‘authenticity’, ‘utility’ (titulus, iudicium libri, utilitas). All four were certainly
together in the Greek. The first edition therefore gives us a small selection of that
Greek and the second gives the rest of it.
Furthermore, the two editions between them give us all the Greek material. If
Boethius in the editio secunda omits to translate any point he tells us so. 201,2,
‘But Porphyry here inserts certain details about Stoic logic. These are foreign to
Latin ears and therefore the point at issue would remain unclear. So I shall omit
them (Porphyrius tamen quaedam de Stoica dialectica permiscet: quae cum
Latinis auribus nota non sit, nec hoc ipsum quod in quaestionem venit agnoscitur
atque ideo illa studio praetermittemus)’. In the smaller commentary Boethius
390 Aristotle Transformed

does not mention the views of Alexander, Aspasius and the others, but only
admits enough of his Greek source to give a simple explanation of the argument
and text: 32,3, ‘The reader should now expect of me only enough for him to grasp
small details one by one following the order of the text, so that he may discover
what lurks in its brevity (nunc a me tantum lector exspectet quantum pedetemptim
minutatimque secundum orationis ordinem textumque sermonis id quod angustia
brevitatis latet intelligat)’. But in the larger work he claims to give us all the
Greek he had previously omitted: 251,1, ‘quod in prima editione. . .omiseram
secunda commentatione complerem’. This plan of splitting up the Greek
comments is his own idea:39 186,2, ‘The best plan for me was to open up the
closely argued meanings of Aristotle through a double sequence of commentary
(quare recte mihi consilium fuit subtilissimas Aristotelis sententias gemino ordine
commentationis aperire)’.
And so we reach the same conclusion as in the case of the in Categorias:
Boethius is translating marginal notes, the explanatory sententiae that he
encountered in his Greek Aristotle: 250,20, ‘The textual sequence is obscure and
it is supplemented by extremely obscure scholia (est quidem libri huius. . .obscura
orationis series, obscurissimis adiecta sententiis)’.40 With the de Interpretatione he
found a great mass of these marginalia and it took him a long time to decipher
them: 421,3, ‘For there are scholia of numerous points heaped up all together and
so I have spent almost two years in a constant sweat of writing comments (nam
et plurimorum sunt in unum coacervatae sententiae et duorum ferme annorum
spatium continuo commentandi sudore consumpsimus)’.41 These comments were

39
But it was probably suggested by the lay-out of his Greek manuscript.
40
‘An obscure course of argument with highly obscure notes added to it’, the translation rejected by
Ebbesen (290, n. 2), is not mine but that of Barnes (79). But it is correct. Ebbesen adds: ‘as far as I
can see, the passage means “the doctrines of the book are very obscure, and on top of that the
manner of presentation is obscure”.’ But, surely, orationis series (= hê taxis tou logou) is not just
‘doctrines of the book’. Also, adiecta agreeing with series does not mean ‘and on top of that’ (= et
praeterea): the real (post-classical) meaning of the word is clarified by Boethius a few lines later:
‘where Aristotle in his brevity is obscure, I, by making some additions (aliquibus additis), may
render the sequence easier through that annexation (adiectione).’ And sententiae is not the ‘manner
of presentation’: the correct sense, ‘notes added’, is warranted by other Boethian passages and is the
same as that clearly seen in Isidore’s chapter-heading de notis sententiarum (‘on reference signs for
locating marginal comments’). See Scriptorium 38, 1984, 340.
41
This statement is meant to spell out the one just before it announcing that the sixth book will
conclude the long commentary composed ‘with great labour’ (magno labore) and ‘delay’ (temporis
mora): the reason for the great labour is the large ‘pile’ of sententiae going with the text, and the
delay is explained as ‘two years’. One should not translate coacervatae sunt (were heaped up together)
as if it were coacervavimus (I have heaped together) matching the first-person consumpsimus as
Chadwick (129) does: ‘Boethius . . . compressed into a single book’. The sweat (sudor), moreover, did
not in my view ‘consist in the deciphering of the tiny handwriting’ (Chadwick, 129); the ancient
scholia were not simply carelessly ‘scribbled in the margins’ (Barnes, 80); and I did not understand
coacervatae to imply ‘compressions’ (Chadwick, 129), but rather the very opposite, ‘heaps’.
 Boethius’ commentaries on Aristotle 391

(7,5) ‘mostly from Porphyry (quam maxime a Porphyrio)’, but Boethius faithfully
‘translates all the other notes (quamquam etiam a ceteris transferentes)’.
Sometimes these are under a commentator’s name (18,26: Syrianus hoc loco. . .)
but sometimes they are anonymous: 19,20, ‘This too is in the margin (illud
quoque est additum)’. The translation of these various marginalia and the
arrangement of them into a continuous commentary according to the order of
Aristotle’s words would seem to be Boethius’ only title to originality.42

3. In Isagogen Porphyrii43
Although Boethius in the editio prima of this work retained Victorinus’
translation of the Isagoge he can hardly have used a commentary by Victorinus:
the list in Cassiodorus’ Institutiones II 18 (p. 128 Mynors) mentions only the
translation by Victorinus but no commentary by him; and apparently Victorinus’
work was nothing more than a concise one-book paraphrase (de Definitione
897a: iam uno libro de bis quinque rebus plenissime disputavimus). Anyway,
Boethius reveals that even in this first commentary he is doing pioneer translation
from Greek: 8,11 Brandt: ‘noêta I have never found in Latin translation; so I have
coined a new word for it, “intellectibilia” (quoniam Latino sermone numquam
repperi, “intellectibilia” egomet mea verbi compositione vocavi)’. He shows too that
even for the editio prima he had ready at hand the Greek text of the Isagoge, for
on occasion he is able to correct Victorinus’ version: 23,17, ‘Porphyry therefore
promises to explain these five terms briefly and in proper measure – but this is
omitted by Victorinus’ (haec igitur Porphyrius, non enim Victorinus, breviter
mediocriterque promittit exponere)’; 34,12, ‘But Victorinus reduces those two
genera to one (Victorinus vero duo superiora genera in unum redigit)’. Brandt
gives a list of 17 parallels, covering both editions, between Boethius and the
Greek of Ammonius;44 and I find that detailed examination would multiply the
number.

42
Even this may be an overstatement if we allow for ancient editorial activity. Such activity is very
visible at in Cat. 263b, where it has corrupted all the extant manuscripts through Latin ignorance of
a Greek title familiar to Boethius (Shiel 1957). In the case of the Boethian Isagoge editorial effort is
again detectable at work on an original text laden with glosses (Shiel 1987). It will be the same
elsewhere. ‘To render Organon he feels obliged to use a hendiadys, ferramentum et quodammodo
supellex (“instrument or perhaps implement”)’: Ebbesen 1987, 187. This duplication is not normal
Boethian procedure, and it is anyway a poor specimen of hendiadys: it looks much more like an
editor’s attempt to cope with a variant gloss: supellex/ferramentum.
43
Boeth. in Isagogen Commenta S. Brandt, CSEL , Vienna 1906.
44
Boeth. in Isag. Brandt, xxiv–xxvi.
392 Aristotle Transformed

All this makes it quite clear that in composing both editions Boethius had at
hand (a) a Greek text of the Isagôgê and (b) a Greek commentary.
Which commentary? Brandt noted that although there was a strong
resemblance between Boethius and Ammonius there were also notable
divergences:45 ‘It would be a mistake to think that there are no discrepancies in
the general consensus of idea and word between Ammonius and Boethius. There
are, and they are serious (minime sane in universa quae inter Ammonium et
Boethium est sententiarum atque etiam verborum congruentia differentiae eaeque
graviores desunt)’. Consequently, though he was inclined to believe that Boethius
followed Ammonius, he was also ready to admit that both might have followed a
common authority: ‘This proviso should be added: if one feels that Boethius had
for his source not Ammonius himself but rather Ammonius’ own source, that
conclusion should not hastily be dismissed (verumtamen is modus iudicio
imponendus est, ut si quis Boethium non Ammonium ipsum, sed eum quem
Ammonius ducem habuisse velit, ea sententia non prorsus improbanda sit)’. On
our argument this seems the correct view. The divergences between Boethius
and Ammonius are much greater even than Brandt supposed. The main plan and
headings of Ammonius’ commentary, which he sets out at the beginning,46 and
his frequent references to Plato, are unknown to Boethius. The similarities, on
the other hand, are explained by their common source.
As in the case of the commentaries on the de Interpretatione, so here too
Boethius seems to have used the same Greek material for both editions. For
example, the prooemium of the editio prima (3,1–16,10) gives the enumeration
and explanation of the six didascalica. The editio secunda does not repeat these
but gives four other introductory points: (a) the division of powers in the soul
and the consequent ‘methods of dialectic’; (b) the origin of logic; (c) is logic a
part or only an instrument of philosophy?; (d) the subject matter of the Isagoge.
In the Greek commentaries these points are always found in the prooemia along
with the didascalica.47 Again, in the second edition (140,13ff ) there occurs a
division of philosophy which differs from that given in the editio prima (8,1); but
both were together in the Greek material, for Boethius was able to say in
the editio prima (9,21): ‘there are other subdivisions of these which I must omit
for the present (sunt earum etiam aliae subdivisiones quas nunc persequi
supersedendum est)’.
45
ibid., xxiv.
46
Amm. in Isag. 23,12 Busse.
47
Similar displacement elsewhere (cf. Wiesner 318–19) indicates that Boethius, or his ancient editor,
was unfamiliar with the full traditional order of introductory headings in a major, continuous
Greek commentary. For this order now see I. Hadot, 120.
 Boethius’ commentaries on Aristotle 393

Other passages likewise indicate that for the first work Boethius picked from
the Greek only such items as he thought necessary for beginners. Brandt48 gives
a list of the points that are not repeated in the second work. They are mainly
detached definitions of terms, e.g. of ‘descriptive explanations (hupographoi
logoi)’ at 42,14. Brandt also observes that the subject-matter of the second edition
seems to him to be really as elementary as that of the first; that Boethius keeps
calling the second work, too, an introduction for beginners; and that there is no
real discrepancy between the first and the second. All this is explained by the
assumption that Boethius constructed both from the same Greek material. Twin
offspring, therefore, of the same Greek parent, the two editions are aptly called by
Boethius a gemina expositio.
The Greek source was obviously post-Porphyrian: 12,20, ‘all beginners in logic
following Porphyry (omnes post Porphyrium ingredientes ad logicam. . .)’. Also it
was brief. The latter part of Ammonius’ Greek commentary consists mainly of
brief glosses on words of the text. But Boethius, by comparison, has even less to
go on. What might have been a one-word gloss, or a diagram, in Greek, he
expands into a paragraph by repetition of Porphyry’s text. For example, his long
explanation of logikôteron at 168,11 could all have been built on a one-word
gloss like endoxos; and 319,15–325,7 could all be taken from a diagram. From the
preface to Busse’s edition of the Isagoge one can see that the oldest extant
manuscripts of the Isagoge (ninth century) have comments of this sort in the
margins, the comments being thickest on the opening pages of the text. It must
surely have been the same in Boethius’ time. His Greek document would have
been a text of the Isagoge with brief marginal scholia. With the aid of some of
these scholia he sets out in the first edition to elucidate the version of Victorinus,
which he allows to stand in order not to confuse beginners by the strange terms
of a new translation. But in the second edition his aim is to give them everything
in his Greek text and marginalia: 135,7, ‘I shall evidently have done good service
if by providing philosophic books in Latin written with all the lucidity of total
translation I shall have no further need to look for them in Greek (multum
profecisse videbor si philosophiae libris Latina oratione compositis per integerrimae
translationis sinceritatem nihil in Graecorum litteris amplius desideretur)’.49

48
Boeth. in Isag. xxi.
49
This passage refers not simply to ‘a complete and unadulterated translation of Porphyry’s Isagoge
(and other basic texts), not of scholia’ (Ebbesen 1987, 290, n. 3); for it immediately introduces Greek
commentary material, not found in the Isagoge itself, about the triple powers of the soul. When
Boethius opened his Greek philosophiae libri his eye fell at once on scholia as well as text.
394 Aristotle Transformed

4. The treatises on the syllogism

(Introductio ad Syllogismos Categoricos; de Syllogismo Categorico libri III; de
Syllogismo Hypothetico libri III )50

Comparison of the main points dealt with in the Introductio51 and in the first
book of the de Syllogismo Categorico makes it clear that both are translations –
and both from the same Greek material: 794c, ‘. . . the commentaries from which
I have taken these points (eos commentarios de quibus haec nos protulimus)’. That
material is mainly a summary of the de Interpretatione, but the final part of it is
obviously a simple Greek diagram of ‘propositions (protaseis)’ of which Boethius
makes a double elaboration (Introd. 769c–792c; de Syll. Cat. I 798c–807a).
This double summary is apparently the summary (breviarium) which Boethius
promised during his commentary on the de Interpretatione.52 Usener, in fact, has
already argued53 that the Introductio seemed to him to be the promised summary,
but Brandt,54 followed by Cappuyns,55 rejected this view on the grounds that (a)
three whole chapters of the de Interpretatione are omitted in the Introductio, and
(b) Boethius’ de Syllogismo Categorico – and therefore the Introductio, which is so
similar – was written before the commentary on the de Interpretatione. But
against this: (a) Boethius did not promise to make the breviarium adhere to all
Aristotle’s chapters but only ‘in some, nearly all’ of them (in quibusdam et fere in
omnibus); (b) Brandt’s method of establishing the chronology of Boethius’ works
by their cross-references is excluded by our argument that these references are
not always Boethius’ own, but based upon his Greek scholia.
Boethius’ Introductio and de Syll. Cat. I are twin translations of a Greek
breviarium, a summary of the de Interpretatione doctrine of categoric
propositions, serving as an introduction to the Analytics. Porphyry himself
seems to have written such an introduction,56 and we even find him duly listing
the de Interpretatione under the title of ‘Categoric Propositions’ (Kata peusin
56,25: ‘Topics,. . .Analytics,. . .Categoric Propositions. . .Categories’).

50
PL 64, 761,876. This paper, devoted to the commentaries proper, paid too scanty attention to the
treatises. But I cannot regard these, however impressive their contents, as simply ‘Boethius’ own
work’ (Stump). They are versions of more substantial scholia (prolegomena, eisagogai) set between
the works of the Organon. That some Greek scholia were of monograph length and sometimes
complete with titles can be seen from the Aristoteles Graecus.
51
cf. Cappuyns, op. cit., 370: ‘According to Bidez’s opinion as reported by A. van de Vyver, Les étapes,
443–6, it would represent a second redaction of de Syll. Cat. I.’
52
in Int. ed. 2a 251,8.
53
H. Usener, Deutsche Litteraturzeitung, 1880, col. 369.
54
S. Brandt, ‘Entstehungszeit und zeitliche Folge der Werke von Boethius’, Philologus 62, 1903, 258.
55
Cappuyns, op. cit., 369.
56
J. Bidez, Vie de Porphyre, Gand 1913, 66.
 Boethius’ commentaries on Aristotle 395

Indeed it seems that through all the three works on the syllogism, as in the
previous commentaries, Porphyry is the main authority in Boethius’ material. We
find here, as before, the views of the ancients (Theophrastus and Eudemus)
followed by Porphyry’s criticism: de Syll. Cat. 813c, ‘. . . others add one other mode
(in the third figure), as Porphyry himself does, following those predecessors
(addunt etiam alii unum sicut ipse Porphyrius, superiores scilicet sequens)’. And, as
before, Boethius apparently found this Porphyrian material in scholia added to the
Analytics as prolegomena: 831c, ‘I found this in very few of the Greeks, expressed
sketchily and confusedly, and never in Latin (. . .apud scriptores quidem Graecos
perquam rarissimos strictim atque confuse, apud Latinos vero nullos repperi)’. 829d,
‘This is what I had to say as introduction to Categorical Syllogisms – all that the
brevity of an introduction would permit (haec de categoricorum syllogismorum
introductione quantum parcitas introducendi permisit expressi)’.
Only these three introductory treatises are extant, but apparently Boethius’
text had other marginalia from which he hoped to compose a complete
commentary on the Analytics:57 829d, ‘. . . if any points are missing we shall deal
with them more insistently in our Analytics (si qua vero desint in Analyticis
nostris calcatius exprimemus)’.
The way Porphyry is mentioned at 813c, ‘as Porphyry himself says (sicut ipse
Porphyrius)’ and at 814c, ‘Porphyry, an author of the highest standing (Porphyrius
gravissimae vir auctoritatis)’ suggests a Greek scholiast later than Porphyry. And

57
An MS ‘Latin commentary on the Analytics’, discovered by Dr L. Minio-Paluello, seemed to him to
be by Boethius: Journal of Hellenic Studies 77, part 1, 1957, 93–102. See my corroboration of his view
in Vivarium 20, 1982, 128–41. I have since examined the language of these ‘Scholia Anonyma
Graeca a Boethio Translata’ in close association with Dr Minio-Paluello, who approved my textual
corrections, and the ascription to Boethius, for inclusion as an appendix in the next printing of AL
iii. With the kind permission of the present general editor of AL , and after meticulous editing by
Dr J. Brams, this supplement has meanwhile been made public in Bull. phil. méd. 26. A chart of the
word-usage (in the same journal) will, I believe, confirm further that the initial guess about Boethian
authorship was a good one; it will also rule out en passant the conjecture (Ebbesen 1981, 9) that the
translator might have been James of Venice – or any of the other translators so far edited in AL .
The Florentine scholia cannot be convincingly regarded as mediaeval ‘excerpts from a whole
commentary that was available in Latin’ (Ebbesen 1981, 3; 1987, 291, n. 15). All the indications are
that the Aristotelian text, and its marginal scholia, and their reference signs, were taken all at once
from Greek by the same translator, and, as Minio-Paluello pointed out to me, the Florentine codex
was specially ruled to accommodate both text and marginalia, all copied in the same hand as if
carefully following an exemplar that was similarly arranged. The contention that the excerpting had
already been done in Greek could be further supported, I believe, from internal evidence in the
scholia themselves. So if we are to have a label for the lost Greek source it is not a ‘commentum
Graecum’ but – to borrow a plural from Boethius himself – commenta Graeca.
It does not help much to call the second book of the Philoponus Greek commentary ‘a Byzantine
collection’ (Ebbesen 1981, 10). Minio-Paluello in his edition of the Boethian scholia in AL iii refers
to it, following Wallies, as ‘pseudo-Philoponus’, but that does not mean that either he or Wallies
considered its contents mediaeval. As in the case of the second half of Alexander’s commentary on
the Topics Wallies thought the second book different in style and authorship from the first. But there
is nothing in it by way of either allusion or vocabulary to suggest that it is not ancient in content.
396 Aristotle Transformed

Dürr’s argument58 that the logical doctrine of the de Syllogismo Hypothetico is

different from that contained in Ammonius’ work of the same title suggests
somebody earlier than Ammonius. Once again we find ourselves thinking of the
school of Proclus.

5. De Divisione59
This too is a translation of a Greek Introduction: 877a, ‘So, as with my other
works, I have translated it fully into Latin as an Introduction (ego quoque id sicut
pleraque omnia Romanis auribus tradens introductionis modo. . .perscripsi)’.

6. Commentaria in Topica Ciceronis60

In this work Boethius undertook to complete Victorinus’ commentary, which,
though occupying four books, covered only a fraction of Cicero’s text.61 Boethius
constantly shows off his possession of the Greek text of Aristotle’s Topics, which
neither Cicero nor Victorinus knew.62
Boethius has a quotation from the Physics,63 but he had found it in his Greek
notes: 1152c, ‘But if anyone is interested in looking at my commentaries he
should know that I have taken over this point from the second book of Aristotle’s
Physics (sed si cui commentarios nostros inspicere vacuum fuerit, sciat haec nos ex
Aristotelis secundo Physicorum volumine advertisse)’. In fact he seems to be
referring back to his commentary on the de Interpretatione (ed. 2a. 190,12ff ),
where he used the same Greek material.

7. De Differentiis Topicis64
Boethius tells us that he takes his differentiae from (a) Greek works and (b)
Cicero’s Topics: 1173c, ‘There are two traditions to be presented, one extracted
from Greek books and the other from the Topics of Cicero (duplex est tradenda
traditio, una quidem ex Graecis voluminibus eruta, altera vero ex M. Tullii Topicis
sumpta)’. But the Graeca volumina turn out to be (1133c) ‘the eight books of my

58
K. Dürr, The Propositional Logic of Boethius, Amsterdam 1951, 13ff.
59
PL 64, 875–92.
60
ibid., 1039–74.
61
cf. Boeth. in Top. Cic. 1041d.
62
ibid., 1051b; 1052b; 1054b; 1091cd; 1098b; 1119b; 1145d.
63
The quotation from the Physics at 1152c is 196b 10–11 (and the mode of quotation suggests that
Boethius took it from a commentary rather than the original text): omission of reference noted by
Stump.
64
PL 64, 1173–216.
 Boethius’ commentaries on Aristotle 397

Latin translation of Aristotle’s Topics (octo voluminibus. . .quibus Aristotelis

Topica in latinam vertimus orationem)’.
Boethius also refers three times to a commentary of his own on Aristotle’s
Topics.65 Hence – exactly as in the case of the Analytics – either he had translated
from scholia a commentary which is now lost or else he was thinking of scholia
which he was on the point of translating.
All the dialecticae rationes used in the de Differentiis Topicis by Boethius,
including the long discussion of magis, minus etc., are elaborated from the simple
Themistian diagram given in contrast to Cicero’s at 1202c–d. Boethius’ other
references to Themistius (1214a, 1215a) are obviously based also on that
diagram.66 The diagram came from Themistius’ commentary on the Topics, now
lost.67 And since it is precisely for this Themistian material that Boethius refers
to his own commentary on the Topics, I take it for certain that the diagram was
a scholion in his Graeca volumina of the Topics.

II

Before drawing our general conclusion about all these commentaries and
treatises, a final doubt remains to be dispelled – the possibility of Latin
intermediaries, the philosophic group of Macrobius and Victorinus. The
Macrobius circle devoutly studied Porphyry and Themistius.68 The circle
included the logicians Vettius Praetextatus and Albinus, as well as Symmachus,
grandfather of Boethius’ father-in-law. In this circle may have been produced the
pseudo-Augustinian Categoriae Decem.69 But Boethius makes it clear that this
school did not help him: he shows no acquaintance with the Categoriae Decem;
he says that he searched for Albinus’ works but could not find them; he did have
Praetextatus’ Analytics but on reading them could easily detect that they were
really translated from Themistius and not from Aristotle, whereas he himself
has the genuine Aristotelian text.70 He could find in Latin no commentaries on

65
1191a, 1209c, 1216d.
66
I did not, however, mean to suggest that this diagram was the only scholion in the book (Stump) but
that its data were repeated, in the common repetitive manner of detached Greek scholia.
67
cf. Themist. in An. Post. 18,26 Wallies. I follow the CAG and Wallies rather than Stump in seeing
here a reference to Themistius’ own commentary (or paraphrase) on the Topics.
68
Courcelle, op. cit., 32–5.
69
Now edited as Paraphrasis Themistiana, in AL I; L. Minio-Paluello, ‘The Text of the Categoriae’,
67–8.
70
in Int. ed 2a. 4,4; 3,7.
398 Aristotle Transformed

the syllogism either.71 He openly displays the few items of Latin he did have,
the Isagoge, de Definitione and in Topica Ciceronis of Victorinus – but their
inadequacy only emphasises the superiority of his own Greek material and his
duty of translating it. Hence the starkly Greek idiom of his language and his
severe lack of reference to Latin models.

III

The general impression produced by this study is that Boethius in composing his
commentaries on the Organon translated Greek notes which he found added to
his text of Aristotle. If this is true, it gives us new insight into the way Boethius
worked.
From the beginning it is evident that he considered the works of the Organon,
including Porphyry’s Isagoge (which Neoplatonic schoolwork put on a par
with Aristotle), as a united whole. He begins, in fact, by listing the order of its
contents, and his aim is to translate and comment on it all as one series: in lib.
de Int. 343,7, ‘In an attempt to transmit and explain its entire content, I have
compiled a continuous sequence of commentary (. . .totius prodere adgressos
atque expedire doctrinam. itaque rectam commentationis seriem conteximus)’.
And the homogeneous character we have noted in his Porphyrian material is
another indication that he found everything together in Greek.
The task he set himself was to put all this Greek, both text and scholia,
into Latin: in lib. de Int. ed. 2a, 79, 16, ‘I shall translate into Latin every work of
Aristotle I can lay hands on, fully translating all their scholia as well (ego omne
Aristotelis opus quodcumque in manus venerit in Romanum stylum vertens eorum
omnium commenta Latina oratione perscribam)’. But the commenta were rather
troublesome. Given the primary labour of translating the Aristotelian text, he
could not deal with all these scholia at once: he therefore picked some for an
elementary first editio and postponed the others to a second. So, in the case of
the Isagoge he began in the first editio with a selection of scholia which he
elaborated into a dialogue. But the dialogue style proved cumbrous for all these
technical notes, and in the second editio he abandoned it in favour of unadorned
literal translation, allowing himself to lapse into the ‘fault of the faithful translator
(fidi interpretis culpa)’.72 For the Categories we possess only one commentary, but

71
de Syll. Hyp. 831c.
72
in Isag. 135,7.
 Boethius’ commentaries on Aristotle 399

the manuscripts call it too an editio prima, and in it Boethius expressly postpones
some Greek notes to a second work.73 However, his Greek material for it seems
to have been more scanty than for the de Interpretatione, where it took him two
whole years of labour to pack the last of the scholia into his second editio. With
the Analytics and Topics he apparently managed to translate only the lengthy
introductory scholia (eisagôgai, prolegomena).
All his Greek text and scholia would fit comfortably into one codex. In fact a
good example of the kind of book he had before him would be our oldest Greek
manuscript of the Organon (Urbinas 35).74 Written in Byzantium in the ninth
century, during the renaissance that followed a three-century decline of
Aristotelian learning, its format probably resembles that of its sixth-century
predecessors. It has been shown that the Byzantine editors did not venture to
interfere with the order of books in the Organon;75 very likely, then, they copied
out marginalia as they found them.
Some of Boethius’ scholia, as we have seen, show a connection with the school
of Proclus in Athens. His whole codex may well have had its origins there.76
Our view of Boethius the translator of scholia is not quite in harmony with all
that scholars have written about him. It dulls the praise of his vast scholarship;
for where he mentions works of Aristotle other than the logic, he is just translating
references among his Greek scholia,77 so that all his knowledge of Aristotle seems

73
in Cat. 160b.
74
But not just this one manuscript. The general Aristotelian manuscript tradition was one of
annotation. The chapter headings in the Aristoteles Latinus editions of Boethius’ translations are
one ancient sign of this, but there are many others. The habit of systematic marginal annotation was
well established in the Greek scriptoria before the time of Isidore. This is confirmed in recent work
by papyrologists such as Turner and Wilson. See my references in Scriptorium 38.
The idea that it could have been a single codex, rather than several, containing the Organon
treatises is palaeographically feasible but remains unsupported. But now combined study of the
Aristoteles Latinus and the new Aristoteles Graecus would be illuminating on this point.
75
F. Solmsen, ‘Boethius and the history of the Organon’, Amer. J. Philol. 65, 1944, 71.
76
There were several Aristotelian schools and a connection with Proclus is difficult to substantiate
because his teaching on the Categories was oral, and so there are no quotations from him in
Simplicius. But the general expressions of indebtedness to ‘the successor of Plato’ in Simplicius and
Ammonius in other works mean that we cannot ignore him. Nor can we ignore the explicit
assertion of Elias about his elaboration of the Kephalaia into the scheme known to Boethius. On the
other hand we cannot force the evidence. ‘If God exists whence evils, and if he does not whence
good things? (Si deus est unde mala? Bona vero unde si non est?’ [Consol. I 4.30]) ‘can be securely
identified as a verbatim quotation from Proclus in Parm. 1056,10–16’, according to Chadwick (129).
But Boethius (in 12 words) is speaking about God (deus), and Proclus (in 46 words) is speaking
about Providence (pronoia), a rather different topic in Greek philosophy. Unde mala?, incidentally,
is a verbatim translation of the question reported by Plotinus at Enn. I 8:1,1 – a tiny reminder that
the arguments of the Consolatio are built up, only more elaborately, from the very same kind of
Greek School material as the Organon commentaries, as Sulowski has so well indicated (see Bieler,
xxv, no. 146).
77
At in Int ed. 2a 27,14; 28,1 he carefully transcribes the Greek of two fragments of the de Anima and
de Iustitia.
400 Aristotle Transformed

to derive from this one codex of the Organon. Again, an attempt has been made
to fix the chronology of his works by their cross-references;78 this, of course, will
not work if similar references were already in his Greek. And scholars have been
disappointed because Boethius in one place made a great promise which he
never fulfilled – to harmonise all the doctrines of Plato and Aristotle;79 but the
promise is probably based on words of Porphyrian Greek – Porphyry in fact did
write a book called On the Unity of Doctrine in Plato and Aristotle.80 It has also
been argued that, because of his similarity to his contemporary Ammonius, he
cannot just have worked in a Roman philosophic school or in the solitude of his
library: he must have been to Alexandria and heard the latest lectures of
Ammonius there.81 But we have explained his similarity to Ammonius by their
common source, the Athenian school of Proclus (Atheniensium scholas);82 and,
given his acquaintance with previous Latin writers on logic and the fact that he
was ‘an accomplished master of both languages (utraque lingua peritissimus)’,
there was nothing to hinder him from translating an annotated Organon in his
beloved bibliotheca far removed (longe positus) from either Alexandria or Athens.
This work of translation83 does not diminish his real greatness. It is precisely
the ground on which his friend Cassiodorus rhetorically praises him (Variae I
45): ‘Though living at a distance you were able to enter the Athenian classrooms,
mingling your Roman toga with the chorus of Greeks clad in pallium, and so you

78
Brandt, op. cit.
79
in Int. II 79,16ff; cf. E.K. Rand, Founders of the Middle Ages, Cambridge Mass. 1928, 215ff;
A. Kappelmacher, ‘Der schriftstellerische Plan des Boethius’, Wiener Studien 46, 1928, 215ff.
80
J. Bidez, Vie de Porphyre, 68.
81
Courcelle, op. cit., 300.
82
This does not necessarily mean that his actual Greek copy (volumina) of the Organon came to him
directly from Athens and not from one of the other imperial scriptoria. Courcelle in concentrating
on Alexandria overlooked the fact that details in many of his other pages carried his readers’
attention rather to Constantinople as the cultural capital to which Symmachus and Boethius turned
for their inspiration and ‘programme des études’.
83
Even Boethius’ first-person words are in some instances clear translations, in others arguably so. ‘I
shall re-read Theocritus today’ (in Int. II 234,10–13) is no real clue to Boethius’ own poetic tastes
(Magee) but rather an intrinsic part of Philo’s Greek explanation of ‘possibility’ by definition with
example, just as in the companion comment (235,6–8) from Diodorus, ‘drowning at sea’ is the
similar Greek example. When meeting with the same example in a more anonymous context
(207,21; 240,10) Boethius, characteristically, changes the Greek poet’s name to Virgil. And the
words he uses to describe his work (scribo, conscribo, perscribo, exprimo, edo, dissero, compono,
digero, etc.) do not imply the modern notion of ‘creative writing’. Admittedly his transfero stands
unequivocally for ‘translate’. But likewise, perscribo, for instance, at in Int. II 79,16 (latina oratione
perscribam) and de Div. 877a (romanis auribus tradens . . . perscripsi) only amounts to ‘translate
fully’.
Boethius’ work contains his own prolix expansions and connections and re-arrangements (or
those of his early editor), Latin names substituted for Greek, personal prefatory remarks, an attempt
at dialogue style, recourse to Latin writers, especially Cicero. To interpret me as thinking that ‘every
line and every word of Boethius’ commentaries was a translation of Greek notes’ (De Vogel) is, I
fear, più romano che il papa.
 Boethius’ commentaries on Aristotle 401

have turned the Greek dogmata into Roman doctrina (sic enim Atheniensium
scholas longe positus introisti, sic palliatorum choris miscuisti togam, ut Graecorum
dogmata doctrinam feceris esse Romanam)’. And there is a patriotic nobility
about his motives. As consul he thinks it his duty ‘to instruct my citizens in the
studies I have laboured over (elucubratae rei doctrina cives instruere)’ (in Cat.
201b) in conscious imitation of Cicero who as senator felt obliged to write his
philosophic ‘transcripts (apographa)’. In time of sorrow such concentrated study
was Cicero’s ‘consolation (levatio)’, likewise for Boethius it is a ‘life’s supreme
consolation (summum vitae solamen)’. Cicero retired to his Tusculan study,
Boethius to his ‘study walls adorned with ivory and glass (bibliothecae comptos
ebore ac vitro parietes)’. Our study of him as a translator emphasises anew his
remarkable role of transmission: through him Aristotelian logic, the equipment
of Neoplatonic paganism, is carried into the Christian Church to be eventually
part of its armour of faith.84

Bibliography

Aristoteles Latinus I 6 xiv n. 3; II 1 ix n. 2.

J. Barnes, ‘Boethius and the study of logic’, in M.T. Gibson (ed.), Boethius, Oxford 1981,
79–80, 88 n. 16.
L. Bieler (ed.), Boethii Consolatio, editio secunda, Turnholti 1984, Praefatio.
M. Cappuyns, review of Shiel (1958), in Bulletin de théologie ancienne et médiévale 8,
1958–1961, no. 604.
H. Chadwick, Boethius, Oxford 1981, 129–31, 153, 299 n. 42, 300 n. 53.
H.D. Dauphin, review of Shiel (1958), in Rev. d’hist. eccl. 54, 1959, 1120.
J. Dillon, review of Chadwick (1981), in Classical Review, n.s. 33, 1983, 118.
S. Ebbesen, ‘Fragments of “Alexander’s” commentary on the Post. Analytics’, in CIMAGL
16, 1976, 89–98; ‘Analysing syllogisms’, in CIMAGL 37, 1981, 1–20; ‘Boethius as an
Aristotelian scholar’, in J. Wiesner (ed.), Aristoteles: Werk und Wirkung II , Berlin
1987, 286–311 (= Chapter 16 below).
I. Hadot, ‘Les introductions aux commentaires exégétiques chez les auteurs néoplatoniciens
et les auteurs chrétiens’, in M. Tardieu (ed.), Les règles de l’interprétation, Paris 1987,
99–122.
P. Hadot, ‘Un fragment du commentaire perdu de Boèce sur les Catégories’, in Arch. d’hist.
doctr. et litt. du moyen âge 26, 1959, 27 n. 74.

84
This illustrates a seasoned historian’s judgment that ‘ancient philosophies, rediscovered, are found
to possess a disturbing vitality, even in modern times’ (Hugh Trevor-Roper, Catholics, Anglicans and
Puritans, 1987, vii).
402 Aristotle Transformed

W. & M. Kneale, The Development of Logic, Oxford 1962, 189.

J.C. Magee, ‘The Boethian wheels of fortune and fate’, in Mediaeval Studies 49, 1987,
525 n. 7.
J. Marenbon, Early Mediaeval Philosophy, London 1983, 29.
L. Minio-Paluello, ‘Boezio: opere logiche’, in Diz. biogr. degli italiani, Roma 1970;
‘Boethius’, in Dict. of Scient. Biogr., New York 1970.
C. Mohrmann, introd. Boezio Consolazione, tr. O. Dallera, Milan 1976, 16.
P. Moraux, Der Aristotelismus I, Berlin 1973,100 n. 12.
L. Obertello, Severino Boezio, Genoa 1974.
A. Pattin (ed.), Simplicius in Praedicamenta, tr. Moerbeke, Louvain 1971, xi.
L.M. de Rijk, ‘On the chronology of Boethius’ works on logic’, in Vivarium 2, 1964, 1–49,
125–62.
R.W. Sharples, ‘Alexander of Aphrodisias De Fato: some parallels’, in Classical Quarterly
28, 1978, 258 n. 158.
J. Shiel, ‘Boethius and Andronicus of Rhodes’, in Vigiliae Christianae 11, 1957, 179–85;
‘Boethius’ commentaries on Aristotle’, MRS 4, 1958, 217–44, reprinted with
postscript in M. Fuhrmann and J. Gruber (eds), Boethius, Darmstadt 1984, 155–83,
and extensively revised in this chapter; ‘Boethius and Eudemus’, in Vivarium 12,
1974, 14–17; ‘A recent discovery: Boethius’ notes on the Prior Analytics’, in Vivarium
20, 1982, 128–41; ‘A set of Greek reference signs’, in Scriptorium 38, 1984; review of
Chadwick (1981) in Cathol. Hist. Rev. 70, 1984, 117–18; ‘Aristoteles Latinus III :
supplementa’, in Bull. phil. médiév. 26, 1984, 119–26; ‘The Greek copy of Porphyry’s
Isagoge used by Boethius’, in J. Wiesner (ed.), Aristoteles: Werk und Wirkung II , Berlin
1987, 312–40.
E. Stump, ‘Boethius’ works on the Topics’, in Vivarium 12, 1974, 77–93.
L. Tarán, Anon. Comm. on Arist. De Interp., Meisenheim 1978, viii n. 10.
C.J. de Vogel, ‘Boethiana’, in Vivarium 9, 1971, 49–66.
 16

Boethius as an Aristotelian commentator*

Sten Ebbesen

By the end of the fifth century ad Greek philosophy had for centuries been
dominated by Aristotelo-Platonic scholasticism.1 The scholastic attitude is well
expressed by Ammonius, who says that it is with good reason that earlier
philosophers have established the tradition of introducing the reading of an
authoritative text with an inquiry into its aim (skopos), utility and authenticity; for
(1) if you do not know the aim, you are likely to drop the book before reaching the
end of it; (2) if you do not know what it is useful for, you are not likely to embark
with enthusiasm on the reading; and (3) even after having been told what it is
good for, ‘we are apt to doubt the utility of a book until we are assured that it is a
genuine work of a classical author known to be generally esteemed, like Aristotle
and Plato – for as far as they are concerned, we assume that whatever they have
said is useful.’2 In periods when the scholastic attitude to the auctores prevails, the
exegesis of the classics of philosophy acquires great importance. By the year 500 a
considerable number of commentaries and auxiliary treatises relating to the works

* Editions of Boethius’ works are as follows. Inst. Arith.: Anicii Manlii Torquati Severini Boetii De
institutione arithmetica . . . (G. Friedlein) Leipzig 1867. in Isag.1 (= ed. prima) and in Isag.2: Anicii
Manlii Severini Boethii In Isagogen Porphyrii commenta . . . (S. Brandt) (= CSEL 48), Vienna and
Leipzig 1906. in Cat.: Patrologia Latina (Migne) 64. in Int.: Anicii Manlii Severini Boetii commentarii
in librum Aristotelis peri hermêneias (C. Meiser), Leipzig 1877–80 (1 = ed. prima in vol. 1; 2 in
vol. 2). Introductio ad Syllogismos Categoricos, de Syllogismo Categorico, de Divisione, in Topica
Ciceronis, de Differentiis Topicis: Patrologia Latina 64. Hyp. Syll.: A.M. Severino Boezio, de
Hypotheticis Syllogismis, L. Obertello (ed.) (= Istituto di filosofia dell’ Università di Parma, Logicalia
1), Brescia 1969. Translations of Porphyry and Aristotle: Aristoteles Latinus 1–6. As a supplement to
Migne’s editions of in Cat. and Syll. Cat. I have used MS Thott 166–168, 20, Royal Library,
Copenhagen (tenth century).
1
On ancient scholasticism, see S. Ebbesen, Commentators and Commentaries on Aristotle’s Sophistici
Elenchi, Leiden 1981, 1.52ff (= Corpus Latinum Commentariorum in Aristotelem Graecorum
(CLCAG ), 7. 1; id., ‘Ancient scholastic logic as the source of medieval scholastic logic’, in The
Cambridge History of Later Medieval Philosophy, N. Kretzmann. A. Kenny and J. Pinborg (eds)
Cambridge 1982, 101ff.
2
Amm. in Isag. 21.

403
404 Aristotle Transformed

of Aristotle and Plato had been written in Greek. Quite a few would be available to
anyone with the will and the financial means to obtain them. But in spite of an
attempt in the fourth century to produce a Latin scholastic library,3 philosophy
had not come to Latium. It had to wait for Boethius. How little his predecessors
had achieved may be gauged from the fact that he has no standard Latin equivalents
for several elementary Greek terms. To render organon (instrument) he feels
obliged to use a hendiadys, ferramentum et quodammodo supellex.4 For phantasia
he uses visum. The term had been coined by Cicero five hundred years earlier, but
Boethius introduces it in a way which shows that he expects his readers to be as
unfamiliar with it as were Cicero’s contemporaries.5 To benefit from Plato’s and
Aristotle’s useful writings a mastery of Greek was still required.
Boethius came to see it as his mission to bring Latin philosophical literature
up to contemporary Greek standards. After an initial attempt to build on the
foundations laid in the fourth century,6 he decided he had to start from scratch,
translating the auctores and then adding the necessary auxiliary works. At its
most ambitious, his plan seems to have comprised:7

(a) a complete set of basic texts, viz. Porphyry’s Isagoge, the whole of Aristotle,
Cicero’s Topics (which, of course, need not be translated), the whole of Plato;
(b) elementary commentaries on each of the basic texts;
(c) in some cases, at least, also a more comprehensive commentary;
(d) in at least one case, also a paraphrase;8
(e) supplementary monographs, including one demonstrating the
compatibility of Aristotelian and Platonic philosophy.

3
Texts relating to the Organon: Isagoge (free translation) + commentary + monograph on definition,
all by Victorinus; paraphrase of Cat. (= ps.-Augustine, Categoriae Decem); paraphrases of An. Pr.
and An. Post. by Themistius, translated by Praetextatus; monograph on hypothetical syllogisms by
Victorinus. It is very doubtful whether there was a translation or paraphrase of Int.; possibly
Apuleius’ Peri Herm. was used. Instead of Aristotle’s Topics, Cicero’s was read; Victorinus composed
a commentary on it. Cf. on these matters, P. Hadot, Marius Victorinus, Paris 1971.
4
Boeth. in Isag.1 10; cf. in Isag.2 93.
5
Boeth. in Isag.2 25: the Greeks call them phantasiai, we may call them visa. Cicero Acad. 1 40: which
they call phantasia we may call it visum.
6
in Isag. ed. prima is keyed to Victorinus’ version of the Isagoge.
7
The basic source is Boethius in Int.2 79–80. See further L.M. de Rijk, ‘On the chronology of Boethius’
works on Logic’, Vivarium 2, 1964, 1–49 and 125–62; L. Obertello, Severino Boezio, Genova 1974 (=
Accademia Ligure di Scienze e Lettere, Collana di monografie I), I, 157ff.
8
Boeth. in Int.2 251: ‘After these twin commentaries we make [sic! one expects: shall make] a sort of
compendium of this book, in such a way that we sometimes, indeed most times, use Aristotle’s own
words, except that when he obscures a matter by brevity of diction, we employ additions to make
the presentation more lucid, so that our style strikes a balance between the brevity of the text and
the diffuseness of a commentary by compressing the diffusely phrased and expanding the too
compressedly written – but that is for later.’ To me this looks like a description of a paraphrase, but
there have been other suggestions. See De Rijk, op. cit., 37–8.
 Boethius as an Aristotelian commentator 405

Death prevented him from fulfilling his plan, but he managed to translate
the whole Organon (with the possible exception of the Posterior Analytics) and
to produce several commentaries and monographs. Most of these works are
still extant.9
The translations of the basic texts are extremely faithful to the originals which
are rendered word by word and morpheme by morpheme10 with a supreme
contempt for normal Latin sentence structure.11 The choice of this procedure
was very deliberate. Boethius was a consummate master of Latin prose, but he
wanted his readers to see the real thing.12 The auxiliary works would provide the
necessary amount of comment and paraphrase. His competence as a translator
grew with time, but he never relinquished the principle of strict fidelity to the
auctor. For this reason the translations tell us very little about his interpretation
of Aristotle.
The commentaries and monographs are more informative, though it is
difficult to evaluate their testimony because almost everything in them comes
from a Greek source. Boethius himself makes no secret of the fact, which
can also be established by means of comparisons with extant Greek works.
Consequently, a mere recitation of Boethius’ words is no sufficient answer to the
question, ‘What did Boethius think about problem x?’ It must also be shown that
his use of just those words reflects a choice, i.e. that several sources were available
to him or that he sometimes chose to modify his source instead of copying it
without change. Thus the fact that he repeatedly stresses that each word has its
meaning thanks to a human decision and not by nature13 cannot be taken to
show that this is a problem he has given personal thought to unless we can
assume that he could have presented another view. If all his sources are lost, we
can never establish such a thing. And this is exactly what has been argued in
recent times. It has been suggested that the only material at Boethius’ disposal
was a copy of the Organon with marginal scholia, and that this collection of

9
See n* above. Surveys in The Cambridge History of Later Medieval Philosophy, Cambridge 1982, 53f,
74ff, 105. Fundamental studies in L. Minio-Paluello, Opuscula, Amsterdam 1972.
10
Thus proba-re = axioun, proba-mentum = axioma, see Aristoteles Latinus 6.1–3, Leiden and Brussels
1975, 121.
11
Thus the genitive absolute khrêsimês ousês tês toutôn theôrias (Porph. Isag. 1,6) becomes an ablative
absolute without participle: utili hac istarum rerum speculatione (Aristoteles latinus 1.6–7: 5,5–6).
Also in the commentaries there are occasional clumsy renditions of Greek phrases, as in Int2 13:
quae sint negationes cum modo propositionum . . . considerator poterit diligenter agnoscere; ‘cum
modo propositiones’ renders hai meta tropou protaseis (cf. Amm. in Int. 8,19; 14,11; 213,33f.).
12
cf. Boeth. in Isag.2 135.
13
Thus in Int.2 23; 54–6; 92–4; Intr. Syll. Cat. 763A; Syll. Cat. 795A; Divis. 886C. cf. S. Ebbesen, CLCAG
1,177.
406 Aristotle Transformed

scholia is no longer extant.14 We may often be able to ascertain the remoter origin
of one of the scholia Boethius knew, but we shall never know whether he deviated
from his direct source in any way and the standard answer to the question ‘Why
does Boethius say this?’ can only be, ‘Because it was in his only source.’
The ‘one source – no thinking’ theory has the support of eminent scholars and
it cannot be refuted by any means that I can think of. But neither can it be proved
by any conceivable means short of finding the supposed manuscript of the
Organon with the marginal scholia. To my mind, the circumstantial evidence
in favour of this theory, though not negligible, is less than convincing.15 The

14
J. Shiel, ‘Boethius’ Commentaries on Aristotle’: Medieval and Renaissance Studies 4, 1958, 217–44,
extensively revised in Chapter 15; id., ‘Boethius and Eudemus’, Vivarium 12, 1974, 14–17; id., ‘A
recent discovery: Boethius’ notes on the Prior Analytics’, Vivarium 20, 1982, 128–41.
15
The case for the marginalia and against Boethius’ using Porphyry’s commentary on the Categories
rests on the following assumptions and arguments:
(1) Boethius’ Greek copy of the Organon is likely to have looked much like Arethas’ from c. 900
(Vat. Urb. 35), which contains rather ample marginalia. This is a doubtful assumption. Marginal
scholia were known in late antiquity, but papyrological evidence does not suggest that a format like
that of Urb. 35 was common then. By contrast, pillaging ancient sources to compose marginal
scholia was no unusual occupation for Arethas and his contemporaries.
(2) Boethius himself says that he uses marginalia. For the meaning of in Int.2 250,20–3 ‘est quidem
libri huius . . . obscura orationis series, obscurissimis adiecta sententiis’ is that the book has ‘an obscure
course of arguments with highly obscure notes added to it’. As far as I can see, the passage means ‘the
doctrines of this book are very obscure, and on top of that the manner of presentation is obscure’.
(3) Boethius used one book only for his source. His commentary on the Categories contains a lot
of Porphyrian, but also some un-Porphyrian material. Therefore he used one book consisting for
the most part of extracts from Porphyry but with some additions from later sources. I see no
compelling reason to grant the first premise, though I admit that he is unlikely to have consulted a
vast literature before writing his commentaries. It may be mentioned that Shiel’s guess that all
Boethius knew of Themistius’ paraphrase of the Topics was a diagram has been disproved by
E. Stump, ‘Boethius’ works on the Topics’, Vivarium 12, 1974, 77–93; cf. S. Ebbesen, CLCAG 7.1,118f.
(4) Boethius was omnivorous. Yet he does not reproduce all of Porphyry’s comments on the
Categories. Ergo he did not possess a complete copy of that work. The passages which Shiel adduces
in support of the first premise (in ‘Boethius’ Commentaries . . .’, 233 and 237, see n. 14 above) cannot
bear the weight of proof. The crucial one is in Isag.2 135: ‘Therefore I think I have accomplished a
great deal if thanks to my rendering the books of philosophy in Latin in a complete and
unadulterated translation, nothing from the Greek writings is felt to be wanting.’ When seen on the
background of the preceding lines it becomes clear that in this place Boethius is speaking about a
complete and unadulterated translation of Porphyry’s Isagoge (and other basic texts), not of scholia.
(5) One medieval Latin manuscript contains An. Pr. with scholia, both translated from the Greek
by Boethius. This MS may be taken to mirror Boethius’ Greek MS . There are strong stylistic reasons
for attributing the translation of the scholia to Boethius, but also some arguments against doing so.
See S. Ebbesen, ‘Analyzing syllogisms, or Anonymus Aurelianensis III – the (presumably) earliest
extant Latin commentary on the Prior Analytics, and its Greek model’, Cahiers de l’Institut du
Moyen-Age Grec et Latin (CIMAGL ) 37, University of Copenhagen 1981, 1–10. Yet, even granted
that the translator of the scholia was Boethius, it does not follow that the medieval manuscript is a
mirror image of one produced by Boethius which again was a mirror image of his Greek manuscript.
In fact, there is good reason to believe that the scholia in the medieval MS are just excerpts from a
larger collection of scholia, possibly a whole commentary, which had been translated from the
Greek and was available to some learned men in the twelfth century. See Ebbesen, op. cit. (this note,
above). So, while the scholia may be remnants of material which Boethius gathered with a view to
composing a commentary on An. Pr., they cannot tell us how much Greek exegetic material relevant
to the text Boethius had access to.
 Boethius as an Aristotelian commentator 407

observable facts are quite as easily explained on the assumption that Boethius
had access to several Greek monographs and commentaries and that he followed
the common practice of using for each work one main source while also
exploiting secondary sources. It is an old discovery that this hypothesis works
well in the case of the extant short commentary on the Categories, the only case
in which we still have what may be the main source. Boethius acknowledges
a debt to Porphyry16 and actually keeps so close to the latter’s extant minor
commentary on the Categories (CAG 4, 1) that it is simpler to assume that he had
direct access to a complete copy of it than to assume second-hand acquaintance
by way of a book which also contained the post-Porphyrian material detectable
in Boethius’ commentary.
Granted that Boethius’ main source was Porphyry’s extant work, we can begin
to examine the way he used it. As it turns out, he follows his predecessor to the
extent of reproducing most of the questions he raised and the answers he gave,
but not to the extent of reproducing long segments of his text in direct translation.
Boethius expanded arguments which he found too compressed while curtailing
or suppressing other passages.17 In fact, he followed the procedure which his
own remarks in this and other works indicate18 – and that procedure involved
making choices. It looks as if it might be worth while to speculate about his
possible motives for choosing as he did.
First of all: why Porphyry? It seems that Boethius turned to Porphyry not
only when composing the commentary on the Categories but whenever there
was a work of his to turn to.19 If Boethius’ choice was not determined by
unavailability of other books, on what was it grounded? I suggest that it was

16
Boeth. in Cat. 160A; see n. 20 below.
17
Expansion: e.g. Boeth. in Cat. 201D–202B vs. Porph. in Cat. 100,12–16; Boeth. 240C–D vs. Porph.
128,13–15. Curtailment: e.g. Boeth. 180C vs. Porph. 86,20–32. Suppression: e.g. Porph. 55,3–56,13;
100,23–7. Cf. J. Shiel, ‘Boethius’ Commentaries . . .’ (n. 14 above). For the work as a whole, expansion
more than neutralises contraction. PL 64, cols 159–26, which correspond to CAG 4.1, pp. 55–142,
would require some 125 CAG pages. Boethius’ in Cat. is comparable in bulk to Philoponus’ comm.
in Cat., and Boethius’ in Int.2 to Ammonius’ in Int.
18
e.g. Boeth. in Cat. 159A; Syll. Cat. 793C; Introd. Syll. Cat. 761C. Boethius uses conventional
phraseology when saying that he will clarify, expand and abbreviate. Cf., e.g., Apollonius Dyscolus
de Pronomine (= GG 2.1.1) 3; Themistius in An. Post. 1–2. But this does not imply that he does not
mean what he says.
19
Porphyry was certainly used for the commentaries on Int., almost certainly for Syll. Cat. (and hence
for Intr. Cat. Syll.) and Divis.; probably also for Hyp. Syll. The planned work on the concord between
Plato and Aristotle presumably was to draw on Porphyry’s book on the same subject. For the Topics,
Boethius turned to Themistius. But then there is no evidence that Porphyry ever commented on
that work. For the Isagoge, Boethius obviously had to turn to a non-Porphyrian commentary; hence
the many un-Porphyrian features of his in Isag.1 and in Isag.2
408 Aristotle Transformed

grounded on the concern for pedagogy to which he often gives expression20 – a

concern which he shared with Porphyry.
It was the general belief of the scholastics that Aristotle himself was
consciously pedagogical and that this is apparent from the structure of the
Organon.21 The books about demonstrative (and, some would add, dialectic)
reasoning, i.e. An. Post. (and Top.), form the core of Aristotle’s work. But
pedagogical considerations motivated him to introduce these books with a
treatise on syllogistic reasoning as such (An. Pr.); and this again with a book on
the proximate constituent parts of syllogisms, viz. propositions (Int.); and this
book again with a treatise on the proximate constituent parts of propositions,
i.e. words (Cat.). This, the most elementary book, was intended to be read first of
all – though the scholastics themselves preferred to warm up with Porphyry’s
Introduction to the Categories (Isagoge).
When scholasticism was in its infancy, several philosophers had subjected the
Categories to unkind scrutiny, pointing out absurd doctrines, important omissions,
internal inconsistencies and inconsistency with other Aristotelian writings. To a
large extent this criticism determined the scholastic exegesis. The problems dealt
with in the commentaries are very often those raised by the early critics. Many of
the supposed deficiencies of the Categories could be explained away by showing
that the criticism rested on a wrong interpretation of the text. But not all difficulties
could be coped with in that way. Porphyry found a way to interpret the vices
as virtues by appealing to Aristotle’s concern for pedagogy. Thus the fact that
Aristotle does not treat of quality immediately after substance but inserted the
chapters on quantity and relation between substance and quality might seem odd
both from a Stoic and from a Platonist point of view. Besides possible ontological
reasons for Aristotle’s order, Porphyry mentions that (1) there are more common
characteristics between substance and quantity than between substance and
quality; (2) in dealing with quantity Aristotle introduced the notions of big and

20
Thus in Isag.2; 161 and 167; in Cat. 159A, 160A–B (see text in n. 53), 250C, 289C; Syll. Cat.
793C–794C (for better text than Migne’s, see De Rijk, op. cit (n. 7) 30 and n. 57 below); Intr. Syll. Cat.
761C–762C; Divis. 877A. Notice the motives for following Porphyry that Boethius himself gives: in
Cat. 160A: ‘This is a suitable interpretation for purposes of introduction and simple exposition: and
for now, at least, this is the one we have decided to propound, in dependence on Porphyry, because
he [or: because his interpretation] is the least complicated and the plainest’. in Int.2 7: ‘The book is
entitled “On Interpretation”. In composing our exposition of it we have drawn on Porphyry for the
most part, though also on the rest, and have rendered [the borrowed material] in Latin. For in our
opinion this expositor stands out both for keenness of insight and for [fine] presentation of the
points in question’.
21
Boeth. in Isag.1 12–15; in Cat. 161B–D; in Int.2 12 (cf. the scholia in CAG 4.5, p. xl). The
Greek sources are numerous. See e.g. Simplicius in Cat. 14–15 & 18,14–16; Philoponus in. An. Post.
1–2.
 Boethius as an Aristotelian commentator 409

small, which are relative terms; so, in order not to leave the reader in suspense,
Aristotle decided to round off with a chapter on relation before going on to quality.
Porphyry thus assigns a pedagogical motive to Aristotle, and Boethius accepts the
explanation.22 He also follows Porphyry in the conviction that lack of agreement
between the doctrine of the Categories and that of other books may be explained
as due to pedagogical considerations. Thus they both say that when Aristotle ends
his list of types of quality with the remark that there may be more than the ones
in the list, this is not a sign of uncertainty, nor can he be blamed for not giving the
better list that he has in the Metaphysics. The reason why he does not give it is that
he does not wish to perplex the freshmen by introducing intricacies which they
need not understand before they have reached a higher level.23 Just in case anybody
should be dissatisfied with this explanation, Boethius adds another one, this time
not from Porphyry, according to which Aristotle said, ‘And there may be other
sorts than the ones I have listed’ because he wanted to incite the reader to start
thinking on his own instead of just being a passive recipient of learning.24 Here,
too, Boethius depicts Aristotle as a conscious pedagogue.
Porphyry held that Aristotle was also pedagogical in a philosophically more
interesting way. The first book of the Organon, he held, is not a book on what
there is. It is about an elementary matter, the rational classification of the words
of the vocabulary which reflects our awareness of the sensible world. The
categories with which Aristotle deals should be understood as widest predicates,
‘predicates’ here meaning words which we may apply to a thing, and primarily
words which may function as names of sensible things. We do by abstraction
form notions of (post rem) universals, but to say that a categorial common name
like ‘man’ signifies a universal Man just means that there are some particular
sensible things about which we may say, ‘This is a man’.25

22
(1) Porph. in Cat. 100,21–3 and Boeth. in Cat. 202C. (2) Porph. 111,12–15 and Boeth. 217A; Porph.
127,3–7 and Boeth. 239A. A similar defence of Aristotle in Porph. 59,34–60,10 and Boeth. 163B–C.
23
Porph. in Cat. 134,25–9; Boeth. in Cat. 252B. A similar defence of Aristotle in Boeth. in Cat. 289C;
cf. Simplicius in Cat. 427–8; Porphyry not available for comparison.
24
Boeth. in Cat. 252C: ‘There is also a further reason: to make us search for more qualities in order
that he should be no mere teacher of his own [discoveries], but also one who incites us to find out
something of our own.’ Cf. Amm. in Cat. 88,20–2; Philop. in Cat. 156,8–11. Amm. and Philop. do
not mention Porphyry’s explanation. Simplicius in Cat. 264,2–4 has it, but does not mention the
explanation of Amm. and Philop. Cf. Boeth. 294A–B where he uses a similar explanation, also found
in Philop. in Cat. 205,26–8. Porph. and Amm. are not available for comparison in this place.
25
See, in particular, Porph. in Cat. 56–8 and 90–1. Read 56,8–13 as follows: ‘Every simple significative
expression is called a “katêgoria” when it is spoken (agoreuthêi) and said of (kata) the thing signified.
For example, let the thing be this particular stone which we feel or see: when we say about it that it
is a stone, the expression “stone” is <a katêgoria, whereas the thing is> a katêgorêma. For it signifies
such a thing and is spoken of (agoreuetai kata) the particular thing that is the stone.’ Cf. Simpl. in
Cat. 11,2–3 and 17,5–7. Boethius does not reproduce the two crucial passages on the meaning of
410 Aristotle Transformed

Porphyry had predecessors in the endeavour to make the Categories acceptable

to Platonists by restricting its scope to the sensible world. But he was particularly
insistent on the point, and Boethius followed him in this.
Critics of Aristotle had observed that Aristotle stood the world on its head by
calling the sensible particulars primary substances. Porphyry would not deny
that if Aristotle were doing basic metaphysics, this would indeed be standing the
world on its head. But, he held, the point is that in the Categories Aristotle is not
working on such a deep level. On the level on which he is working, it is perfectly
reasonable to call sensible particulars primary substances while calling secondary
the abstract universals whose only being consists in being predicable of the first
ones, and to leave such intelligible entities as God or Mind totally out of account.
Boethius agrees.26 That he did not only copy but also understand this Porphyrian
view may be inferred from the fact that he manages to say the same things as his
source without reproducing it verbatim. He even manages to improve upon the
master’s formulations. Thus it is Boethius who says that ‘secondary substances do
not have their being except in being predicable’ and that ‘what causes secondary
substances to be is nothing but their predicability of primary substances’.
Porphyry uses a less pregnant formulation, saying that the particular animals
and men are the causes of being for the commonly predicated ones.27
So Aristotle was right in abstaining from deeper matters when introducing
people to philosophy. Porphyry extended the principle of abstention from deeper
matters and concern for the novices from being a principle of interpreting Aristotle’s
logic into a rule guiding his own activity as a commentator – at least when he was
writing for beginners, as was the case when he composed his Introduction to the

katêgoria, katêgorêma, katêgoreisthai (Porph. in Cat. 56,6–13 and 58,16–18). Yet cf. Boeth. in Cat.
162D: ‘Since there are ten primary sorts of things, there also had to be ten simple words that might
be said of the things underlying them; for whatever signifies is said of the thing it signifies.’ Syll. Hyp.
p. 210 (Obertello): ‘The categorical proposition, then, [viz. “man is an animal”] makes us realise that
an actual man (ipse homo) is an animal, i.e. may receive the name of “animal”. . . . So the categorical
proposition is a declaration to the effect that the thing which it puts in the role of subject is a
recipient of the name of the thing predicated’. Introd. Syll. Cat. 768C: ‘the subject is that which
receives the expression of the predicate.’ See also Boeth. in Cat. 243C where he is eager to make
it clear that the fact that genus is entailed by species means that: ‘Wherever there is a species, the
name of its genus also turns up without delay; but where a genus is, the word for some species does
not necessarily follow.’ The words immediately before strongly resemble Iamblichus, ap. Simpl.
in Cat. 230,32–231,1; but ubicumque-sequitur is not paralleled there. Porphyry is not available
for comparison.
26
Porph. in Cat. 90–1. Boethius in Cat. 183–4.
27
Boeth. in Cat. 185D: ‘Secondary substances are such that whereas they are not in a subject they are
said of a subject. Hence they do not possess their being except by being predicated of something.
Now, secondary substances are predicated of primary ones. Hence, what causes secondary substances
to be is their predicability (praedicatio) of primary substances.’ Porph. in Cat. 90,33–91, 1: ‘The
particular animals . . . which in fact are the causes of being for the commonly predicated ones.’
 Boethius as an Aristotelian commentator 411

Categories and the extant commentary on the same work. On the whole, he remains
faithful to the programme announced in the beginning of the Introduction: to give
a concise account of Peripatetic lore from a logical point of view, abstaining from
the deeper issues, such as the mode of being of universals, keeping in mind that he
is writing an introduction.28 Though Boethius, when commenting on this passage,
could not quite keep himself from telling what Porphyry abstained from saying,29
he elsewhere makes the Porphyrian programme his own.
In fact, Porphyry tried to make Aristotelian logic function on a minimum of
assumptions, notably these:30 (1) that man can recognise certain individuals in
the sensible world; (2) that man can name these individuals; (3) that man can
recognise a similarity between several individuals; (4) that man can decide to
use a certain sound as a name of any individual which is similar to others in
respect of a certain feature; (5) that man can recognise a similarity between
certain names; (6) that man can decide to use a certain sound to name any
individual name which is similar to others in respect of a certain feature.
These are the basic assumptions needed for an interpretation of the Categories.
They allow for the creation of our primary object-language with its universal
names, the classification of these names in ten categories, and for the creation of
a second-order language, which is the subject of de Interpretatione and which is
necessary for talking about the first-order language which is the subject of the
Categories.
Some more assumptions are needed to deal with propositions. These are
basic: (7) that any two men confronted with the same sense-data will isolate the
same individuals and recognise the same similarities; (8) that the thoughts or
‘concepts’ to which these ‘recognitions’ give rise can be stored and recalled to
consciousness when required; (9) that simple concepts can be combined into
complex ones; (10) that man can invent words indicating the internal syntax
of complex concepts; (11) that one man can teach another to use words in
the same way as himself; (12) that by uttering the words associated with the
simple or complex concept presently actualised in him one man can actualise
the same concept in a fellow-man who speaks the same language. These further
assumptions allow men to communicate with each other by means of propositions
analysable into subject, predicate and their mutual relations, without the need of
pointing at the object of discourse.

28
Porph. Isag. 1.
29
Boeth. in Isag.1 24ff, in Isag.2 159ff.
30
cf. S. Ebbesen, CLCAG, 1,141ff. It is pertinent to notice that Boeth. in Int.2 is one of the really
important sources for Porphyry’s thought.
412 Aristotle Transformed

Of course, this is not the whole story, but Porphyry did his best to explain
Aristotelian logic without a host of other assumptions, and he tried to be a good
pedagogue by not introducing an assumption before it was needed. Thus, in spite
of the crucial role of concepts in his Aristotelian semantics, he did not formally
introduce concepts in the little commentary on the Categories. Concepts could
wait until the reader got to the de Interpretatione. Boethius followed him in this.
But he did more than just follow Porphyry.
Boethius demonstrates his understanding of Porphyry’s thought by moving
the presentation of the fundamental theory about the origin of our language to
the very beginning of his commentary on the Categories. Porphyry operated
with two labellings of things, two ‘impositions of names’.31 By the first imposition
man has created his object-language, by the second his meta-language. Using
and presenting the model requires some delicacy. For instance, it may be wise
to de-personalise the first impositor as much as possible so as not to end up
with one of the following ridiculous scenarios (with which some people have
ended up):

(a) One wise impositor at some time decided what things ought to be called.
(b) A congress of terminologists convened to invent language. Having done so,
they went home to meet again at a later time in order to perform the
second imposition.32

Boethius avoids these absurdities. Porphyry had de-personalised the

impositor by calling him ‘man himself ’, not ‘some man’.33 Boethius achieves the
same by saying that only the race of men could put names on things and that
the human mind applied words to the things.34 On the other hand, it is equally
absurd to let an abstract mind perform the naming. Some particular man must
have been the first to say ‘man’. So it is reasonable enough when Boethius – not
quoting Porphyry – speaks of the man who first said ‘man’.35 It is less reasonable
that he sometimes (not in the commentary on the Categories) suggests that the
imposition must have resembled a christening ceremony, the impositor saying,
‘Let this be called a man’.36 Apart from the fact that experience shows that there

31
Porph. in Cat. 57–8.
32
See S. Ebbesen, op. cit., 1,178.
33
Porph. in Cat. 57,21.
34
Boeth. in Cat. 159A. Cf. in Int2 55,1–7.
35
ibid. 183D.
36
Boeth. in Int.1 46; in Int.2 64; cf. next note. Boethius may have had Greek models for the christening
ceremony. Cf. Anonymous Commentary on Aristotle’s De Interpretatione (Codex Parisinus Graecus
2064), L. Tarán (ed.), Meisenheim 1978, 3 (= Beiträge zur klass. Philol. 95).
 Boethius as an Aristotelian commentator 413

are other ways of introducing new words, the picture Boethius paints of the
imposition ceremony is based on the unnecessarily strong and unconvincing
claim that all names start as proper names. The model of the first imposition is
designed to show that our use of generic names is dependent on a confrontation
with the sensible particularised world. To avoid solipsism, the imposition must
in some way be a public event, but there is no need to deny the possibility of a
telescoped procedure, so that the impositor can say, ‘Let such a thing be called a
horse’. Boethius himself may on second thoughts have found the christening
ceremony unconvincing. In the first version of his opuscule on categorical
syllogisms, the Aristotelian claim that the nominative is more truly a name than
the oblique cases of nouns is defended with the remark that the man who first
introduced the word circus probably said, ‘Let this be called circus’.37 In the
revised version, in which he does much to eradicate crude expression and
thought, this explanation has been dropped. Instead we find the rather more
sophisticated remark that since we start by sensing things which are present
to us, and it is clear that men have assigned their words to the things which
we perceive with a sensation in the present, it is reasonable to say that
properly speaking verbs always have present signification, so that Aristotle’s
characterisation of past and future verb forms as non-verbs is justified.38
On the whole, then, Boethius’ account of the theory of imposition is good.
Moreover, he puts the notion of second imposition words, or names of names, to
use in a place in which his main source did not. Discussing the transitivity of
predication he has to deal with the example ‘Man is a species, Socrates is a man,
ergo Socrates is a species.’ Boethius solves the paralogism by pointing out that
‘species’ does not occur in the definition of man, and so ‘man is a species’ cannot
be a predication ‘in quid’. No, he says, the fact is that in this proposition the
function of ‘species’ is to indicate that ‘man’ is predicable of individuals only ‘and
[“species”] is, in a sense, a name of names’.39 That is, the proposition may be
paraphrased as ‘the word “man” is predicable of individuals only’, and ‘species’ is
the name of those names which are predicable of individuals only. Boethius
owes this solution of the paralogism to a Greek thinker,40 but not to Porphyry’s
minor commentary on the Categories. His departure from the usual source

37
Boeth. Syll. Cat. 796A; cf. preceding note.
38
Boeth., Intr. Syll. Cat. 765D–766A. Notice that in Intr. Syll. Cat. Boethius is less dependent on his
Greek source(s) than in Syll. Cat. Thus he expands the sections on the parts of speech, 796C–D, so
as to present the whole Latin inventory of eight parts, one being interjections; see 766B–C.
39
Boeth. in Cat. 176D–177A.
40
Possibly Iamblichus. See Dexippus in Cat. 26; cf. S. Ebbesen, op. cit., 1, 231.
414 Aristotle Transformed

probably means that he found the explanation attractive – as indeed it is in the

theoretical framework within which he is working.
The cautious formulation that genus and species are ‘in a sense (quodammodo)’
names of names may indicate his own and/or his source’s reluctance to call
‘species’ a name of names because that expression was primarily used of the
designations of the parts of speech – noun and verb.41 But there may also have
been a more serious reason for his cautiousness.
At this point, after so many kind words about Boethius’ intelligence, I wish to
make a digression about an intricate matter which he did not manage to sort out
and which shows a limitation of his ability to use in one place what he had
learned in another context.
When on his own, Boethius was not much at ease when having to explain the
sort of relations that obtain between such non-categorial terms as ‘species’ and
such categorial terms as ‘man’. This is very clear in his two works on topical
argumentation (in Topica Ciceronis, de Differentiis Topicis). For these he had
no good Greek sources but had to produce his own mixture of a Greek theory
which he knew from Themistius – an eloquent man, but not the greatest of
philosophers – and of Roman rhetorical tradition. The result is astonishing in
several respects. One cause of wonder is the total absence from these works of
the notion of names of names, although it would appear to be highly relevant
in the context, as some of Boethius’ medieval commentators saw.42 For in the
topical works he operates with second-level words or concepts (it is not quite
clear which), such as genus, species, definition. Associated with each of these
is a set of axioms, such as ‘That to which the definition of the genus does not
apply is not a species of that of which the definition in question is a definition.’
This particular axiom is supposed to bail for the soundness of the inference,
‘An animal is an animated sensible substance, a tree is not an animated sensible
substance, ergo a tree is no animal.’ Directly the argument is supposed to hinge
on the axiom, indirectly on the notion of definition. In other words, to decide the
question whether trees are animals we may construct an argument founded on
the notion of definition and on a particular instance of a definition, the definition
of animal.43 But Boethius has a worry.44 Are not definition and definiendum
really the same thing? Did not the definiendum, animal, occur as a term in the

41
See Boeth. in Cat. 159B–C, corresponding to Porph. in Cat. 57,29–58,3.
42
See N.J. Green-Pedersen, The Tradition of the Topics in the Middle Ages, Munich 1984.
43
Boeth. Diff. Top. 2, 1187A–B.
44
Boeth. in Top. Cic. 3, 1083Cff; 1091B; cf. 1,1055ff. I owe my awareness of the problems of these
passages to N.J. Green-Pedersen, op. cit.
 Boethius as an Aristotelian commentator 415

proposition we wanted to prove? Is not then our use of the definition of animal
a case of using a term to prove something about itself? To explain that nothing is
wrong Boethius uses singularly obscure language, a vice to which he was not
addicted. I think what he wants to say is that the definition ‘animate sensible
substance’ may be considered in three ways: (1) as the thing or the sign of a thing
which is really identical with animal though notionally different; (2) as a phrase;
in this sense the definition is also a thing ‘in so far as phrases are things’; (3) as a
definition; it is in this last sense that it is a locus (topos), i.e. a source of arguments
(sedes argumentorum, aphormê epikheirêmatôn), and in this sense it is not a
thing. This is suggestive, but it certainly does not attest to a clarified conception
of the semantics of non-categorial terms.
Boethius knew next to nothing about Stoic logic. Porphyry in his Aristotle
commentaries often compared Stoic terminology and tenets with Aristotle’s.
Boethius rarely bothered to include such passages in his own commentaries. He
found them both irrelevant and obscure.45 Real knowledge of Stoic logic had
disappeared long before his time, but fragments – often distorted fragments – of
Stoic theory would occur in all sorts of technical literature, not least in rhetorical
writings. I think Boethius’ confusion about the status of definitions was
occasioned or increased by his use of a source that contained some distorted
version of a Stoic distinction between corporeal and incorporeal entities, such as
words vs. their meanings; or between incorporeal somethings (tina) and nothings
(outina), such as predicates vs. universals.
Boethius’ works on topics also contain a distinction between argumentatio,
which is a proof qua (pronunciation of a) string of words, and argumentum,
which is the sense of the argumentatio.46 This must reflect a Stoic distinction
between an incorporeal epikheirêma and a corporeal epikheirêsis, comparable to
many other Stoic distinctions between corporeal entities with names in -ia or
-sis and their incorporeal effects or meanings which have names in -ma.47 The
purpose of the distinction was no longer clear to Boethius.48 Only very tentatively

45
See Boeth. in Int.2 24; 71; 201.
46
Boeth. in Top. Cic. 1,1050B; 1053A–1054A; Diff. Top. 1174D; 1183A.
47
cf. e.g. katorthôsis-katorthôrma, v. Arnim, SVF III nos. 85 and 524. ôpheleia-ôphelêma, harmatia-
harmatêma are other examples. This sort of distinction also underlies Porphyry’s between katêgoria,
katêgorêma (see n. 25 above).
48
This appears both from the fact that he makes little use of the distinction and from the discussion
in in Top. Cic. 1,1053A–1054A. The first evidence of the distinction argumentum-argumentatio is
Cicero Part. Or. 45: ‘you seem to be looking for argumentation, i.e., the unfolding of argument.’ Cf.
Cic. Inv. 1. 40,74. See also Fortunatianus Ars Rh. 2, ¶23 and 28 (= Halm, Rh. Lat. Min. 115; 118);
Victorinus in Rh. Cic. 1,29; 31; 40 (= Halm, op. cit., 231; 232; 240; 247); Cassiodorus Inst. 2, p. 105
(Mynors). I have not found the distinction between epikheirêma and epikheirêsis in any Greek
416 Aristotle Transformed

would I suggest that there may be a connection between the two detached
fragments of Stoic theory. The original idea may have been that just as physical
topos (place) is an incorporeal something, so is a topos in argumentation theory.
Its contents are other incorporeal somethings, viz. argumenta (epikheirêmata),
which constitute the meaning of argumentationes (epikheirêseis), our verbal
formulations of proof.
Boethius understood Porphyry much better than he understood the Stoics.
He understood Porphyry’s de-ontologising of logic and his economy of
assumptions so well that on occasion he refused to follow his teacher when the
master forgot his own principles.
Porphyry suggests that the order in which Aristotle deals with the principal
categories, viz. substance – quantity – relation – quality, may be due to the fact
that (1) being a body is prior to being a qualified body, and being a body implies
having dimensions, and so implies having quantity; (2) when the dimensions
pre-subsist, bigger and smaller, which are terms of a relation, accrue to them.
Boethius reproduces this explanation. But he performs a tiny change. The
statement that the relation of bigger and smaller accrues to the pre-subsisting
dimensions becomes, ‘quantity being posited we also have the relation of bigger
and smaller’.49 The removal of the ontologically loaded word ‘pre-subsisting’ may
be significant.
Porphyry sinned in an even graver way against his own principles when he
suggested that Aristotle was wrong in claiming that the known may exist prior to
the knowledge of it. Perhaps, Porphyry said, it is best not to look at knowledge as

source. But notice the quotation of Porphyry in Syrianus’ commentaries on Hermogenes, Rabe, Rh.
Gr. 16.1, p. 93 and 16.2, p. 14: ‘Speech has a soul (psukhê), as it were, and a body. The discovery of
the concepts may justly be considered the soul of speech, their externalisation (hermêneia) its body.’
Compare Boeth. Diff. Top. 1, 1174D: ‘However the argument and argumentation are not the same.
For the force (vis) of what is stated (sententia) and the structural property (ratio) which resides in
an utterance (oratio) when something in doubt is being proved is called the argument; the actual
externalisation (elocutio, = hermêneia) is called argumentation. Therefore, argument is the power
[or: virtue (virtus)] and sense [or: mind (mens, = nous or psukhê?)] of the argumentation and is
what is stated, whereas argumentation is the unfolding in utterance of the argument.’ Similarly in
Top. Cic. 1,1053B: ‘For either the externalisation and the very string of propositions . . . is called
argumentation, whereas the sense (mens) and statement (sententia) involved in the syllogism is
called argument.’ Syrianus also has a definition of epikheirêma closely related to the Ciceronian/
Boethian. Compare Boeth. Diff. Top. 1174D: ‘An argument is a ratio that provides faith in a matter
about which doubt has been raised’ with Syrianus, op. cit. 16.1,57 ‘an argument (epikheirêma) is a
logos adduced to produce faith in the question at hand’.
49
Porph. in Cat. 100; 111; 127, matched by Boeth. in Cat. 202B–C; 216D; 239A. Compare Porph.
111,9–10 ‘when length, width and depth presubsist the bigger and the smaller accrue to them, and
they are relative’, with Boeth. 216D ‘quantity being posited the bigger [maius, Migne’s text has
‘magis’] or smaller must be. Therefore, as relation immediately follows upon quantity, it was correct
to put the relatives next in order after quantity.’
 Boethius as an Aristotelian commentator 417

something residing in men, but as residing in the nature of the things that may
be known, since the eternal mind understands all things that are, whereas the
knowledge of the separate things always descends to men. So that when there is
something sensible, there is a universal sensation; and when there is something
that may be known, there is a universal knowledge, the terms of the relation thus
being simultaneous.50
Porphyry here uses the Neoplatonist device of grounding the possibility of
communication between distinct entities in a higher-level union of these distinct
entities. And he introduces the eternal mind so conspicuously absent from the
rest of his exegesis of Aristotle’s logic. Boethius deserts his main source at this
point. Only at the end of a longish defence of Aristotle’s opinion does he mention
that some, including Porphyry, disagree, and then goes on to quote Porphyry,
excusing himself for doing so with the remark that this will not take up much
space.51 But the work he quotes from is not his usual source, Porphyry’s minor
commentary on the Categories. It must be the big Categories commentary,
addressed ad Gedalium. The quotation is likely to be second-hand, but this
is immaterial. The question is: what did Boethius gain by switching source?
Apparently very little, for the quotation says much the same as the minor
commentary and there is no appreciable gain in terms of brevity. Yet he gains one
thing: he avoids introducing the Eternal Mind in a work of logic, for the extract
from the big commentary contains no mention of the Eternal Mind.
But is it really so remarkable that Boethius is Porphyrian, and sometimes a bit
more than Porphyry himself? Yes, it is. For not everybody was so. In particular
Iamblichus was not. He had introduced the two-level commentary, consisting
for one part of the pedestrian Porphyrian exposition, for another of an
‘intellectual’ interpretation of the same matters. Iamblichus had the peculiar
idea that Aristotle had plagiarised the Pythagorean Archytas when writing the
Categories, and that the true interpretation of the doctrine of categories had to
be Pythagorean. Boethius knew of the existence of Iamblichus’ commentary.52
We cannot be certain that he had read it. But if he had not, he had read
some other commentary depending on Iamblichus, for he knew the sort of
philosophising about the categories which it contained. And he liked it.
In his own commentary Boethius, after presenting Porphyry’s explanation of
what the Categories is about, adds that he has chosen to follow Porphyry because

50
Porph. in Cat. 120,33–121,3.
51
Boeth. in Cat. 233B–D.
52
See ibid. 162A; 224D and 225B.
418 Aristotle Transformed

he is so easy to understand for a beginner in philosophy. But, says Boethius, I

have in mind to write another work in which I shall treat of three questions, one
of which is what the Categories is about. The explanation that I shall give there
will be very different from the one I give here, much profounder and in
accordance with the advanced theory of the Pythagoreans.53
It is not obvious what the other two questions are that Boethius intended to
write of elsewhere. Various solutions have been proffered.54 I would like to point
out that there are two more references in Boethius’ commentary to a future work.
In one place he promises to discuss the authenticity of the work on categories
that Iamblichus believed to be by the old Pythagorean Archytas, the contemporary
of Plato.55 In another place he promises a refutation of people who criticise
Aristotle’s list of categories for being too long, too short, or badly ordered.56
These may be the two missing questions. At any rate, there is every reason to
believe that Boethius wanted to treat them in the same work as the first one.
These passages show that Boethius had a clear notion of levels of interpretation
and that his exclusion of Neoplatonic speculation from the elementary
commentaries was deliberate. But isn’t he cheating his readers by offering them
an interpretation which he himself considers only second-best? Hasn’t he a
better defence for his procedure than the claim that the profound theory is too
difficult for novices? I think he has, though he does not directly confront the
question. He does, however, in his Introduction to the Categorical Syllogisms, tell

53
ibid. 160A–B. Migne’s text is corrupt. De Rijk, op. cit. (n. 7 above) 133 quotes a Brussels MS . I here
translate the text of MS Thott 168, 20, fols 2v-3 in the Royal Library of Copenhagen. In the essentials,
there is agreement with the Brussels MS : ‘But we have in mind to discuss three questions, one of
which is what the Categories is about and there we shall enumerate the interpretations that different
people have put forward and tell which one we agree with. Nobody ought to be surprised that it will
be in conflict with the interpretation proffered here, when he sees how much profounder the new
one is so that it could not have been properly grasped by the minds of the beginners for whose
elementary instruction we have written the present work. We must, then, use an elementary
(mediocris) exposition to stimulate and, as it were, line up at the very doorstep of this discipline
those whom we prepare to admit to this interpretation. So, the reader of both works will realise that
the reason for our change of interpretation is that there the interpretation offered in the other
exposition is tuned to Pythagorean science and perfect doctrine, whereas here it is tuned to the
simple [mental] motions of beginners.’
54
De Rijk, op. cit., 137 suggests skopos, khreia, epigraphê. A scholiast in MS Thott 168, 20, fol. 2v says:
‘This should not be understood as if he had produced another exposition, for what he promises
here is what he clearly expounds in part 1 of the second edition of [his commentary on] Peri
hermeneias, where in fact his closing remark is: “For the question what this book is about and
concerning its title and the matter of its certain attribution to Aristotle, the above must suffice”.’
These last words in quotes come from Boeth. in Int.2 13,9–11. Above ‘diversorum sententiis’ the
scholiast wrote: Aspasii, Alexandri, Theophrasti, Aristotelis, Andronici. The scholiast is hardly right
in thinking that Boethius is referring to in Int2 7–13, but it is very probable that Boethius there used
some of the material he had intended for use in the promised work.
55
Boeth. in Cat. 162A.
56
ibid. 180C.
 Boethius as an Aristotelian commentator 419

his readers not to think that in teaching them logic and its way of dealing with
language he wants them to unlearn the grammar they were taught at school.
What the reader is invited to do is to consider the same matter from a new angle
which gives a profounder understanding. Looking at the constituents of speech
from a grammatical and a logical point of view may be compared to looking at a
line or a surface from a mathematical and a physical point of view. The several
disciplines which study the same object stand in no conflict. On the contrary, he
says, a true understanding of nature can only arise as a result of using more than
one approach.57
A similar attitude to the relation between logic and profounder theories
would hold that Aristotelian logic as interpreted by Porphyry is a perfectly
legitimate and useful discipline. We must have a theory of the linguistic behaviour
associated with our awareness of the sensible world, and it is an advantage if
this theory carries few ontological assumptions. The profounder theory adds
a new dimension to our understanding of the world. It does not render the
shallow theory superfluous. What was the profound work to be like? I think we
have enough information to gain an impression of what a true Pythagorean
interpretation of the categories would be for Boethius. The story would run
more or less like this:58

57
Boeth. Introd. Cat. Syll. 761D–762C. Notice that 761D Idem namque – 762C explicata cognitio is a
replacement of the original exhortation to look up Boethius’ Greek models; Syll. Cat. 794C (I quote
from MS Thott 166,20 fol. 1v: ‘And, if we are really not good enough for those people, then they had
better taste the works (commentarii) from which we have drawn this. Attracted by their agreeable
taste of subtlety those people, unbridled and untamed lovers of polemics though they are, may find
peace in accepting the unassailable authorities of the ancients. But anyone without knowledge of
the Greek language must invest an effort in [the reading of] our works or similar ones, if such there
are. Hence the law which this preface lays down is that nobody may enter our court as a spectator
if he intends to blame us because he will not understand. But in order not to waste our time with
empty prefaces, let us start as soon as we have repelled the danger lest somebody blame our style for
its sterility. We do not seek the complexitites of eloquence, we seek to make ourselves plain; if we
manage to do that, we have achieved our aim, no matter how unrefined our language may be.’
58
That, broadly speaking, ‘Pythagorean’ must mean ‘after the manner of Iamblichus’ is obvious, as
pointed out by De Rijk, op. cit., 136ff. A more precise idea of what Boethius would understand by
‘Pythagorean’ in connection with the Categories may be obtained from the following observations.
(a) Simplicius and Syrianus connect the list of exactly ten categories with Pythagorean speculation
(Archytas). See Simpl. in Cat. 13,21–3; 51,3–4; 68,22–8 (= Ps.-Archytas, Szlezák (ed.), 57); Syrianus
in Hermogenis P. staseôn (Rabe, Rh. Gr. 16.2) 58. So does Nicomachus Intr. Arithm. 2.22 (pp. 122–3
Hoche): ‘The number 10, most perfect, as the Pythagoreans think, in accordance with which the ten
relations received their quantity, as we saw a moment ago, and the ten so-called categories, and the
divisions and relations of the extremities of our hands and feet, and lots of other things.’ Boeth. Inst.
Arithm. 2.41 (p. 139 Friedl.) expands this as follows: ‘. . . because of the perfection of the number 10,
as the Pythagoreans thought. . . . Therefore it is obviously the Pythagorean 10 that we find in the
Aristotelian, and still earlier in Archytas’, division of the ten categories. Well, Plato too, who took
such interest in Pythagoras, uses the same system of division [secundum eandem disputationem
dividit; presumably, kata ton auton logon, and thus an indication that this addition to Nicomachus
comes from a Greek source], and Archytas the Pythagorean established these ten categories before
Aristotle, though some doubt this. The ten partitions of our limbs and a host of other things have
420 Aristotle Transformed

Philosophy is the love of wisdom, and this wisdom is the divine mind, which
may also be called the primordial number. It contains the multitude of intellectual
ideas but is characterised by non-distinction between mind, mental grasping
and the mentally grasped. It gives the measures of being to all things. On this
exalted level there is no need of names. But on the lower level of soul it is different.
The soul is not the things themselves. It has ‘pictures’ of the things. These pictures
are concepts or names. They exhibit the things as structured by the primordial
number which (possibly by means of oppositions like even/odd) has structured
the things in the perfect number of ten categories – ten like the fingers of the
hands and the toes of the feet. The bottom is reached when the soul’s immaterial
names are embodied in matter, in sound. Only wise men who see the Mind and
the nature of things can choose the appropriate matter for the name-forms or
meanings, and thus perform this final act in the process that creates human
language. But even wise men do not always see the same aspects of the things to
be named. They are like artists painting the same motif. Even good artists
produce different pictures. Consequently it is no wonder that different peoples
may have different words for the same thing, or even that the same people may

the same origin, but there is no need to spell it all out in detail.’ (b) Boethius’ contrasting Porphyry’s
interpretation with the profounder Pythagorean one presupposes that the latter treats the categories
as real, intelligible entities. And so it did according to Simpl. in Cat. 68,22–8; 91,14ff. (c) Simplicius
and Dexippus consider it a hall-mark of Pythagorean exegesis of the Categories to hold that there is
a natural link between things and names. See Simpl. in Cat. 13,21–6; 40,5–13; 105,3–4; Dexippus in
Cat. 16,33–17,3 (notice the kick at Iamblichus in 17,4–6). (d) Proclus’ exposition of Pythagorean
semantics in his scholia on Cratylus 16 (ed. Pasquali pp. 5–6) is very closely related to the
(Iamblichean) account of semantics in Simpl. in Cat. 12,12–13,11. That semantics does postulate the
natural link between names and things, and introduces the notion of original unity in the (Divine)
Mind. It also links up perfectly with the distinction between two senses of phusei presented in
Proclus, op. cit., 17, pp. 7–8 Pasquali and Amm. in Int. 34ff. (e) Boeth. in Int.2 21–3, actually sketches
this sort of semantics, though without attributing it to any particular school. A central passage,
22,6–13: ‘But if you go back to nature and consider it carefully, you will realise that when there is a
thing, there is also the concept of it – if not among men, at least in him who in the divinity of his
own substance is ignorant of nothing in the thing’s own nature. And, if there is a concept, there is
also a word. But if there is a word, its letters are there; they may be unknown but this does not affect
the nature of the word’. Cf. L.M. de Rijk, ‘Boèce logicien et philosophe: ses positions sémantiques et
sa métaphysique de l’être’, in L. Obertello (ed.), Congresso internazionale di studi Boeziani – Atti,
Roma 1981, 141–56. (f) From Nicomachus’ Arithm. 1 1–6, which Boethius himself had rendered in
Latin, he would know to connect the theory of divine exemplar ideas with the name of Pythagoras,
and also the ‘Pythagorean’ explanation of what philosophy is. This explanation occurs with
ascription to Pythagoras in Amm. in Isag. 9,7–23, in a list of definitions of philosophy. The
corresponding passage in Boeth. in Isag.1 7, contains no list, but just the ‘Pythagorean’ explanation,
without ascription. (g) Addressing Philosophy, Boethius in the Consolation 1 pr. 4,38–9 says: ‘Day
after day you keep dropping the Pythagorean motto, “Follow God” into my ears and my thoughts. It
was not fitting for me to seek the protection of the most abject spirits when you were preparing me
for such excellence that you might make me similar to God.’ (h) The passage on levels of cognition
in Cons. 5 pr. 4,25ff is consistent both with the ‘Pythagorean’ semantics and with Boethius’
programme of not penetrating to the profoundest level when writing about logic for beginners.
 Boethius as an Aristotelian commentator 421

have more than one name for the same thing. The various names are the
expression of the givers of names attending to different aspects of the forms of
things. Language is ‘by positing’ (thesei) in the sense that the phonic stuff
constituting the matter of each name was chosen by somebody. By nature
(phusei) in the sense that the matter was not chosen at random but in imitation
of the form to be conveyed by it. And thus there is a path leading back from
material names to concepts and thence to things themselves and the original
unity. Following that path the philosopher complies with Pythagoras’ admonition
hepou theôi (follow God).
Darkness fell over the world of learning soon after Boethius’ death. But his
writings survived and were influential in the early phases of the medieval
renaissance of philosophical studies. By keeping, for the most part, a clear
distinction between the pedestrian and the profound level of interpretation, and
by only handing the first down to posterity, Boethius greatly contributed to the
sanity of medieval logic. Contrast this passage from Iamblichus:59

Since the power of the one from which [i.e. the one] everything that has quantity
is engendered extends self-identical all through things and limits each thing by
proceeding from itself, insofar as it extends totally indivisibly all through things,
it establishes the continuous, and insofar as it exercises its procession as one,
indivisible procession without delimitation. But insofar as it makes a stop in its
procession at each of the forms and insofar as it limits each of them and makes
each of them one, insofar does it produce the discrete. In virtue of the one and
most principal cause which simultaneously comprises these two acts, it produces
the two quanta. And in virtue of its total identity everywhere, both in each of the
parts and in them all, it effects the continuous, but in virtue of the identity of
each of them to themselves and of its being in its totality in each one it engenders
the discrete. And in virtue of the mutual union of the intelligible quanta it
establishes the continuous; but in virtue of their mutually separated union it
establishes the discrete. And by virtue of its static act it creates the discrete, by
virtue of its proceeding act it creates the continuous. Since, then, it is both static
and proceeds, it engenders both. For the power of the intelligible measures
simultaneously contains both those quanta which are static and those which
proceed, in one and the same.

By not giving the early medieval logicians a Pythagorean commentary on the

Categories Boethius saved them from being led astray by such gibberish.

59
ap. Simpl. in Cat. 135,10–26.
422
 17

An unpublished funeral oration on

Anna Comnena
Robert Browning

The Byzantinist has one advantage over the student of classical antiquity – unless
the latter happens to be a papyrologist. With a little diligence and a minimum of
good luck he can easily unearth unpublished texts and find himself producing
an editio princeps. And however often one has turned over the leaves of a
manuscript and laboriously read words which have remained unread for perhaps
five centuries or more, it never loses its thrill. Yet one must admit that the
advantage is less than it seems. The classical scholar’s texts are usually worth
reading from some point of view, while what the Byzantinist finds is so often
empty rhetorical verbiage. Byzantine funeral orations are notorious for their lack
of information on the life of the deceased. Yet they never tell us absolutely
nothing if we read them alertly, and they are sometimes remarkably informative
on the ideas and values of the times. When the subject is a major figure of
medieval Greek literature about the details of whose life we are very much in the
dark, even the most trifling addition to our knowledge is welcome. It is this
thought which encourages me to present a hitherto unknown Byzantine writer of
the middle of the twelfth century – George Tornikes, Metropolitan of Ephesus –
and to dwell in particular on his funeral oration on Anna Comnena.
I shall try in my remarks to avoid involvement in technicalities of dating and
prosopography which are of exclusively Byzantine interest. George Tornikes,
Metropolitan of Ephesus, is in fact not a wholly unknown figure. Six of his letters
were published by Spyridon Lampros in 1879 in his edition of the works of
Michael Choniates, Metropolitan of Athens, to whom he wrongly supposed
these letters to be addressed. And the same scholar thirty-seven years later
described in some detail the unique manuscript of Tornikes’ works (Neos
Hellenomnemon 13, 1916, 13–22). His description has been superseded only in
the last few months by that in the magnificent first volume of the new catalogue

423
424 Aristotle Transformed

of the Vienna Greek manuscripts by Herbert Hunger. Yet the great reference
works on Byzantine literature – those of Krumbacher and Beck – do not mention
him. Historians of the period, like Chalandon and Paolo Lamma, pass him over
in silence. And even the diligent Brockhoff, in his dissertation on the history of
Ephesus in the Middle Ages (Jena 1905), fails to list him among the bishops of
that city. On the few occasions when he is mentioned, he is tacitly or explicitly
identified with a namesake who was professor of Rhetoric in the Patriarchal
School in Constantinople at the end of the twelfth century. Indeed it is to this
mistaken identity that we owe the publication of the six letters by Lampros more
than eighty years ago.
The corpus of Tornikes’ works survives in a single manuscript of the early
fourteenth century, now in the National Library in Vienna, which is a treasure-
house of unique twelfth-century texts, cod. Vindob. phil. graec. 321. The corpus
comprises twenty-five letters to named and usually easily identifiable addressees,
three prooemia to inaugural or ceremonial lectures delivered by Tornikes at the
Patriarchal School, a confession of faith made in connection with a well-known
theological controversy in the middle fifties of the twelfth century, a letter drafted
for the emperor Manuel I to Pope Alexander III , and a very long funeral oration
on Anna Comnena, with which I am now principally concerned. None of these
texts has been published except in trifling excerpts.
From his works one can gather something of Tornikes’ life and career. On his
father’s side he belonged to a wealthy Macedonian or Thracian family, allegedly
of Armenian origin, which in 1047 provided an unsuccessful claimant to the
imperial throne. His mother was a niece of an archbishop of Bulgaria, perhaps of
the great Theophylact. We hear of a brother named Leo, who was for a time in
Athens, doubtless on government service, and of a cousin Euthymius, his special
protégé, who held a junior appointment on the staff of the Patriarch. As such
posts were generally the prelude to a distinguished ecclesiastical career, we may
tentatively identify this Euthymius with Euthymius Tornikes, bishop of Patrae,
who was expelled from his see by the Latins in 1204. This prelate was a nephew
of Euthymius Malakes, bishop of Neae Patrae, theologian and scholar, and
lifelong friend of Eustathius, for whom he composed a still surviving funeral
oration. So George Tornikes may also have been related to Malakes and have
known Eustathius as a young man.
We do not know when he was born, nor much about his early life, except that
he was an intimate of Anna Comnena and her family during her long years of
retirement from public life after the death of her father Alexius Comnenus in
1118. He seems to have been a member of the literary and scholarly circle
 An unpublished funeral oration on Anna Comnena 425

gathered round her, of which I shall have more to say later. He probably also held
junior teaching posts in the Patriarchal School. At any rate in 1146–7 we find
him appointed didaskalos tou psaltêros, the junior of the three Professors of
Theology, and we have the proem to the inaugural lecture which he delivered on
that occasion. Unfortunately, unlike many inaugural lectures of Byzantine
teachers, it tells us nothing of Tornikes’ earlier career, but consists largely of
laudations of the Patriarch Cosmas II Atticus, to whom he owed his appointment.
Some years later he became didaskalos tou euangeliou and head of the Patriarchal
School, and later still, but probably before the end of 1154, he was appointed
hupomnêmatographos and so became responsible for drafting the official
documents issued by the Patriarch. Literary skill and a command of rhetoric
were essential qualifications for such a post in the twelfth century, just as they
were indispensable to the corresponding official in the imperial service, the
Keeper of the Inkstand (ho epi tou kanikleiou). At this point in his rapid career
Tornikes could reasonably look forward to an archbishopric in the provinces;
this was the usual reward of a successful career in the patriarchal School or
in the personal staff of the Patriarch. But he seems to have got on the wrong side
in the complex ecclesiastical politics of the time, and we have a number of
despondent letters from this period describing the hostility shown to his former
patron, the Patriarch Cosmas II Atticus, whose right to a Christian burial he
stoutly defended when the former Patriarch was accused not only of showing
undue sympathy towards members of the imperial family then in disgrace, but
also of being a crypto-Bogomil. The leader of the opposition to Tornikes at this
time was his colleague Soterichus Panteugenes, with whom was associated his
successor as Director of the Patriarchal School, Michael ho tou Thessalonikês.
Tornikes got his revenge in time, when Panteugenes and Michael were
condemned in one of the most celebrated heresy trials of the century. But in the
meantime he suffered a setback. In due course, however, probably after a change
of patriarchs in 1154, his chance came. He was put on the shortlist for the
Archbishopric of Corinth – the vacancy was possibly due to the death of Gregory
of Corinth, though the chronological problems are complex – but withdrew on
the advice of a highly placed and no doubt well-informed friend. A little later,
possibly in the second half of 1155, he was appointed Metropolitan of Ephesus.
Like so many others who had spent their lives in the brilliant and sophisticated
but narrow world of the capital, he found the transference to a position of high
responsibility in the provinces a traumatic experience. His letters from Ephesus,
like those of Michael Choniates from Athens, Michael Italicus from Philippopolis,
Basil Pediadites from Corfu, are marked by self-pity and lack of sympathy
426 Aristotle Transformed

towards the provincials. Captured by the Seljuks in 1090, recaptured by the

Byzantines shortly afterwards, harried by pirates, particularly the notorious
Čaqa, occupied by the Crusaders in 1147, Ephesus was but a shadow of its former
self. We hear in Tornikes’ letters of pieces of mosaic falling on the bishop’s head
as he officiated in the great church of St John the Theologian, around which the
medieval town had grown up. Heresy was rife. The local population was ignorant
and unfriendly. The Byzantine governor and his court were at Philadelphia, at
the end of a long journey over the mountains. In several of his letters Tornikes
talks of leaving his diocese and returning to the capital, but they cannot be
readily dated. He was probably still in Ephesus in May 1157, and he may have
remained there for years. In early summer 1166, however, we find him entrusted
by Manuel I with the drafting of a reply to a letter from Pope Alexander III , in
the course of negotiations of which we learn something in Boson’s contemporary
Life of that Pope. He must then have been in Constantinople, enjoying the closest
confidence of the emperor. This is his last appearance on the stage of history. By
February 1170 the Metropolitan of Ephesus is one Nicolaus.
So much for the man. Now to his funeral oration on Anna Comnena. I
propose to say something first on the biographical information which it
furnishes, then on the light it throws on the intellectual life of the period, but it
will not be easy to keep the two themes clearly separate. But first of all a word on
the construction of the speech. The overall pattern is that established in antiquity
for a basilikos logos – prooimion, patris, genos, genesis, anatrophê, paideia,
epitêdeumata, praxeis (classified under the four virtues), tukhêi – the rules are
most clearly formulated by Menander. But there is interference from a different
genre, the epitaphios logos. The fact that the subject is a woman in any case strains
the pattern of the basilikos logos. And Tornikes’ verbosity, as well his personal
involvement, lead to an almost independent development of certain of the
constituent parts. So the final result is not only very long – 13,000 to 14,000
words, only slightly shorter than Isocrates’ Panathenaicus – but also somewhat
complex. Tornikes begins by apologising for the delay in the delivery of his
oration and expressing his astonishment that no one else in Constantinople has
delivered a monody on the princess, the subject being so unique. He then turns
to the praise of her parents, Alexius I and Irene Ducas. Then come Anna’s birth,
upbringing and education, her appearance, her betrothal to Constantine Ducas
and his untimely death. Then the account of her subsequent betrothal to
Nicephorus Bryennius leads to a long digression on the virtues and exploits of
the princess’s husband. Next comes a long development on Anna’s devotion to
her mother and the lessons she learned from her. The theme of her care for her
 An unpublished funeral oration on Anna Comnena 427

parents leads to the account of the death of her father, and the change which it
brought about in Anna’s circumstances. Her misfortunes and the constancy with
which she met them are described with a wealth of literary imagery – the
Homeric cormorant, the wimple of Leucothea, etc. – but little precision. Next
comes an account of how she spent her enforced retirement in intellectual
pursuits, and gathered about her a group of scholars whose work she inspired
and directed, and of her own philosophical views – these are expounded at great
length. Embedded in this section is an account of the deaths of her mother and
her husband. Then Tornikes speaks of Anna’s literary work, her letters – which
do not seem to have survived – and her History. By a skilful transition – her wide
technical competence, her knowledge of medicine, both theoretical and practical,
her care of the sick, her care of her sister during her illness – he leads to the story
of the princess’s last illness, her taking of the veil, and her death. He concludes
with a development on Anna’s daughter – probably Irene, with whom he was in
correspondence – and on the many groups and classes who will sincerely mourn
her loss.
Anna Comnena was born in the Porphyry Chamber of the Great Palace at
dawn on Saturday, 2 December 1083. The date of her death is unknown: probably
after 1148 is the nearest one can get on the internal evidence of her History, and
the narrative sources for the middle of the twelfth century say nothing about her.
The new text gives a little more precision, but not much more. It was delivered,
says the title, when Tornikes was still hupomnêmatographos. This, as we have
seen, gives a date before the beginning of 1156, and some years after 1147. Some
time had elapsed, he says, since the princess’s death. The impression conveyed to
the reader is that he is thinking of months rather than years. I should be inclined
to date Anna’s death to the years 1153–5. Father J. Darrouzès informs me in a
letter that he has independently reached the same conclusion. Perhaps we get
some confirmation from a long passage in which the speaker compares Anna’s
relationship to the brilliant men of her age with that of a comet to the fixed
stars. In both cases the novelty of the event attracts attention. Now Halley’s comet
put in one of its regular appearances in April 1145, and was visible from
Constantinople; in February 1147 there was another comet, visible the world
over; and finally in May 1155 a comet appeared, visible throughout Europe. I
should like to think that men’s eyes had been raised to it in wonderment just
before our oration was delivered.
In any case, Anna must have been over seventy years of age, and have spent
some thirty-five years in retirement in her apartments in the monastery tês
Kekharitômenês when she died. She preserved to the last, says Tornikes, her rosy
428 Aristotle Transformed

complexion. We might at this point glance at our author’s description of the

princess’s appearance, the only one to survive, and one of the few eye-witness
descriptions of living persons in Byzantine literature (see Appendix to this
chapter, extract 1).
Anna’s own account of her education is celebrated (Alexiad, pr. 2; 15,7 fin.)
And she continually thanks her parents for having taken such care over it. The
story we find in the oration is slightly different, and perhaps more interesting.
Anna was trained by her parents in virtue, but not in logoi. They did not object
to her studying philosophy, which could be safely christianised and in any
case tended to moral improvement, but they were strongly opposed to her
studying its essential forerunner, grammar, which could not be christianised and
was morally neutral to say the least. It was dangerous enough for men, far more
so for women – and let us remember that Anna was no more than thirteen years
old at this time. However, the princess outwitted the vigilance of her parents by
taking private lessons in grammatikê from one of the palace eunuchs, like a
maiden having a clandestine assignation with her lover (see Appendix, extract
2). Later, after her marriage to Nicephorus Bryennius, she continued her
education openly, studying grammar under a palace eunuch – apparently a
different one this time – and philosophy under philosophers who combined
eloquence, a philosophical ethos, and old age. The final stage in her education
was after her father’s death and her own retirement from public life, when she
gathered around her some of the most learned men of her time, and with them
studied Aristotle, Plato, Euclid and Ptolemy – except such of his doctrines as
Christianity rejects – rhetoric and history.
Anna Comnena is eloquent on her woes in her History, but she never
makes it clear exactly what they were. We know from other sources that she
spent the latter part of her life, after the death of her father, in the convent tês
Kekharitômenês in Constantinople in semi-retirement, though of course she did
not actually become a nun until she was on her death-bed – a fact which we
learn from the present oration. The chronicler John Zonaras, whose history
ends with the year 1118, tells us in some detail how Anna, who hated her
younger brother John II , and still hoped to place the imperial crown on the
head of her able, but good-natured and unambitious, husband Nicephorus
Bryennius, quarrelled bitterly with her brother over their father’s death-bed,
and was the prime mover in a plot to assassinate him shortly after his accession
(Zonaras, 18,28–9). The same story is told by Nicetas Choniates, writing at
the beginning of the thirteenth century. If true, it would be sufficient explanation
of the obscurity in which the rest of Anna’s life was passed. Anna’s own
 An unpublished funeral oration on Anna Comnena 429

account of the death of her father gives an idyllic picture of a united family,
though she does mention that John went off to the Great Palace before
their father had breathed his last. And Anna’s striking coldness towards her
brother, whom she scarcely mentions in her History, has often attracted
attention. Zonaras’ story cannot be dated, unfortunately. We know that he was
still alive and working on his great compilation of Canon Law in 1159, but
his Chronicle may have been completed long before that. What does Tornikes,
an intimate friend of the family, have to say? He was clearly aware of an
account of the events discreditable to Anna, and implies that it had obtained
wide credence, though of course he somewhat baldly denies it (see Appendix,
extract 3). The probability is that he had the text of Zonaras before him. This is
not a very convincing defence of Anna, but then perhaps a funeral oration was
not the place for polemical discussion of the events of more than a generation
before.
Anna’s nephew Manuel I, whom she perhaps scarcely knew, seems to have
done nothing after his accession in 1143 to heal the breach with his formidable
sexagenarian aunt. He gets a single brief and cold mention in this speech. In any
case Anna and Manuel would never have seen eye to eye. Manuel was a
Latinophile and a womaniser, and he was an enthusiastic believer in astrology, a
science in defence of which he wrote a treatise. The condemnation of astrology
which Tornikes in the present speech attributes to Anna may well have been
directed mainly at her brilliant but erratic imperial nephew. So far as it is possible
to date Manuel’s defence of astrology, it is certainly later than 1147, and was
probably written not long before 1156.
When death finally came to the princess as she was engaged in consoling one
of her sisters – we do not know which – after a recent bereavement, she had been
cut off from the sophisticated world of the court for thirty-five years. But she had
not been out of contact with the intellectual movement of the time. It was in the
closing years of her retirement, as we know, that she composed the history of her
father’s reign. The present oration furnishes precious information on another of
her activities, her philosophical circle. Let us look once again at the passage in
question (see Appendix, extract 4). It seems that Anna not only studied herself – a
procedure which would involve lectures or supervisions by specialists – but also
organised and inspired, and no doubt paid for, the work of others. In particular
we hear of the encouragement of Aristotelian commentators, and especially of
the exegesis of works on which no commentary had survived from antiquity.
And the name of Michael of Ephesus is mentioned as one who engaged in these
activities under Anna’s patronage.
430 Aristotle Transformed

Now the long series of commentaries on the works of Aristotle, which begins
in the first century bc with Andronikos, the rediscoverer of the Aristotelian
corpus, continues without a significant break down to Stephanus of Alexandria
and his pupils David and Elias in the seventh century.1 Then, with the loss to
the Arabs of the school of Alexandria, the tradition seems to dry up. John
Damascene in the eighth century was profoundly influenced by Aristotle and his
commentators, but by his time it had become a closed body of thought; there was
no further exegesis or development. Surprisingly, exegesis of Aristotle begins
again in the eleventh or twelfth century with two men whose work survives,
Eustratius, Metropolitan of Nicaea, and Michael of Ephesus. The former
commented upon certain books of the Nicomachean Ethics and sections of the
Organon, the latter on other books of the Ethics, sections of the Organon, the
Rhetoric, the Physics, the Politics, and a number of the zoological and
anthropological works. Eustratius is a well-known figure, a pupil of John Italus,
who made a brilliant career in the Church, wrote many works of dogmatic and
polemical theology, took part in the discussion with Petrus Chrysolanus in 1112,
and found himself charged with heresy in 1117 as a result of an anti-Armenian
tract which he had composed. He recanted, but was probably suspended for life.
We do not know when he died. The date cited in the handbooks, c. 1120, is the
result of a somewhat labile construction by Draeseke.2 The date of his birth is
equally unknown; it could be as late as 1060. As for Michael of Ephesus, nothing
at all is known of his life, and up to now all that could be done by way of dating
him was eleventh/twelfth century. Tatakis in his recent book on Byzantine
philosophy supposes Michael to be a contemporary of John Italus and a
predecessor of Eustratius.
The present text fills out the picture in much more detail. Michael of
Ephesus’ commentaries belong to the years of Anna’s retirement after 1118.
And they were probably completed by 1138, since after that year Anna was
mainly engaged in the composition of her History, originally intended as a sequel
to that written by her late husband. Karl Praechter in a well-known paper3
remarks that anyone looking at the list of works of Aristotle commented upon in
late antiquity or early Byzantine times is struck by three gaps – the Politics, the
Rhetoric, and the zoological and anthropological works. Eight hundred years

1
[For differing views on the relations of these three philosophers, see Chapters 14 and 1 – Ed.]
2
Byzantinische Zeitschrift 5,1896, 319–36.
3
K. Praechter, review of Michaelis Ephesii In libros De partibus animalium . . . Commentaria,
M. Hayduck (ed.), CAG 22.2, in Göttingische gelehrte Anzeigen 168, 1906, 861–907.
 An unpublished funeral oration on Anna Comnena 431

before Praechter wrote these words, the same point seems to have struck Anna
Comnena, who had the resources and the connections to remedy the shortcoming.
Michael of Ephesus was breaking entirely new ground in his commentaries on
the zoological and anthropological works and on the Rhetoric and the Politics.
And he was doing it as part of a co-operative scholarly undertaking conceived
and guided by Anna Comnena. The list of his own commentaries which he gives
illustrates how systematically the plan of Anna or of her advisers was being
carried out.4
What of Eustratius? In her History Anna speaks of him with high praise
(14,8). In the proem to his commentary on EN 1 he tells us that it was composed
at the behest of a highly placed personage whom he does not name. And in the
proem to that on EN 6 he addresses his patron as basilis theosebês, basilis
philologê, basilis philagathê kai philokalê. So far as I can discover, the identity of
this princess – the word need not mean the wife of a reigning emperor in
Byzantine usage – has never been cleared up. It is very tempting, and an obvious
working hypothesis, to suppose that she was Anna Comnena, and that Eustratius’
exegetical work belongs to the years of his theological disgrace. In the
commentary on EN 6 he speaks of himself as an old man, for what that is worth.
And it is significant that he was apparently asked to comment only on EN 1 and
6, while the other books of the Nicomachean Ethics were commented on by
Michael of Ephesus and possibly by a nebulous Aspasius5 – there is some
variation in the titles in the manuscripts. No part of the Ethics was the object of
comment by more than one of these scholars. This suggests that their
commentaries were all prepared as part of the same co-operative enterprise. It
may well be that Eustratius was the real inspirer of the whole project.
If some measure of probability can be accorded to our hypothesis thus far,
Anna Comnena played a role in the revival of Aristotelian scholarship in the
Byzantine world. It has long been a commonplace that the renaissance of
Aristotelian exegesis depended ultimately on the renewed interest in and grasp
of the ancient philosophical tradition displayed by men like Michael Psellus and
John Italus in the eleventh century. But this remains a vague formulation. The
tendency of Psellus and John Italus was Platonist or Neoplatonist rather than
Aristotelian. The detailed connection and organisational link is missing. The
more one examines Byzantine literature, the more one becomes convinced that
it never appears spontaneously; it needs a salon, patronage, institutionalised

4
Göttingische gelehrte Anzeigen 22, 149, 8ff.
5
[In Chapter 18, Mercken credits the second-century Aspasius – Ed.]
432 Aristotle Transformed

forms. We know of some of the literary circles of the earlier twelfth century. The
Patriarchal School housed one such circle, whose interests, theology apart, lay in
the fields of grammar, rhetoric and belles-lettres; Michael Italicus is one of its
leading figures. Another centred upon the sebastocratorissa Irene, widow of
Manuel’s elder brother Andronicus. Poetry was one of its main interests;
innumerable occasional poems of Theodore Prodromus can be connected with
it; Tzetzes commented on Hesiod and Homer and wrote his Theogony for Irene;
and she was the patroness and dedicatee of Constantine Manasses’ verse
chronicle. We can now add to these Anna Comnena’s philosophical circle,
numbering among its members Michael of Ephesus and probably Eustratius of
Nicaea. It is worth recalling in this connection that it was in Constantinople in
the thirties of the twelfth century that James of Venice became acquainted with
the Physics, the Sophistici Elenchi, and other works of Aristotle, which he
subsequently translated into Latin.6 It has been assumed that he found them
studied in the University. But we must now reckon with the possibility that he
had contact with Anna Comnena’s Aristotelian circle. It may be significant that
the Sophistici Elenchi, one of the works translated by James of Venice, was first
commented upon by Michael of Ephesus. It may well be, too, that other works of
philosophical and mathematical exegesis will prove to have been composed
under Anna’s watchful eye in the suite in the Kekharitômenê convent, at the head
of a valley overlooking the tranquil waters of the Golden Horn. One would look
in the first place among the nameless Byzantine scholiasts on Euclid and Ptolemy,
who continued the tradition of Proclus and Simplicius. John Tzetzes’ unpublished
commentary on the Canons of Ptolemy, in cod. Paris. gr. 2162, was probably
composed too late to qualify, but one cannot be certain.
In so far as one can attach political labels to such things, Anna Comnena’s
circle belonged to the ‘outs’. She herself was in semi-disgrace to the end of her
days; Tornikes’ complaint that no one but himself had thought of commemorating
her death bears this out. Eustratius was a man whose career was finished; the
Patriarch Cosmas II Atticus, who promoted Tornikes, and whom Tornikes
defended after his death, was dismissed after just over a year in office, charged
not only with Bogomil sympathies, but also with supporting Manuel’s disgruntled
relatives.
From time to time in her history Anna mentions this or that point of
philosophy. However, the long account of her tenets given by George Tornikes

6
cf. L. Minio-Paluello, Traditio 8, 1952, 265–304.
 An unpublished funeral oration on Anna Comnena 433

provides a more comprehensive view than we can derive from the History. I
shall outline the main points briefly, as they are not of much importance in
themselves. Anna may not have really had much of a head for philosophy,
Tornikes was probably no philosopher, and I am certainly quite incompetent
in this sphere. Anna sees as her task to fit the tradition of ancient philosophy
to the requirements of Christian dogma. The series of syntheses which had
been made in the past no longer satisfy her entirely. And the tradition of ancient
philosophy means to her, as to all Byzantines, Aristotle and Plato, both seen
through the spectacles of commentators of late antiquity, who were either pagan
Neoplatonists like Simplicius, or Aristotelian heretics like John Philoponus.
In general, she says, she finds Aristotle most satisfying. But she objects to his
agenêsia (significantly the word seems first attested in Simplicius), that is, to
an uncreated universe which leaves no room for providence and in which
everything must be automaton. This is precisely the point in dispute between
the schools of Athens and Alexandria in the early sixth century, which led John
Philoponus in 529 to publish his de Aeternitate Mundi contra Proclum. To
avoid this difficulty she accepted Plato’s dêmiourgos; but she rejected totally
to par ’autôi tôn ideôn mêkhanêma, the Aristotelian criticism of which she
regarded as valid. She admired Plato and his followers, the Neoplatonists,
and approved many of their doctrines, such as that of the anousion agathon,
but preferred Christian explanations of such matters, particularly those of
Pseudo-Dionysius the Areopagite. However, she borrowed from the Timaeus
the circles of the other and the same – men had been confusing themselves
by interpreting the Timaeus literally for a millennium – and she accepted
Plato’s doctrine of the divisions of the soul, but of course rejected his
metempsychosis. From Aristotle she borrowed the concept of entelechy
(actuality). She favoured a double entelechy (dittê entelekheia) in regard to the
soul – this points to her familiarity with the second book of the de Anima – and
she frequently illustrated this by the analogies of the lyre-player and the
steersman. This leads to another passage of the de Anima which, unfortunately
for Anna, modern authorities generally regard as a momentary lapse on the part
of Aristotle. But John Philoponus in his commentary took it seriously, and the
word for steersman used by Tornikes is Philoponus’ and not Aristotle’s. And both
the topic and the imagery alike figure in the surviving writings of Michael Psellus
and John Italus. Above all, she was determined to find a physics and metaphysics
which would not destroy the basis of ethics, and this to Anna meant an after-life
in which our actions are judged, rewarded, and punished. Belief in destiny –
heimarmenê – was to her something worthy of beasts rather than men. It is
434 Aristotle Transformed

significant that it was precisely the Physics and Metaphysics which were neglected
in the orthodox Christian – Aristotelian synthesis represented by Leontius of
Byzantium and John Damascene.
I doubt if much can be made of Tornikes’ garbled and rhetorical account. But
it is clear that Anna and her colleagues were bent on constructing a philosophical
system, and not merely on glossing texts. They were not alone in this in the
twelfth century. Michael Anchialus, the future Patriarch, was appointed Professor
of Philosophy – hupatos tôn philosophôn – about 1165, after the office had
apparently been in abeyance for some time. In his inaugural lecture, which I have
edited elsewhere,7 he lays great emphasis upon the study of Aristotle and upon
the support which a somewhat eclectic Aristotelianism can give to religion. He
attacks at great length the followers of pagan theology, who believed matter to be
uncreated. This is just the point which gave offence to Anna. And it is also, we
recall, one of the charges made in 1082 against John Italus, who was also accused
of reviving the errors of ancient philosophers on the nature of the soul, and
of regarding profane literature as a fountain of truth. In 1156 Soterichus
Panteugenes, whom we have already met briefly as an enemy of Tornikes, was
condemned, together with a number of leading men of letters and teachers in
the Patriarchal School, for heretical interpretation of the liturgy of the Eucharist.
Nicolaus of Methone, the leading theologian of the age, expressly connected the
heresy of Panteugenes with the Platonic doctrine of Ideas, which, he declared,
had been refuted by Aristotle. And Nicolaus was no novice in philosophy. He
wrote, probably before 1156, an anaptuxis tês theologikês stoikheiôseôs Proklou in
which he says that he is arguing against men of his own age, who call themselves
Christians, but who are befuddled by pagan learning, and so epiprosthen tôn
oikeiôn tithentai ta allotria. It has been plausibly suggested, too, that Michael
Glycas the chronicler, whose heretical doctrines began to attract attention early
in the reign of Manuel I, was trying to construct a systematic cosmology of
neoplatonising character.
System-building was not the monopoly of hellenisers and heretics at this
time. Euthymius Zigabenus wrote his Dogmatic Panoply and Nicetas Acominatus
his Treasury of Orthodoxy to combat the rationalist heresies so attractive to the
twelfth-century mind. Neilus Doxopatres, an exalted functionary on the
patriarchal staff, who became a monk and went off to Norman Sicily shortly
before 1143, composed a great dogmatic summa, on a scale quite unprecedented

7
Balkan Studies (Thessalonika), 2, 1961, 173–214.
 An unpublished funeral oration on Anna Comnena 435

in the Byzantine world. It was in five books, of which only the first two –
comprising between them 466 chapters – survive unpublished.
In a book written seventy years ago8 Fyodor Uspenskij argued that throughout
the twelfth century a long argument went on between Nominalists and Realists
in the Byzantine world, parallel to and not unconnected with that going on at the
same time in the Latin world. By and large the Realists lean on Plato, the
Nominalists on Aristotle.
We need a new study of the ground covered by Uspenskij seventy years
ago. The time is scarcely ripe yet. Too many of the relevant texts still slumber
unread in the dust of libraries. But we can begin to see why the tradition of
Aristotelian exegesis was taken up again so vigorously after a lapse of nearly four
centuries. And this unpromising-looking funeral oration gives a glimpse of the
commentators at their work. To be quite fair, Anna Comnena and her circle were
probably not quite the first to revive the tradition. Theodore of Smyrna, successor
of John Italus as hupatos tôn philosophôn, who survived until the closing years
of Alexius I’s reign, if not longer, and who appears among the intelligentsia of
the underworld in that lively Byzantine satire the Timarion, had composed an
exposition of the physical doctrines of the ancients, in which he seems to have
dealt with Aristotle at some length. His work is unpublished. But this was only
a small beginning. Without the resources at Anna’s command the movement
might well have petered out. In fact more than a century and a half elapsed
before the next great ‘wave’ of Aristotelian commentaries in Byzantium. We
clearly owe much to the drive and inspiration of this astonishing lady, whose
memory is enshrined not only in the immortal history of her father’s reign, but
also in the austere volumes of the Commentaria in Aristotelem Graeca.9

Appendix

1. Her eyes, of well-proportioned size and shining with happiness, neither

constantly moving and turning about – the signs of an unstable and giddy
soul – nor immoderately motionless and sluggish – the marks of a slothful
disposition – moved with liveliness and ease when she wished to survey her

8
F. Uspenskij, Ocherki po istorii vizantijskoj obrazovannosti, St Petersburg 1891 (1892), 146–245.
9
Since this article was first published, the surviving works of George Tornikes have been edited for
the first time by Father J. Darrouzès in Georges et Démétrios Tornikès: lettres et discours, Paris 1970.
The translation in the Appendix is from Darrouzès text, not from that printed in the original article.
436 Aristotle Transformed

surroundings, but were more often steady and fixed. Her eyebrows were curved
like the rainbow; her nose was straight and slightly curved towards her lip,
protruding slightly and harmoniously; her lips were compressed like rosebuds,
recalling the scarlet cord of the Song of Songs. Her whole complexion was white
as wool, while her cheeks were tinged with a rosy blush, which she retained into
old age. Her whole countenance formed a well-rounded circle, as if traced with a
compass. Her head was firmly set upon her backbone by a neck of proportionate
length, her shoulders did not heave irregularly, her feet were agile, her hands
even more so. All her members were fair in themselves and fair in relation to one
an other. Her body was like a lyre or a well-tempered lute, a good instrument
fashioned for a good soul (fol. 25r).
2. As for profane learning, they suspected it of being insidious, just as wise
mothers of children often distrust match-makers, lest they inspire in maidens
dishonourable passions. For they admired and welcomed the subtlety of its
logical procedures and the virtuosity of its dialectic, as well as its manifest
revelation of the secrets of nature, and, among some of its practitioners (the
understanding) of the creation of the universe by the Prime Mover and his
providence, as revealed to them by God. But the literary study which precedes
these disciplines, the principal object of which is poetry and its subject-matter
polytheism, or rather atheism and fables recounting the ill-starred loves of gods,
the violation of virgins, the abduction of youths, and other exploits as disgraceful
in word as in deed, this study they rightly thought dangerous even for men, and
for women and girls excessively insidious, and believed that their eyes and ears
should remain undefiled by them, since it is through these that passions enter
the soul.
Such were their thoughts, truly worthy of imperial persons and followers of
Christ. But she – for she listened attentively to those who gave displays of their
wisdom every day before her father the Emperor, and was roused to emulation
by them – what did she do? She knew the judgment of her parents concerning
profane wisdom and accepted it. But as soldiers who have learnt of an ambush
laid by the enemy but cannot return home by any other route brave their
ambuscade, after arming themselves well and preparing themselves to hold out
with courage to the end, so she armed herself against the deceitful fictions and
the stories of dishonourable passions and braced the weakness of her soul so that
it might not be taken by surprise or carried away by the magic potion of Circe or
the songs of the Sirens. She closed and opened her ears in accordance with the
guidance of reason and, advancing with Odysseus’ magic herb, braved grammar
and poetry, taking care not to be detected by her parents, and surreptitiously and
 An unpublished funeral oration on Anna Comnena 437

with caution taking lessons from eunuch servants who were not ignorant of such
matters. And like a maiden who takes a furtive glance at her bridegroom through
some chink, so she had furtive meetings with her beloved grammar (fol. 24v).
3. But his son, who was also Emperor, while his father was still breathing, took
up the sceptre of power and went off to the palace, as indeed he had to do, since
time presses in such situations, and many unexpected claimants to imperial
power plot against an Emperor’s heirs if there is the least hesitation. She however,
who according to those who say anything lightly was her brother’s rival for
power, although she knew that her father had just left this world, forgot, along
with her children and her husband, the imperial title and joined with her mother
in mourning, as they sat alone on the floor with bared heads around the Emperor
instead of the chamberlains, the officers, and the company of valorous guardsmen
(fol. 27v).
4. So she gathered together all the most eminent representatives of the logical
sciences – and they were numerous and remarkable. For this was one of the
achievements of the age of Alexios, that among those who taught the young by
daily exercises he honoured the most distinguished by gifts and imperial
dignities, which were given to men of letters as to others. First were those who
were philosophers by their knowledge and their way of life, making this their
prime goal rather than money-making or commerce, wealth or office, but rather
gathering knowledge from books and spreading it in turn among the souls of
those who desired it, and pouring into their ears great wisdom in simple words.
Then came those who were at one and the same time men of the world and
philosophers and eloquent of tongue, combining wealth of knowledge with
elegance of exposition, teachers as brilliant by the content of their thought as by
its outward expression. By these men she was initiated into such of the doctrines
of the Stagirite and Plato, Euclid and Ptolemy, as are not banished from society
by the laws of the Christians. Neither did she reject political or epideictic oratory,
nor did she disdain the utility of history, which provides the subject-matter
for rhetoric and most of its verisimilitude. For she lived with poetry, wept with
the tragedians and laughed with the comedians, refreshing herself from her
philosophical labours by such distractions; some things she laughed at, others
she lamented and wept over; she praised the laughter of Democritus and admired
no less the tears of Heraclitus. The works which philosophers of our time
addressed to her bear witness to her love of learning, works concerning those
writings of Aristotle on which commentaries had not been written until her
time, but the explanation of which was transmitted orally in every kind of form,
without certitude and with little zeal. For books are an inviolable storehouse of
438 Aristotle Transformed

words, and writing is an indestructible memorial of thought, while the ear is

often violated by oblivion as by robbers. I have myself heard the wise man from
Ephesus attribute the cause of his blindness to her, because he spent sleepless
nights over commentaries on Aristotle at her command, whence came the
damage done to his eyes by candles through desiccation (fol. 29v).
 18

The Greek commentators on Aristotle’s Ethics

H.P.F. Mercken

The main commentaries

I shall examine the most important of the Greek commentaries on Aristotle’s

Nicomachean Ethics. One is a composite commentary (CAG 19.1, 158–86, and
20 G. Heylbut (ed.), 1889 and 1892; CAG 22.3, M. Hayduck (ed.), 1901), compiled
in twelfth-century Constantinople, and translated into Latin in thirteenth-
century England by Robert Grosseteste. Although the date of compilation is
late, the composite work incorporates some of the earliest Greek commentaries
on Aristotle that we possess, produced by Aspasius and other Peripatetics in
the second century ad. This early commentary is juxtaposed with twelfth-
century work, including that of Eustratius of Nicaea and Michael of Ephesus.
The compilation is all the more significant because of the shortage of commentary
on Aristotle’s Ethics from the heyday of commentary writing which stretches
from Alexander in the third century ad to Simplicius in the sixth.
There are further commentaries by Aspasius on Books 1 to 4 and part of Book
7 of the Nicomachean Ethics (CAG 19.1, G. Heylbut (ed.), 1889, 1–157).
The remaining Greek literature, which I shall not discuss, consists of: (i) an
epitome of Aristotle, including his ethics, written by Arius Didymus in the first
century bc , and partly preserved by Stobaeus.1 (ii) Alexander of Aphrodisias,
Ethical Problems (CAG suppl. 2.2, I. Bruns (ed.), 1892, translated by Robert W.
Sharples, 1989), from the beginning of the third century ad, not in commentary
form. (iii) Pseudo-Andronicus of Rhodes On Emotions (Peri Pathôn), also
translated by Grosseteste, found in MS Peterhouse 116, A. Glibert-Thirry (ed.),
Pseudo-Andronicus de Rhodes PERI PATHÔN, Édition critique du texte grec
et de la traduction latine médiévale, Corpus Latinum Commentariorum in

1
cf. P. Moraux, Der Aristotelismus bei den Griechen, vol. 1, Berlin and New York 1973, 316–434:
‘Areios Didymos, der Abriss der aristotelischen Ethik.’

439
440 Aristotle Transformed

Aristotelem Graecorum (CLCAG ), supp. 2, Leiden 1977. (iv) Paraphrase of the

Whole Philosophy of Aristotle by Georgius Pachymerius from around 1300.
(v) The paraphrase commentary ascribed unreliably in one MS to the otherwise
unknown Heliodorus of Prusa (CAG 19.2, G. Heylbut (ed.), 1889), which may
date from any time up to 1366.2
The contents of the composite commentary are diverse: commentaries on
Books 1 and 6 by Eustratius, metropolitan bishop of Nicaea, in the years marking
the transition from the eleventh to the twelfth century; anonymous scholia on
Books 2, 3, 4, and 5, probably from the last quarter of the second century ad ;
another commentary on Book 5, as well as one on Books 9 and 10, by Michael of
Ephesus, professor at Constantinople and contemporary of Eustratius; an
anonymous commentary on Book 7, which is in all likelihood even later than
those by Eustratius and Michael; finally a commentary on Book 8 by Aspasius, a
renowned Peripatetic of the first half of the second century of our era. The
composite commentary thus comprises texts from five authors, spread over
eleven centuries: Aspasius, Michael of Ephesus, Eustratius of Nicaea, and two
anonymous scholiasts.
There has been one earlier edition, the Aldine edition of 1536: Eustratiou kai
allôn tinôn episêmôn hupomnêmata eis ta deka tôn tou Aristotelous Êthikôn
Nikomakheiôn biblia meta tou hupokeimenou . . ., with colophon: Venetiis, in
aedibus haeredum Aldi Manutii, et Andreae Asulani soceri, mense Julio.
M.D.XXXVI . It lacks the anonymous scholia on Book 5.
The extant Greek manuscripts containing the commentaries of Eustratius
‘with others’ can be divided into two main classes, the first of which contains
basically the compilation which Grosseteste translated; the second class, lacking
the early anonymous scholia, contains in their stead most or all of the remaining
commentary of Aspasius.3

The Greek manuscripts of the composite commentary

with the early anonymous scholia

It is a curious fact that there are only two extant Greek manuscripts with exactly
the compilation which Grosseteste must have had in his possession. They are
Oxford, Corpus Christi College 106 and Oxford New College 240/241. Being the

2
cf. R.A. Gauthier, in R.A. Gauthier and J.Y. Jolif, Aristote, L’Éthique à Nicomaque 2, Louvain and Paris
1970, Introduction, 106–7.
3
See G. Heylbut, CAG 19.1, v–vii, and 20, v–ix.
 The Greek commentators on Aristotle’s Ethics 441

only ones to contain the anonymous scholia on Book 5, they have consequently
been used for the edition of these scholia by Heylbut, who gave them the sigla
F and G, respectively.
There are, in addition, two manuscripts which differ only from the former
and from Grosseteste’s (lost) manuscript by the omission of the anonymous
scholia on Book 5. This omission may have been justified by the consideration
that Michael of Ephesus had incorporated most of these glosses in his
commentary on the same book. These manuscripts are Milan 95 Ambrosianus B
sup. and Paris Coislinianus 161. The latter is the only manuscript on which
Heylbut, who gave it the siglum B, based his edition, for which he also used the
Aldine edition. The Aldine edition contains the same text as Coislinianus 161
(=B). Heylbut gave it the siglum a.

The Greek manuscripts of the composite commentary with

the remaining commentary of Aspasius

Of Aspasius’ commentary on the Ethics, only two parts have survived, viz. one on
Books 1–4 and one on the greater part of Books 7 and 8.
The Greek manuscripts containing Eustratius ‘with others’ of the second class
differ from those of the first in the following ways:
(1) They do not contain the older anonymous glosses on Books 2–5.
(2) They add the commentary of Aspasius on Books 1–4 after Eustratius on
Book 1.4
(3) They add capita moralia, taken from the Quaestiones, attributed to
Alexander of Aphrodisias,5 after Eustratius on Book 6.
(4) They add Aspasius on Book 7 after the anonymous scholia on the same
book, with the exception of Vatican gr. 269 and Paris gr. 1927. The omission in
these two manuscripts may have been influenced by the consideration that
Aspasius’ commentary on Book 7 covers only the second half of that book.
To this second class belong Florence Laurentianus gr. 85,1 (=N); Modena gr.
197 (II .G.4); and Berlin (ex collectione Carolii Morbii) fol. 58,59 (Acc. 1889,
304).6 I deliberately count also the incomplete manuscripts Vatican Barberinus

4
The text of this commentary is acephalous. It starts with tên dialektikên enioi ephasan . . . on p. 3,11
of Heylbut’s edition (CAG 19.1) and follows immediately upon Eustratius’ commentary. As a result,
the presence of Aspasius’ commentary on Book 1 in these MSS has often been overlooked. Heylbut
himself omits it in his description of Laurentianus gr. 85,1, on p. vii of CAG 20.
5
I. Bruns (ed.), Supplementum Aristotelicum 2 pars 2, Berlin 1892, 117–63.
6
The last MS is not mentioned in the Praefatio of CAG 19.1.
442 Aristotle Transformed

gr. 223 (formerly II 44); Leiden Vossianus fol. n. 12; Vatican gr. 269 II ; and Paris
gr. 1927 in this class, because they also conform to the description of that class in
so far as they can exhibit it, that is, only partially.
Finally, Vatican gr. 320 is fragmentary in that it contains only the commentary
of Michael of Ephesus on Books 9 and 10.
In summary, all extant Greek manuscripts mentioned, in so far as they are
complete, contain Eustratius on Books 1 and 6, Michael on Books 5, 9, and 10,
the recent anonymous scholiast on Book 7, and Aspasius on Book 8; one main
class contains in addition the anonymous glosses on Books 2, 3, and 4 (and 5 for
the manuscripts F and G); the other main class contains instead (again in so far
as the manuscripts are complete) capita moralia of Alexander of Aphrodisias
and the whole of the extant commentary of Aspasius minus the first folio (except
on Book 7 in two manuscripts, Vat. gr. 269 and Paris gr. 1927).
One cannot but wholeheartedly agreed with V. Rose that the second class of
manuscripts containing Eustratius ‘with others’ resulted from a deliberate
substitution of Aspasius for the anonymous glosses, and that the first class of
manuscripts (and hence the translation of Grosseteste) represents the original
compilation. The only part of Aspasius’ commentary which belongs to the
original compilation is that on Book 8. The decisive argument is that the text of
this book in the manuscripts of the compilation deviates considerably from the
one in the manuscripts containing Aspasius’ commentary alone, whereas the text
of the other books of Aspasius’ commentary, whenever present in the manuscripts
of the compilation (that is, in the second class of these manuscripts), is very close
to that in the manuscripts containing Aspasius’ commentary alone, the main
difference being that the latter’s text is not acephalous.7

The Renaissance translation of the composite commentary

In 1541 Giovanni Bernardo Feliciano published in Venice a new translation of

the Nicomachean Ethics and of the Greek commentaries. He included the same
items Grosseteste translated, with the exception of the anonymous scholia on
Book 5, but adding some parts from Aspasius. Feliciano mixed the anonymous
scholia on Books 2–4 with those of Aspasius on the same books. He attributed

7
V. Rose, ‘Ueber die griechischen Commentare zur Ethik des Aristoteles’, Hermes 5, 1871, 71 and 73.
Rose’s article (61–113) contains a wealth of information and is amazingly accurate, especially if one
considers the time at which it was written. One can only regret that until recently subsequent
scholars, in particular the editors of the Greek commentaries, have not made sufficient use of it.
 The Greek commentators on Aristotle’s Ethics 443

the worthless scholia on Book 7 to Aspasius and the commentary of Michael of

Ephesus on Books 9 and 10 to Eustratius, while mentioning that some attributed
these to Michael. He did not include a translation of Aspasius on Books 1 and 7.
Later editions are Basel, 1542; Paris (Roigny), 1543; Venice (heres Hier. Scoti),
1589; and Helmstadt (Sam. Rachelius), 1662. They contain practically the same
text, although the Paris edition shows some corrections, including emendations
based on the Greek text.

Eustratius on Books 1 and 6

Eustratius is the best known of our commentators.8 He was born c. 1050 and
died c. 1120, as Draeseke has conclusively argued. He received an education in
philosophy from John Italus, the well known successor of Psellus as the head of
the school of Constantinople (hupatos tôn philosophôn). John Italus, whose
interest in Aristotle counterbalanced the Platonic tendency of Psellus, was much
admired by his contemporaries for his dialectical ability, but his attempts to
apply philosophical reasoning to the mysteries of Christian revelation led to his
downfall. A synod in 1082 condemned nine articles, which he admitted having
defended, as being filled with pagan atheism, forbade him to teach either
privately or publicly, and relegated him to a monastery. The synodicon of
Orthodoxy Sunday (first Sunday of Lent) retains eleven anathemata against
John Italus and his doctrines, the first of which is directed against his attempts to
discuss dialectically the mystery of the Incarnation and of the hypostatic union
of two natures in Christ. Eustratius, who was merely a deacon at the time, and
principal (proximos) of the school of Sporakin, is cited with other disciples of
John Italus, all of whom succeeded in convincing the synod of their innocence

8
For the whole compilation, see F. Schleiermacher, ‘Ueber die griechischen Scholien zur
Nikomachischen Ethik des Aristoteles’, Sämmtliche Werke, Abt. 3, Bd. 2, Berlin 1838, 309–26;
V. Rose, ‘Ueber die griechischen Commentare zur Ethik des Aristoteles’, Hermes 5, 1871, 61–113;
and R.A. Gauthier in R.A. Gauthier and J.Y. Jolif, op. cit. in n. 2, 100–5 and 121. For Eustratius, see
J. Draeseke, ‘Zu Eustratios von Nikaea’, Byzantinische Zeitschrift 5, 1896, 319–36; S. Salaville,
‘Philosophie et théologie ou épisodes scolastiques à Byzance de 1059 à 1117’, Échos d’Orient 29,
1930, 132–56; P. Joannou, ‘Eustrate de Nicée. Trois pièces inédites de son procès (1117)’, Revue des
Études Byzantines 10, 1952, 24–34; id., ‘Die Definition des Seins bei Eustratios von Nikaia. Die
Universalienlehre in der byzantinischen Theologie im XI . Jh.’, Byzantinische Zeitschrift 47, 1954,
258–368; id., ‘Der Nominalismus und die menschliche Psychologie Christi. Das Semeioma gegen
Eustratios von Nikaia (1117)’, Byzantinische Zeitschrift 47, 1954, 369–78; id., ‘Le sort des évêques
hérétiques réconciliés. Un discours inédit de Nicétas de Serres contre Eustrate de Nicée’, Byzantion
28, 1958, 1–30; K. Giocarinis, ‘Eustratius of Nicaea’s defense of the doctrine of Ideas’, Franciscan
Studies 24, 1964, 159–204; A.C. Lloyd, ‘The Aristotelianism of Eustratios of Nicaea’, Aristoteles: Werk
und Wirkung II , J. Weisner (ed.) 1987, 341–51.
444 Aristotle Transformed

in the matter at hand and signed a letter in which they anathematised their
master’s heretical doctrines.
Eustratius soon became an influential figure. Only a few years after the trial of
Italus, he wrote a dialogue and a treatise on the question of the Icons against
Leon, metropolitan bishop of Chalcedon. Leon had been relentless in accusing
Emperor Alexius I Comnenus of sacrilege and iconoclasm, for the latter had
caused sacred vases and golden and silver ornaments of church gates to be
melted down for coin to sustain his wars. For this Alexius had obtained the
consent of the clergy. Eustratius earned the emperor’s favour and protection by
writing against Leon, and his own election to the metropolitan seat of Nicaea
was no doubt due to his connection with the emperor.
In 1111, Eustratius, as an official court theologian, expounded with others
in the presence of Alexius the Byzantine position against Petrus Grossolanus,
archbishop of Milan and representative of Pope Paschalis II , during one
of the many fruitless attempts to reconcile East and West after the schism of
1054. Eustratius’ position against the use of unleavened bread for Mass and
against the doctrine teaching the ‘procession’ (proodos) of the Holy Ghost from
the Son is found in several small opuscula which originated during these
discussions.
In 1114, Eustratius accompanied Alexius to his summer residence in
Philippopoli, where the emperor, duly supported by Eustratius, conducted a
series of conferences against the monophysite heresy of the ‘Armenians,
Manicheans, or Bogomils’. Eustratius on this occasion wrote a dialectical
discourse against the Armenian Tigranes on the two natures of Christ, a model
of its kind, and the rough draft of two treatises on the same subject. The latter
two were which Eustratius had explained in an unorthodox way. These opuscula
were taken away from him – so he defended himself later – by some unknown
person before they were finished; they were then copied and brought into
circulation without his knowledge or consent. The theologians of the court
thereupon accused Eustratius of heresy. He had defended the eternal subservience
of the human element of Christ to God the Father. Trying to save his favourite,
the emperor presided personally over the synod of April 17, 1117, together with
the patriarch of Constantinople. There, Eustratius addressed himself to the
emperor, the patriarch, and the synod in a written confession (exomologêsis), in
which he told of the fate of the incriminated writings and revoked and
condemned all errors in any form contained in them. They were, he explained
apologetically, the result of his ignorance and carelessness, but he had never
believed in them. A few days later the synod was urged by the patriarch to show
 The Greek commentators on Aristotle’s Ethics 445

clemency, but neither the effort of Alexius nor that of the patriarch could save
Eustratius from suspension a divinis for life.
The arguments that won the case for his enemies are preserved in the
discourse of Nicetas of Serrai at the last meeting of the synod: Eustratius was
a heresiarch, founder of a new heresy, and for such there had never been a
precedent of pardon, unless the point had not been defined by a Council, a
condition which did not apply to Eustratius’ case; moreover, the heresy had
already been present in his first works on the question of the Icons, so that it was
not true that he had never believed in it; finally, he had violated his signature on
the letter anathematising John Italus’ teaching.
Eustratius was suspended as a heresiarch in spite of imperial protection. One
can only speculate about the political intrigues surrounding the trial, but the
doctrinal implications are fairly clear. As in the case of John Italus, not only were
specific doctrines condemned, but through them also a certain spirit. Of the
twenty-four propositions summarising the errors found in Eustratius’ two libelli
against the Armenians, the last two are directly aimed at the use of dialectics in
theology. The twenty-third condemned proposition affirms the necessity of
relying on the power of dialectical reasoning (tês tekhnês ton logon kai tês
epistêmês tên epikheirêsin) in order to speak truly about the Incarnation, whereas
the last one reads: ‘That Christ in all of his sacred and divine sayings reasons in
Aristotelian syllogisms (sullogizetai aristotelikôs)’.
There is no doubt that the action was aimed at the dialectician in the
theologian Eustratius, follower of John Italus. Eustratius had inherited from
Italus the method of not relying on authorities – they are at most starting points
for a dialectical discussion – but on formal syllogistic reasoning for solving
doubtful problems. His dialogue on the question of the Icons presents
Philosynethes (the friend of tradition) and Philalethes (the friend of truth) in
what has all the characteristics of a scholastic debate, with questions and answers
and a frequent recourse to syllogisms in their several forms. The two characters
may well be interpreted as representing in abstracto the position of Leon and
Eustratius, respectively.
The treatise on the same subject is an excellent example of how Eustratius
uses dialectics in theology. It ends significantly with a formula stating that there
is still more to say by means of arguments (logismois) that ‘scratch’ (epixainousin)
the very problem, but that this will suffice to the intelligent.
The same dialectical method is characteristic of Eustratius’ discourses against
Petrus Grossolanus and of his treatise against the Armenians. This is in fact
reflected in the title of the latter: ‘Refutation of those who attribute one nature to
446 Aristotle Transformed

Christ, our most true God, on the basis of logical, physical, and theological
arguments, from which it is demonstrated necessarily (ek logikôn kai phusikôn
kai theologikôn epikheirêseôn ex hôn deiknutai anankaiôs) that Christ, my Saviour,
consists of two natures.’ The work, which introduces itself as an ‘exercise’ on an
already accepted truth, indeed contains pages and pages of sorites.
That Eustratius understood the accusation in this sense is brought out by a
passage from his exomologêsis:

He who says that one must syllogise about the indemonstrable principles of
faith, and about those things about which the saints [the recognised Fathers of
the Church] did not syllogise – as some have suspected us to say –, that he be
rejected, not only as impious, but also as stupid and silly.

The very phrasing of this curious anathema betrays Eustratius’ naturalistic bent
of mind. He seems to allow for the possibility of an unintended heresy on his
part, but not for that of having attempted a logical impossibility.
Eustratius, then, was a theologian whose use of philosophical arguments,
especially of Aristotelian syllogisms, became detrimental to his career.
His surviving philosophical work, apart from a short scientific treatise, is
limited to three commentaries, each of them explaining one book of a work of
Aristotle: one on Book 2 of the Posterior Analytics,9 one on Book 1 and one on
Book 6 of the Nicomachean Ethics.
The scientific treatise concerns metereological, cosmological and astronomical
matters and is addressed to ‘the empress Lady Anna of Alania’, the spouse of
Michael VII Ducas, who himself had been a pupil of Psellus.10 It was composed
at her request. It is extant in two manuscripts, a Venetian and Parisian one.
However, the Parisian MS lacks the beginning, has three chapters more than the
Venetian one, and is on the whole more scientific in style and content than the
latter. Since certain passages are similar to loci in Michael Psellus’ de Omnifaria
Doctrina and Solutiones quaedam ad imperatorem Michaelem Ducan, and the
treatise shares with the Solutiones a common scheme of dividing the matter into
short chapters and a dedication to a member of the imperial family, Psellus may
be the direct source of Eustratius’ little work. It looks as if the Parisian text is a
collection of material that Eustratius found available at the Academy, whereas
the Venetian version seems to be adapted to the requirements of educating the

9
CAG 21.1.
10
Edited by Paola Polesso Schiavon, ‘Un trattato inedito de meteorologia di Eustrazio di Nicea’, Rivista
di studi bizantini e neoellenici, n.s. 2–3 (xii–xiii), Atti del primo congresso nazionale, Ravenna, 23–5
maggio 1965, Rome 1965–6, 285–304.
 The Greek commentators on Aristotle’s Ethics 447

empress. Eustratius may have obtained the material in a number of ways, either
as a student or specifically at the occasion of the empress’s request. From all this
it appears that the treatise probably belongs to Eustratius’ early years and that he
was acquainted with the Byzantine imperial family before the rise of Alexius I
Comnenus. In any case, he was a person to whom women of that family would
turn for instruction.
The commentary on the Posterior Analytics confirms Eustratius’ interest in
Aristotle’s theory of syllogistic demonstration and continues the tradition of
John Italus. The commentaries on the Ethics too reveal an acquaintance with
Aristotle’s logic, especially by frequently working out arguments in their explicit
syllogistic form, often with their Aristotelian label of figure and mood attached
to them.
This procedure, common to the commentaries on Books 1 and 6, marks
Eustratius as the author of both. There is no doubt that they are the work of the
same man, and that this man is Eustratius. This is well attested by the manuscript
tradition. The generally accepted view has rightly ignored Schleiermacher’s
arguments against a common authorship of both partial commentaries on the
Ethics. I mention them only because our Latin text provides us with the answer
to Schleiermacher’s strongest objection.
After having touched on some general differences between the two books,
which can easily be explained by a lapse of years and by a growth in both
knowledge and maturity of the author, Schleiermacher claimed to have found
serious discrepancies in (a) the explanation of Aristotle’s ‘exoteric’ logoi (exôterikoi
logoi) given in the two commentaries and (b) their prologues. As regards (a),
Book 1 (111,21ff ) divides Aristotle’s treatises into acroamatic and exoteric as
follows: the former were written for the benefit of an audience, his pupils,
whereas the latter were addressed to correspondents outside the school. Book 6
(298,30–31), however, explains that the logoi expressing the common view of the
multitude outside the scientific tradition are called exoteric by Aristotle.
The second inconsistency (b), according to Schleiermacher, involves the
prologue of both commentaries. Book 6, on the one hand, starts with a formal
dedication in the usual, elaborate, highflown Byzantine style, in the second
person, addressed to a God-fearing, logos-loving, goodness-and-beauty-loving
basilis (royal lady). Robert Browning’s hypothesis (see Chapter 17 above) that
this was Anna Comnena, if not certain, is at least the best conjecture.
Schleiermacher offered as candidates the queen of Cyprus and Maria of Alania,
spouse of Michael VII Ducas, for whom Eustratius wrote the scientific treatise.
But that work dates from his early youth.
448 Aristotle Transformed

In his dedication, Eustratius, conforming to the accepted literary canon,

belittles himself as much as possible; he invokes his old age and recalls that
some time ago (pro khronou tinos) the royal lady had requested from him an
explanation of the first book of Aristotle’s Nicomachean Ethics, which task he
then performed in all due obedience. The prologue of Book 1, on the other hand,
only mentions a certain person of the highest worth, who had requested the
work; a refusal was out of the question, since this person had been a benefactor
in many a difficulty. Eustratius could not have referred to the royal lady in this
way, Schleiermacher argues, even if she had not wished to be named, and
certainly not without some sort of apology, when he finally broke the silence in
Book 6, for previously having mentioned her in such a remote manner.
The first argument can easily be disposed of. The lapse in time between the
two commentaries may be held responsible for the fact that Eustratius did not
recognise the expression ‘the exoteric logoi’ when he commented upon it for the
second time. These are not two incompatible accounts of the technical expression
‘exoteric logoi’. Only in Book 1 does the commentator understand the expression
as denoting a class of Aristotle’s works; in Book 6 he understand logoi in their
most common sense of ‘opinions’ or ‘arguments’ and the expression ‘the exoteric
logoi’ as referring to ‘exoteric beliefs’.
The second argument is a serious one. Some curious facts emerge, however,
with respect to the passage in Book 1 referring to the solicitor of the commentary
(in EN 1,13–23). First, the reference is indeed surprisingly vague, a fact which
remains puzzling even if Eustratius is not the author. Further, this is not an
independent dedication or preface, but it is grammatically conjoined to a main
sentence as a genitive absolute and in content interrupts the line of thought –
which is to place, by way of a general introduction, Aristotle’s Ethics in the
framework of the whole of his philosophy. Finally, the passage is completely
absent in Grosseteste’s translation, and must have been lacking in his Greek
manuscript.
There is no doubt, then, that this passage is an interpolation, an attempt to
make the beginning of Book 1 conform to the dedication of Book 6.11 It restricts
the commentary to the first book of the Nicomachean Ethics, whereas at this
point the introduction still envisages the whole of the Ethics. This conclusion is
confirmed by the interpolation, equally unknown to Grosseteste, of the words
hou kai tên exêgêsin apêithêmetha (in EN 4,9), ‘an interpretation of which has

11
Thus V. Rose’s intuition is vindicated: ‘despite the alteration of the prologues which has to be
assumed after a comparison of them’ (op. cit., 70).
 The Greek commentators on Aristotle’s Ethics 449

been requested from us’, referring to Book 1, to which at this point the
commentary is restricted. Both references to a request for a commentary in the
beginning of Book 1 are interpolations. The original text kept silent on that
point.
This clarification of an otherwise puzzling fact illustrates how useful the
comparison of the Greek text with the Latin translation of Grosseteste can be.
If Eustratius continued in his commentaries on the Ethics to make explicit use
of the Aristotelian syllogistic forms, he nevertheless showed himself more a
Platonist than an Aristotelian in his explanation.
This was made possible by the type of commentary Eustratius had chosen.
His exegeses of Books 1 and 6 are commentaries in the fullest sense of the word.
They are not primarily concerned with explaining particular passages, phrases,
or words in the text, but with interpreting the very scope, design, and impact of
Aristotle’s Ethics. Eustratius interprets the work from his own theological-
philosophical point of view, which is that of a Christian.
This Christian outlook pervades both commentaries. The prologue of Book 1
presents the Old Testament figures Abraham, Isaac, Jacob and Job as exemplary
heads of of households, and Moses and Jesus Naue (Joshua) as exemplary
statesmen. It contrasts the ancient sages (pagan thinkers) with ‘ourselves’
(Christians), where they ‘call the end of human life happiness, whereas we call it
beatitude’. It describes this end as the ‘union with God in the indwelling One,
which is called the flower of the Intellect by the great Dionysius’. Finally, the
prologue concludes by invoking the assistance of ‘the Cause of all good’ in this
(the commentary on Aristotle) as well as in all else. The commentary on the
opening sentence of the Nicomachean Ethics reminds the reader that ‘we are
created after the image of God’ and that it is ‘our’ aspiration to ‘behold the Word
and through It the Father’, since ‘nobody comes to the Father but through Me’.
Other passages in Book 1 refer to the Christian martyrs and to the hermits;12 the
life of the latter is even identified with the ‘life of contemplation’ of EN 1095b19.13
The prologue of Book 6 ends, as does its predecessor, by invoking God’s assistance.
The commentary on Book 6 contains a remarkable passage, several pages long,
against the carnal philosophy of the ‘pseudoprophet’, Mohammed.14
Eustratius’ Christian outlook, however, is cast in a philosophical mould. This
philosophy can aptly be described as a ‘Christian Neoplatonism’. Ethical growth

12
in EN 35, 33–7 and 7, 9–12.
13
cf. ibid. 7,9–12 with 34,23–8.
14
ibid. 272,3–277,17.
450 Aristotle Transformed

is depicted as a journey upward, during which the soul goes through four stages:15
(1) the moderation of passions; (2) the mortification of passions, leading to
‘apatheia’; (3) the participation in the Intellect; and (4) the union with God, who
is the Primary Good, identified with ‘the Good at which everything aims’ of EN
1094a3.16
Eustratius sees the life of the passions, that is, the life of enjoyment or pleasure,
as the natural life of the child, but one becomes practised in the moral virtues in
order to leave that life behind for one of civic virtue, a middle stage between the
life of the passions and that of contemplation. The Nicomachean Ethics is, for
Eustratius, an exploration of the virtues of this middle sort of life, the life of the
political man. To the following stages belong the ‘purgative or purifying’ life and
the life of contemplation.17 Aristotle provides only a few scattered hints in the
earlier books, and devotes only a few chapters in Book 10 to describing the life
of contemplation, the highest sort of life. Eustratius thus has relative freedom to
wax neoplatonistically eloquent about man’s pathway toward participation in the
true intellect and in the uninterrupted activity of reason, and finally toward the
mystical union with God.18 Eustratius sees the life of the intellect as being
particularly pursued by the Christian hermit, who seeks the happiness of unity
with God ‘in solitude’. When Aristotle comments at 1101b18, ‘We call the most
godlike men blessed and happy’, Eustratius finds the occasion to say:

These are the ones who have completely broken their relation with matter and
found life outside the flesh and the world, above all visible things; their intellect
is united with the divine natures; it converses with them and is exalted together
with them; they live always in act, while reflecting the first light in an
uninterrupted way.19

This strong support of the mystical and monastic life, in the context of a
commentary on Aristotle’s Nicomachean Ethics, was bound to influence western
writers who already tended to see mysticism and monasticism in a favourable
light.
Eustratius also develops a strong defence of Plato’s theory of Forms, in the
context of his commentary on EN 1.6.20 He thinks that when Aristotle criticises
the form of the Good in that chapter, he is inconsistent, since he has already

15
ibid. 4,25–38. See also below.
16
ibid. 6,5–8.
17
ibid. 34,19–31.
18
ibid. 4,25–8.
19
ibid. 106,19–32; cf. 63,29–33.
20
ibid. 39,24ff. See K. Giocarinis, ‘Eustratius of Nicaea’s defense of the doctrine of Ideas’, 159–204.
 The Greek commentators on Aristotle’s Ethics 451

conceded that there is ‘a good at which all things aim’, in the first chapter of the
book. He clearly interprets Aristotle’s rather semantic remark (the word ‘good’
means that at which something aims) as a definition of the transcendent Good.
Later, Eustratius offers another definition of good: good as a quality of beings,
good as the ‘order’ of beings; good is the universal ordering, the harmony, the
cosmos; and evil is lack of order and disharmony.21 The order of the universe, for
the Neoplatonist, is found both in the immanent forms, taken to be identical to
the Aristotelian species, genera, and perhaps categories, and in the Logos which
transcends those forms. The link between the two meanings is given by Eustratius
in a third definition of the good: the good of beings is a participation in the
transcendent good: participatio primi boni. There is then, pace Aristotle (EN
1096a29–34), one science of all good things qua good, which is nothing other
than metaphysics, the science of all beings qua being. Eustratius reserves for this
science the name of ‘theology’. The ultimate aim of the intellect is to know God,
to know the order, and to participate fully in the universal unity.
As Eustratius sees it, Aristotle’s criticisms of Plato are mistaken in that they
impute to Plato transcendent concepts, universals in the Aristotelian sense,
whereas Plato intended to assert the existence of causes. A.C. Lloyd attempted in
a recent article to ‘place’ Eustratius in the whole of the Aristotelian tradition
precisely with respect to his attitude towards the Platonic Ideas and more generally
with respect to his view of universals:22 ‘Eustratius shows himself an able
philosopher in the tradition passed on from Psellos to John Italos . . . This consists
in the standard combination of Aristotle and Neoplatonism, in which Aristotle’s
universals are concepts with a non-mental existence or hupostasis . . . in the
individual forms of individual substances and their accidents.’ This interpretation
of Aristotle – Lloyd calls it a form of ‘conceptualism’ – is, he claims rightly,
not specially characteristic of the Neoplatonists: ‘It would be difficult to find a
philosophically trained Byzantine who did not take it for granted.’ It originated
with the Alexandrians, ‘but Eustratius’ version of the logico-metaphysical problem
is in some respects more subtle’ than theirs.‘No one acquainted with late Byzantine
Aristotle commentators will look for originality in their suggestions’, but ‘in their

21
in EN 46,29–31.
22
A.C. Lloyd, op. cit. in n. 8 above. The author bases his account on the parallel passage of in An. Post.,
CAG 21.1, 195,5ff and in EN, CAG 20, 40,22ff, but starts from the former commentary. He does not
mention Giocarinis’ article (op. cit.), which is restricted to in EN , but says that Eustratius there, in
following ‘ “three ways of understanding wholes”, . . . uses some of his material (at 40,37–41,2) less
well than in the An. Post. commentary (196,14–16) – if he is not plain confused.’ Lloyd does refer to
P. Joannou’s work (see n. 8 above), adding some critical remarks to the latter’s ‘Die Definition des
Seins . . .’
452 Aristotle Transformed

style of exposition and their selection from traditional points of view.’ Eustratius
shows himself a follower of the Neoplatonists in keeping close to Proclus’ locus
classicus23 of ‘three ways of understanding’ universals or ‘wholes’: (i) before the
parts (hôs pro tôn merôn), (ii) out of the parts (hôs ek tôn merôn), (iii) in the part
(hôs en tôi merei) and in defending ‘not just the existence of the Ideas as archetypes
but their function, explicit in Proclus, as efficient causes’.
Eustratius’ commentary on EN 6 contains a further defence of Plato, namely,
the defence of the Socratic definition of virtue as knowledge, against EN
1144b17–21. Here Eustratius explicitly refers to the ‘Socrates’ of Plato’s Republic.24
To call Eustratius a ‘Christian’ Neoplatonist, however, is to deny that he was
simply a Platonist or Plotinian Neoplatonist. His long and passionate defence of
Plato in Book 1 concludes paradoxically: ‘We do not say this as proponents of the
Ideas, since those who wish to abolish the Ideas may have no difficulty refuting
these words.’ Eustratius no doubt felt that several aspects in regard to his
exposition of Plato’s theory of Ideas had to be reinterpreted in a Christian light,
e.g. the place and function of the Demiurge, the relation of the Ideas to God, the
relation of first being to the One, and the doctrine of emanation. He probably
wanted to avoid the fate of Psellus, who had been accused of being a follower of
Plato against Christ. Eustratius must have felt also that such a reinterpretation of
Plato could hardly be undertaken in a commentary on Aristotle’s Ethics. He
agrees with Aristotle that the problem of the Idea of the Good is a metaphysical
– Eustratius says ‘theological’ – one and falls outside the scope of ethics.
This leads to another observation concerning Eustratius as a commentator.
Whereas he identifies, in a truly Platonic fashion, metaphysics and theology, he
never so much as hints at a possible distinction between what has later been
called ‘speculative’, ‘philosophical’, or ‘natural’ theology on the one hand and
‘positive’ theology, based on revelation, on the other.
To summarise, Eustratius’ commentaries on Books 1 and 6 of the Ethics suffer
from the typically Byzantine defect of being long-winded and verbose and they
are not impressively systematic in structure, but they are very important as an
interpretation of the Nicomachean Ethics from the point of view of a Christian
Neoplatonist. By the same token, however, they are defective as a tool for a better
understanding of Aristotle’s genuine thoughts. Readers with different points of
view have appreciated the work differently. Schleiermacher, for instance, has
nothing but praise, while Gauthier finds much to blame.25

23
Proclus Inst. 67. Cf. Joannes Italos Quaestiones Quodlibet, P. Joannou (ed.), 7–8; 16–17.
24
in EN 400,22–401,20.
25
cf. R.A. Gauthier, L’Éthique à Nicomaque2 . . ., vol. 1, part 1, 105.
 The Greek commentators on Aristotle’s Ethics 453

Such was the work and the life of the man whom Anna Comnena has
described as ‘skilled in the sacred and the profane sciences, more confident in
dialectics than those who frequent the Stoa and the Academy’.26

The ancient anonymous scholia

The anonymous ‘commentary’ on Books 2, 3, 4 and 5 differs greatly from that of

Eustratius. It is referred to in the book titles of the manuscripts as skholia. It is
indeed a collection of scholia, that is to say explanatory notes of various content,
which follow the text of the Ethics very closely, instead of a commentary
concerned with the overall philosophical interpretation of the work. The scholia
explain the meaning and use of certain words, give a wealth of historical and
literary information by way of illustration (including quotations from classical
sources, some of which are otherwise unknown or lost), or are philosophical in
content – that is, they clarify an argument, explain technical notions, compare a
Peripatetic thesis with or defend it against the view of a rival school: Plato, the
Epicureans, the Stoics, or others. Scholia of the last type are especially useful for
a better understanding of the text of Aristotle, since they are thoroughly
Aristotelian in spirit. The scholia contain a considerable number of references to
Plato, who is often cited in support of Aristotle, to other works of Aristotle, to
Theophrastus, and to the Stoics, who are openly attacked.
It is always difficult to date scholia. They may have accumulated over many
years and be the work of several scholiasts; they are a collection of notes and
excerpts from various sources. The evidence in this case, however, strongly
suggests that the scholia on Books 2, 3, 4, and 5 were compiled by one man, or at
least in one school, just before the end of the second century of our era. They
show no Christian influence, except perhaps in Book 5, where one finds the three
words ‘So also Absalom’ at the end of the story of Phoenix’s liaison with his
father’s concubine.27 They form, no doubt, a late gloss, which it was easy to add at
the end of the paragraph.28
There are no quotations or references to any author later than Lucian. This
negative argument finds a strong confirmation in a positive time index, contained

26
Anna Comnena Alexias xiv,8.
27
in EN 229,22. For the story of Phoenix see: Iliad 9. 452; of Absalom: II Sam. 16:21–2.
28
Rose’s supposition (op. cit., 82) that this was an addition by Robert Grosseteste can only be
understood by realising that he did not have access to the Greek text of the anonymous scholia on
Book 5. Paul Moraux’s remark, ‘his statements about the origin of this addition would have to have
been checked, however’ (Der Aristotelismus bei den Griechen, vol. 2, 329 n. 146), has no ground.
454 Aristotle Transformed

in Book 5. The passage defends the Aristotelian doctrine that corporeal and
external goods are to a degree necessary for happiness against the Stoic doctrine
which stresses the all-sufficiency of virtue. In the very beginning of Book 4, the
scholiast had already presented an argument against the division of virtue as the
only good, vice as the only evil, and the rest, including wealth, as being neither
good nor bad, but only potentially good or bad, depending on the use one makes
of it. Here the doctrine of ta adiaphora, things indifferent, that is to say,
intermediary between good and bad, is attacked again. Justice, whether
distributive or retributive, must be useless for the proponents of the adiaphora-
theory, ‘among whom previously Aristonymus belonged, but now even some
who pretend to be Platonists, while surreptitiously introducing this doctrine,
among whom also Atticus appears to belong’.29 The attacks on Aristotle by the
Middle Platonist Atticus, including the above extreme doctrine, are well known
from Eusebius. Atticus was active under Marcus Aurelius (ad 161–180).
The strict Aristotelian character of the scholia suggests also that their final
form dates from the period before the school was to all intents and purposes
absorbed in Neoplatonism; that is, before the end of the third century.
On the other hand, the references to Lucian and Atticus place the time of final
composition of our ‘commentary’ definitely after Aspasius, the Peripatetic
commentator of the first half of the second century. Since Aspasius’ commentary
has some material in common with the anonymous scholia, the latter must contain
excerpts from the former, unless they have both drawn from a third source.

The excerpts from Adrastus’ monograph

A. Kenny suggested in 1987 the identification of the anonymous scholiast on

Books 2–4 as Adrastus of Aphrodisias.30 That Adrastus wrote a book on the
Nicomachean Ethics is asserted by Athenaeus.31 However, Adrastus lived in the
first half of the second century and can hardly have made a reference to Atticus

29
in EN 248,24–6 (italics mine).
30
Anthony Kenny, The Aristotelian Ethics. A study of the relationship between the Eudemian and
Nicomachean Ethics of Aristotle, Oxford 1978, 37 n. 3. Kenny overplays his hand in trying to make
the anonymous scholia fit the report on Adrastus’ work (see next note), for he first insinuates that
Athenaeus speaks of six books of Aristotle’s Ethics and then goes on to argue very cleverly that the
second-century collection of scholia extended over exactly six books. The facts are that Athenaeus
mentions only one book, be it a sixth one, of Adrastus on the Nicomachean Ethics and that only four
books of the ancient anonymous scholia are extant.
31
Athenaeus 15,673e–f: ‘for our fine Adrastus has published five books on the historical and literary
questions in Theophrastus’ Characters and a sixth on those in the Nicomachean Ethics of Aristotle.’
 The Greek commentators on Aristotle’s Ethics 455

or have quoted Lucian. Moreover, the scholia are clearly a compilation from
various sources. On the other hand, a good deal of them show unmistakably the
characteristics ascribed to Adrastus’ monograph, Peri tôn kath’ historian kai lexin
zêtoumenôn (On historical and literary questions), in that work. This has been
abundantly shown by P. Moraux.32
Concerning Adrastus’ chronology, Moraux has established that his
commentary on Plato’s Timaeus dates from before the year 147, whereas Galen’s
reference to the commentaries of Adrastus and Aspasius on Aristotle’s Categories
from about the middle of the century suggests that Adrastus was a perhaps
younger contemporary of Aspasius’. This suggestion arises from the fact that he
is named with, and before, Aspasius and that Galen ‘heard’ in his youth an
unnamed pupil of the latter’s in Pergamon.
Adrastus is best known for his influential commentary on the Timaeus,
in which he incorporated contemporary theories of mathematical astronomy
(eccentric spheres or epicycles) into the Platonic-Aristotelian cosmology, and for
his On the Order of Aristotle’s Works, quoted more than once by Simplicius and
being chiefly philological in character. Nothing remains from his commentary
on the Categories. Galen recommends it, together with that of Aspasius, for
new students in philosophy, and Porphyry mentions that Plotinus used their
hupomnêmata among those of others, without explicitly mentioning the
Categories. An important fragment from what in all likelihood was a commentary
on Aristotle’s Physics has been transmitted by Simplicius through Porphyry.
Adrastus’ five books on Theophrastus’ Characters and the sixth one on the
Nicomachean Ethics of Aristotle have already been mentioned, together with
their historical and literary nature. Moraux was rightly impressed by the richness
and scholarly tenor of the anonymous scholia that are of this kind, in contrast
with the fairly indifferent character of those that are not.
Moraux identified and described in detail the passages that qualify as excerpts
from Adrastus’ work.33 He did this by going through Heylbut’s Index nominum
(and probably also the Index verborum) in so far as it refers to the pages
containing the anonymous scholia. These are the scholia exhibiting a philological,
literary or historical bias. They differ from those that contain discussions with
other schools of philosophy, offer philosophical analyses, or simply paraphrase
Aristotle’s text. The excerpts from Adrastus show an interest in Atticism, a good

32
Paul Moraux, Der Aristotelismus bei den Griechen vol. 2, covers Adrastus, 294–332: chronology,
294–5; commentary on the Timaeus, 296–313; On the Order of Aristotle’s works, 314–17;
interpretation of Aristotle, 317–23; monograph on the Ethics, 323–30.
33
327–30.
456 Aristotle Transformed

acquaintance with Theophrastus’ ethical works, but above all with Attic tragedy,
with comedy, and with historical sources.
In my further account of the ancient scholia book by book, I shall confine
myself to identifying the (groups of) scholia that may be considered as excerpts
from Adrastus’ monograph. The reference to Lucian may be seen as a later
addition in an otherwise Adrastan context, but the one to Atticus must belong to
a scholion written entirely by the compiler himself. Although nothing excludes
the possibility that some gloss has been added later on, the only evidence for
such an addition is contained in the note on Absalom.

The scholia on Book 2

The scholia on Book 2 have the following characteristics. First, they do not form
a comprehensive commentary. The text of the scholia is less than twice the length
of Aristotle’s text, although each gloss gives a rather detailed explanation of or
comment on an expression in the text. Secondly, two alternative explanations are
often proffered. This suggests a conflation of two scholia from different sources.
Thirdly, there are occurrences of a definition of a word being followed, a few
pages later, by the introduction of another definition, based on a different
meaning. This again suggests a compilation from different sources. Fourthly,
sometimes the same thought is repeated in different words, no doubt because the
scholiast had not perceived the identity in meaning or else wanted to preserve the
difference in form: another indication of a compilation from different sources.
As to their content, the glosses frequently interpret the text of Aristotle by
adding a word or a phrase that was, supposedly, understood but not expressed by
Aristotle or that lays stress on an otherwise inconspicuous expression of his.
Other glosses give synonyms, offer definitions, or explain by means of typically
Aristotelian notions or principles. Elsewhere concrete examples or literary
quotations illustrate the meaning of the text. They are most likely taken from
Adrastus.34 Finally, the scholiast expresses several times approval of a thought or
an expression in the Ethics.35

34
Moraux (329 n. 146) rightly judged that the attribution of the verses from Iliad 18.109–10 to
Heraclitus (in EN 129,1–3) was a mistake of the copyist (probably the compiler) who was confused
by the philosopher’s name at 128,32 (in fact EN 2.3, 1105a8).
35
e.g., in EN 122,9: kalôs de touto (good). The approval is often expressed by means of the words kalôs
(good) (see also in EN 125,16 and 138,24), but not always: see for instance in EN 126,21: panu
anthrôpinôs (very humane).
 The Greek commentators on Aristotle’s Ethics 457

Plato is referred to three times, but the first reference is taken from Aristotle’s
text.36 The scholiast remarks on the strength of EN 1104b24–5 that the doctrine
which defines virtue as apatheia is older than the Stoa, and notes Aristotle’s
correction: virtue is not an unqualified apatheia, but a qualified one, that is, only
in regard to faulty passions. Some Platonist is reported to have applauded the
Stoic view as a Platonic thesis.37
On the whole, the short collection of scholia on Book 2 stays very close to
both the letter and the spirit of Aristotle’s text and shows some concern with the
rival school of the Stoa. The attitude toward the Academy is ambiguous as was
that of Aristotle himself: a mixture of admiration for Plato and criticism of the
Platonists.

The scholia on Book 3

The scholia on Book 3, although more numerous than those on Book 2, have
similar characteristics. However, the literary and historical references are, even
in proportion, much more prominent. The scholiast has no doubt made extensive
use of Adrastus’ commentary.
In the first place, explanations are given for such references in the text of
Aristotle. Alcmaeon killed his mother in Euripides’ play (EN 1110a28) because
his father had ordered it. The scholiast cites two different sources to explain why
Alcmaeon obeyed the unusual order, thus revealing that he had no access to the
text of the play: according to one source the motive was pity (not to distress his
father), according to the other, which quotes a verse from the play, it was fear (to
avoid his father’s curses).38 Aeschylus is said to have divulged the Mysteries (EN

36
in EN 127,5; 136,28; 137,8.
37
ibid., 127,7.
38
The whole scholion is probably Adrastus’. The text (in EN 142,22ff ) is not very sound. Heylbut,
following the Aldine edition (= a), drops touton (attested by B) before hôi on line 22, but Robert
Grosseteste read touton, phasin, hôi (hunc aiunt cui). This is the commonly accepted view, against
which the scholiast cites another authority (line 24). Heylbut has: allôs. parentithetai ton <par’>
Euripidêi Alkmaiôna hôs. . . . Both a and B read Euripidên, the preposition is wanting in B. Again,
Grosseteste seems to have had the correct reading: allos palin tithetai ton par’ Euripidêi Alkmaiôna
hôs . . . (Alius rursus ponit apud Euripidem Alcmeona ut . . .). What has happened is clear. The
forgotten preposition, whose disappearance has been the cause of the accusative Euripidên, must
have been corrected in the margin, from where it has been taken to replace palin. There are many
disturbances in the text, for instance, the last three words of the preceding sentence are completely
turned around in a and Grosseteste’s original as against B, this time followed by Heylbut. That the
scholiast and not just the second source is likely to be Adrastus is suggested by the construction
phasin, which we find elsewhere used in similar circumstances, e.g. in the discussion about the
meaning of ponêros with the accent on the last syllable at in EN 1113b14 and of the Atticism
ponêros with the accent on the first syllable, where we read phasi . . . alla phamen . . . (in EN 155,4–8).
458 Aristotle Transformed

1111a10) in several of his plays, named by the scholiast, who refers to the
historian Heraclides Ponticus for the story of Aeschylus’ trial and acquittal and
cites his epitaph. The error of Merope (EN 1111a12) is explained with reference
to Euripides’ lost play Cresphon. The Homeric quotations in Aristotle (EN
1116a23–6 and 34–5) are identified and completed.
Further, Aristotle’s sayings are illustrated by concrete examples, most of which
are mythological or historical: for instance, Zopyrus, Odysseus, and Dolon, each
in a different way, did something distasteful in order to obtain a certain good,
whereas Anaxarchus and Zeno bravely suffered torture without being swayed
into committing a disgraceful act. The Iliad is a favourite source for such
examples. Further quotations in various contexts are: one from a third lost play
by Euripides, the Erechtheus; two from the comedian Epicharmus, one of them
from a play called Hêraklês ho para Pholôi; one from Hesiod’s Hai megalai Êoiai;
one from Isocrates’ Panegyric; and one from Lucian’s Somnium. There is a
reference to Pittacus, ‘one of the seven wise men’, and one to the lyric poet
Tyrtaeus. A considerable number of historians are appealed to, among others
Herodotus, Cephisodorus, Anaximenes, and Ephorus.
All the material discussed in the last two paragraphs, including the quotation
from Lucian, coincides with whatever is known of Adrastus’ interest in his book
on Aristotle’s Nicomachean Ethics. There are also several references to Plato. The
most important one concerns Aristotle’s argument that man is responsible either
for both the good and the evil he does or for neither; that is, either both good and
evil actions are done intentionally (hekousiôs) or neither are. The scholiast notes
that this dilemma is directed against Plato. Thus, the Socratic paradox is ascribed
to Plato himself.39
Whereas Epicharmus is quoted as placing man’s good in labour, Callicles ‘in
Plato’ and Epicurus are cited as placing it in pleasure. In this connection
Protagoras’ thesis of relativism is brought in, but the question is resolved in an
Aristotelian manner: each person, according to his acquired hexis, will strive
either after the genuine good or after a seeming good.40 Also cited is Epicurus’
division of desires into necessary and natural, not necessary and natural, and
neither necessary nor natural,41 but Aristotle is said to have used such a
distinction before the Epicureans.

39
in EN 143,26ff.
40
ibid. 153,19–154,8.
41
ibid. 171, 23–8.
 The Greek commentators on Aristotle’s Ethics 459

There are references to the Topics and to other works of Aristotle’s, including
one to his treatise On Pleasure,42 probably his treatment of the subject in Book
10 of the Nicomachean Ethics.
One scholion appears to be a gloss on the one that precedes it: ‘Fate (to
heimarmenon) also would be classified by these people under nature, for it is
neither unalterable nor necessary.’43 The present quotation and many other
discussions about nature reveal the preoccupation of the scholiast with the Stoa.
Earlier on, in the beginning of the book, a discussion on whether ‘voluntary’
and ‘involuntary’ have each one or more senses seems to reflect imperfectly a
similar discussion that has found its way into problem 11 of the Ethical Problems,
attributed to Alexander of Aphrodisias.44
Finally, there is a distinct interest in linguistic usage, especially in Atticism,
e.g. when the use of ponêros, with the acute accent on the first syllable, is
explained as an Atticism.45 In the discussion on the meaning of ouden pro ergou,
EN 1113b27, the Attic custom is appealed to, together with Isocrates.46 Such
passages again show Adrastus’ hand.47
To sum up, the scholia on Book 3 show the same characteristics as those on
Book 2; they contain, however, more references to and quotations from various
literary and historical sources, most likely culled from Adrastus, and they are
engaged in philosophical discussions, not only with the Stoa, but also with Plato,
Epicurus, and several minor schools.
Concerning the philosophical interpretation of Aristotle’s text, it is worth
mentioning that F. Becchi has pointed out the positive role attributed by the
anonymous scholiast to the passions in the constitution of moral virtue.48 Becchi
draws attention, for instance, to in EN 130,25: ‘For passions (pathê) are the
starting points for actions’, which he contrasts with EN 1139a31, where deliberate

42
ibid. 154,15–16.
43
ibid. 150,2–4.
44
See R.W. Sharples, ‘Ambiguity and opposition: Alexander of Aphrodisias, Ethical Problems 11’,
Bulletin of the Institute of Classical Studies 32, 1985, 109–16.
45
in EN 155,1–17. This passage contains already mentioned quotations from Hesiod and Epicharmus
of Syracuse.
46
From the Panegyric, as is already mentioned. The work is not named, however, probably because it
was well known.
47
We find the quotation from Lucian neatly tacked on the last of these scholia. If it is an addition from
a pupil of Adrastus’, some of the other ‘Adrastan’ notes may be his too.
48
F. Becchi, ‘Variazioni funzionali nei Magna Moralia: La virtù come impulso razionale al bene’,
Prometheus 6, 1980, 201–26, at 225. Becchi has developed the concept of ‘functional Aristotelianism’
in a series of articles in Prometheus between 1975 and 1983 on Peripatetic ethics, from the
anonymous author of Magna Moralia to Plutarch’s de Virtute Morali. This is an Aristotelianism that
has adapted itself to the problems raised by its confronting rival schools. Becchi therefore questions
the labelling of certain key notions in late Peripatetic ethical writings as Middle Platonist.
460 Aristotle Transformed

choice (proairesis) is said to be the principle of action. We should see in this, he

maintains, a functional adaptation of Aristotelianism in order successfully to
answer rival theories. He cites similar passages in the Magna Moralia, Stobaeus,
Aspasius and Alexander of Aphrodisias, and thinks on tenuous grounds that the
change goes back to Theophrastus.49 The scholion, however, does not contradict
1139a31, because Aristotle himself stresses the dual nature of deliberate choice
(1113a11), a mixture of deliberation and striving (orexis). The latter is the generic
term for the activity (or its faculty) of the irrational part of the soul that ‘in a way
partakes of reason insofar as it listens to it and lets itself be persuaded by it’
(1102b30). It is this non-rational element that the scholiast calls passion (pathos),
as is clear from a scholion in Book 4 (in EN 180,24–5): ‘deliberate choice is
deliberative striving (=EN 1113a11), and striving is a passion.’ Aristotle never
calls striving (orexis) (nor willing, boulêsis) a passion, but its species appetite
(epithumia) definitely is one of the passions. To extend the scope of passion to
the whole domain of striving may be a matter of ‘functional’ adaptation; it is
certainly a question of interpretation.

The scholia on Book 4

The scholia on Book 4 resemble those on 2 and 3. The existence of two scholia on
1126a21,‘For revenge quells anger’ – the second one being out of place – confirms
the hypothesis that this commentary is also a compilation of glosses from various
sources. On the other hand, there are signs that certain strings of scholia are the
work of one scholiast, for instance, the many references to the second book of
the Nicomachean Ethics; one scholiast seems constantly to have kept his eyes on
the summary treatment there of the subject matter of Book 4; and the scholiast
of in EN 178,30 explicitly refers to an earlier note of his on Book 3: ‘As we also
pointed out above in the section on courage . . .’50
The following information will give some idea of the kinds of scholia one
encounters here.
Let us consider as a sample the first three glosses on Book 4.51 The first gloss,
on EN 1119b21–6, is a methodological one: that, concerning which a virtuous

49
It is doubtful that these philosophical discussions have their source in Adrastus, whereas the
reference to the moral works of Theophrastus – the only hard evidence Becchi cites – concerns only
the character of Simonides, in a scholion that is almost certainly from Adrastus (in EN 180,16–17).
50
in EN 168,5ff.
51
ibid. 176–7.
 The Greek commentators on Aristotle’s Ethics 461

man is praised, is the subject matter (the-things-about-which: ta peri ha) of his

virtue, according to Aristotle; for instance, for liberality it is the giving (and to an
extent the acquisition) of money.
The second gloss, on EN 1119b26–7, is a semantic one: khrêma means properly
(kuriôs) nomisma (money) but is also used loosely for pragma (any act, deed,
affair, or thing). The philological character of this gloss points to Adrastus.
The third gloss, on EN 1120a4–5, is a long and interesting one. It explains hôn
de esti khreia as meaning hôn gar esti khrêsis connects khreia (use); with khrômai
(I use); and establishes the difference between khrêsis (use) and to khrêsimon (the
useful): how we use something is its use, whereas what we thus use is useful. The
point, however, is not a linguistic but a philosophical one. The subject matter of a
virtue belongs to the class of the useful, whereas virtue and vice belong to the
class of uses. The subject matter of each virtue is put to a good use if it is used well
(kalôs) and to a bad use if used badly (kakôs). This invalidates the argument of
those (the Stoics) who maintain that virtue alone is good and vice alone evil on
the ground that no bad use can be made of virtue nor a good use of vice, but that
wealth and other things – called potentially good – are neither good nor evil, on
the ground that they can be used either way. According to the scholiast, the
argument shows only that virtue is not evil and vice not a good. Moreover, it leads
to the conclusion, not that wealth etc. are neither good nor evil, but that they are
both. Finally, if what is used badly is either bad or neither good nor bad, Socrates,
who was ‘used badly’ by the Athenians, would be either bad or neither good nor
bad. This last note is probably an argument ad hominem against the Stoics, whose
devotion to Socrates was well known. If this is the case, then the following lines
are an addition: ‘So would Anaxarchus, who was “badly used” by the tyrant
Nicocreon; so also would Zeno’, an echo taken from the gloss in Book 2 that
connects the fate of these two philosophers.52 Given the philosophical and
polemical nature of this passage, it is most likely not taken from Adrastus as such,
but it contains philological and historical material that is typical of his work.
One further gloss, also relating to the Stoics, seems to have two sources:53

(i) For not all lupê (grief, pain) nor all pathos (emotion, passion) is contrary to
right reason, as some (tines) think, but one can also be grieved ‘in the right
way’, that is, reasonably; but nothing that is done reasonably is sinful
(hêmartêmenon).
(ii) This is the lupê which the Stoics instead of lupê call sustolê (contraction).

52
ibid. 143,1–3.
53
ibid. 180,12–15.
462 Aristotle Transformed

The gloss is on EN 1121a2: ‘He will be grieved (lupêsetai), but moderately and in
the right way.’ The first part (i) contains the traditional complaint against the
Stoics’ attitude toward pathos, the word being understood in the Peripatetic
sense; the second (ii) is a reply drawn from a Stoic source. The word tines in
(i) is probably an emendation.
One gloss testifies to the existence of the variants ê (or) and kai (and) of EN
1126a28 in the early manuscript tradition of the Ethics.
An explanation of Aristotle’s terms for ‘avaricious people’, to which the
scholiast adds two terms of his own, and a further differentiation of the wicked
man (akolastos) into eight types, each with its proper definition, show a
Theophrastan interest in character types and terminology.54 These scholia are
almost certainly taken from Adrastus, as are all those on Book 4 that will be
mentioned from here on, except those referring to Plato’s Republic.
Various quotations from and references to comedians occur in two passages
or scholia.55 The first contains quotations from Ecphantides, ‘the oldest poet of
the ancients’, and from Aristophanes’ Wasps. It announces also a quotation from
Myrtilus’ Titanopanes, but the verses that follow are by Eupolis. Heylbut rightly
suspects a lacuna here. The second passage quotes from Araros’ Kampuliôn and
from Plato Comicus’ Elders, and refers to Aristophanes’ Thesmophoriazusae. In
addition to the comedians, it refers to Clitarchus, the glossographer, and to the
historian Xenophon.
Elsewhere another historian, Callisthenes, is referred to en têi prôtêi tôn
Hellênikôn,56 and the poet Simonides is said to be remembered as a lover of
money by Theophrastus in his On Customs and On Wealth.57
Finally, the glosses on Book 4 contain many references to Homer’s Iliad and
to Plato’s Republic.
Kenny remarks that the last scholion on Book 4 ‘appears to promise a
discussion of continence by Aristotle in the book after next, rather than as we
would expect after two intervening books’.58 However, the scholiast says only that
Aristotle will discuss continence ‘after the next book’, adding: ‘for he stretches out
the whole treatise on justice over the book after this one.’ The note is a simple
paraphrase of Aristotle’s closing sentence: ‘Neither is continence a virtue, but
something mixed. This will be shown later on. But let us now speak of justice.’

54
ibid. 182,25–9 and 6–13.
55
ibid. 186,10–20 and 200,8–18.
56
ibid. 189,13.
57
See above.
58
A. Kenny, The Aristotelian Ethics, Oxford 1978, 37.
 The Greek commentators on Aristotle’s Ethics 463

The scholia on Book 5

That the anonymous commentary on Book 5 is a compilation of scholia, like its

predecessors, is clearly expressed in its title and is further substantiated by
its selectiveness. Only certain phrases from Aristotle’s text are explained or
commented upon. There is, for instance, no introduction proper. The first gloss
bypasses Aristotle’s opening sentence and concentrates on the methodological
implications of the second sentence. Another sign is again the occurrence of
mutually exclusive explanations of one and the same notion or phrase – for
instance, the twofold explanation of EN 1132b9, with the typical connective
ê houtôs (‘or thus:’).59 Finally, several glosses appear out of place.60
Stylistic elements, as well as matters of content, connect the scholia on Book
5 with those on 2 to 4. The use of ‘such is what he says’ (in EN 211,6) continues
a practice which is found, for instance, in glosses on Book 4.61 The frequent
approval of Aristotle’s expressions is another phenomenon that Book 5 has in
common with preceding glosses.62 I have already mentioned the scholion on
Book 5 which attacks the Stoic indifferents-theory and related a similar gloss in
the beginning of Book 4 to it.63 This is the scholion that contains the time
reference to the middle-Platonist Atticus.
The overall characteristics of these glosses show that they belong to the same
compilation as those on Books 2 to 4: they contain a wealth of material, no doubt
taken from Adrastus’ work, as is obvious from the following examples. There are
two quotations from Euripides.64 Three lines are quoted from the elegiac poet
Theognis in relation to the proverb of EN 1129b28–9 and the source of this
information is given as Theophrastus’ On Customs; but, the gloss continues, the
same Theophrastus quotes this proverb in the first Book of his Ethics as occurring
in Phocylides. Another scholion explains that Bias is ‘one of the seven wise men’

59
in EN 221,24.
60
in EN 234 contains the same gloss twice with only slight variations (lines 7–18 correspond to 27–
34) and in EN 254,28–255,1 belongs with in EN 254,12 (before alla pleuritin). That this is not just a
matter of scribal errors is manifest from in EN 246,11, where the editor notes ‘disturbed order’ in
his critical apparatus and refers to in EN 244,36. The gloss in EN 246 is obviously conceived in the
way it appears there: words of two different passages are cited and commented upon together. The
second part of the gloss cannot be detached and moved back. There is no ‘disturbed order’ in the
sense of a mistake in the manuscript tradition. The only disturbing factor is that the earlier passage
is already commented upon by another gloss. This shows the piecemeal character of the compilation.
61
in EN 202,18 and 203,14.
62
As was mentioned in relation to Book 2, see above, n. 35. An example is eulogôs prosethêke (in EN
246,13). See also for Book 4: in EN 203,8.
63
See above.
64
in EN 240,29–30. They are said to belong to the Bellerophon, but Welcker thought they came from
the Alcmaeon (Nauck FTG 2, 381, n. 2).
464 Aristotle Transformed

and cites two others, Pittacus of Mytilene and Chilo of Sparta, each with one of
his famous sayings. One scholion refers by name to Aristotle’s Magna Moralia,
whereas another refers to the same work simply as allakhou (elsewhere). The
former scholion also attributes the verse of EN 1132b27 to Hesiod and quotes
the full sentence as it appeared in his The Great Works. There is also a reference
to Aristotle’s Politics.
The examples which Aristotle gives of what is just by convention in EN
1134b21ff are glossed as follows. The agreement between Sparta and Athens
to fix the ransom for prisoners of war to one mina is illustrated by a quotation
from the historian Androtion. That a goat but not two sheep shall be sacrificed
is not attested by any historical account, according to the scholiast. Finally,
that sacrifices be offered to Brasidas is said to be a decision of the citizens of
Amphipolis, who celebrated him as a ‘hero’.
The scholia on Book 5 again contain references to Homer’s Iliad,65 and to
Plato’s Republic. All of the scholia touched upon in the last three paragraphs may
be traced back to Adrastus, with the exception of those referring to works of
Plato and Aristotle.

Michael of Ephesus on Books 5, 9 and 10

Michael of Ephesus is the author of the second commentary on Book 5, as well

as of the skholai on 9 and 10. This is attested by the titles in the manuscript
tradition and confirmed by the character of the commentaries themselves: they
agree in method, style, language, and areas of interest with the other works of the
same commentator.66
Michael was a native of Ephesus, according to his own testimony in the
commentary on Book 10 of the Ethics (first mentioned by Schleiermacher),
where he calls Heraclitus a fellow citizen.67
Almost nothing is known of the man, apart from what can be found in or
inferred from his own testimony, as Praechter remarks; and that information is
very scarce and of little importance. As a result, the date of Michael of Ephesus’

65
For instance, the one to Phoenix’s intercourse with his father’s concubine, which is followed by the
insertion ‘So also Absalom’.
66
For Michael of Ephesus, see K. Praechter, review of Michaelis Ephesii In libros De partibus animalium
. . . Commentaria, M. Hayduck (ed.) CAG 22.2, in Göttingische gelehrte Anzeigen 168, 1906, 861–
907; id., review of Comm. in Ar. Graeca, in Byzantinische Zeitschrift 18, 1909, 535–7 (= Chapter 2
above); id., ‘Michael von Ephesos und Psellos’, Byzantische Zeitschrift 31, 1931, 1–12.
67
in EN 570,21.
 The Greek commentators on Aristotle’s Ethics 465

activity has been a matter of considerable puzzlement, until, in 1962, R. Browning

drew the attention of the scholarly world to the funeral oration of Anna Comnena
by a George Tornikès, metropolitan bishop of Ephesus.68 In 1970 followed the
edition and translation of the oration.69
The passage concerning Michael of Ephesus reads:70

I myself heard the scholar from Ephesus impute to her the cause of his blindness,
i.e. his having devoted entire nights without sleep to his commentaries on the
words of Aristotle by her command; hence his eye-troubles, brought on by the
candles through desiccation.

Quite rightly, Darrouzès notes:71

The scholar of Ephesus (a native, not the metropolitan bishop) cannot be other
than Michael, who in fact commented on EN 5, 9, 10, on the Parva Naturalia, on
PA, MA and IA . The testimony of Tornikès, who knew the scholar personally
and seems to treat him as no longer alive, allows us at least to fix Michael’s
activity in the first half of the [twelfth] century . . .

This piece of evidence put an end to fruitless speculations concerning the dating
of Michael’s life and works. Still, it is worthwhile investigating why so competent
a scholar as Praechter went wrong in his paper ‘Michael von Ephesos und Psellos’,
which was mainly devoted to this subject. Praechter placed Michael’s activity an
entire century earlier, since he thought that most of his commentaries on
Aristotle were written before 1040. He thought that they were used in, and so
must precede, a certain anonymous summary of logic and of the four branches
of the quadrivium,72 the oldest dated manuscript of which was completed in
January 6548 (ad 1040). This summary (Anonymous Heiberg) contains passages
which are textually identical with Michael’s commentary on the Sophistici
Elenchi. Anonymous Heiberg contains a number of expressions (‘he says’, ‘he
inferred’) which show that the common passages are excerpts from a commentary
on the Sophistici Elenchi. These expressions are present also in Michael’s
commentary, which moreover contains cross references and other characteristics,
fully warranting the assumption that the common passages form an integral and
therefore original part of it. The commentary in question is certainly Michael’s

68
R. Browning, Chapter 17 above.
69
Jean Darrouzès (ed.), Georges et Démétrios Tornikès: Lettres et discours. Introduction, texte, analyses,
traduction et notes, Paris, CNRS , 1970, 220–323.
70
283,9–12, translated also at the end of Chapter 17.
71
282, n. 70.
72
Anonymi Logica et Quadrivium cum scholiis antiquis, J.L. Heiberg (ed.) (Det Kgl. Danske
Videnskabernes Selskab. Historisk-filolog. Meddelelser, xv, 1), Copenhagen, 1929.
466 Aristotle Transformed

work: it agrees in method, style, and language with his other commentaries on
Aristotle.
Now, Michael refers explicitly in his commentary on the Sophistici Elenchi to
his own commentaries on Aristotle’s logic (more specifically the Analytics and
the Topics), Rhetoric and Physics. Further, a comparison between the two shows
that his commentary on the Parva Naturalia preceded the one on the Sophistici
Elenchi. This, Praechter argued, is an important conclusion, since the commentary
on the Parva Naturalia contains a list of commentaries on Aristotle which
Michael had already written at that time, i.e. those on de Partibus Animalium, de
Motu Animalium, de Incessu Animalium, de Generatione Animalium, and the
second half of the Metaphysics, while promising a new one on de Coloribus (a
spurious treatise). The inclusion of the commentary on the Metaphysics shows
that Michael did not restrict himself here to Aristotle’s physical works. It is
therefore remarkable that he makes no mention of his work on the Organon or
on the Rhetoric, let alone on the Ethics or the Politics. Praechter concluded that
young Michael’s first interest focused on physics and resulted in a group of
commentaries on Aristotle’s physical treatises, not including the Physics. The
commentary on the Physics must have been written later. This first group includes
also the commentary on the Metaphysics. A second group relates to the Organon
– including at least the Prior Analytics, Topics and Sophistici Elenchi – Rhetoric
and Physics. All these works, Praechter argued, must have been written some
time before January 1040. The commentaries on the Ethics and Politics, written
later, form a third group.
Praechter may be right about the internal relationship of Michael’s work, but
he was wrong to suppose that Anonymous Heiberg depended on Michael. By an
exhaustive study of the origins of Michael’s commentary, Sten Ebbesen has been
able to distinguish earlier recensions by Michael of his commentary on the
Sophistici Elenchi.73 Taking these into account, Ebbesen has shown that
Anonymous Heiberg does not depend on Michael, but that, at least in part, they
draw from a common tradition of earlier scholia. Some of the scholia which they
have in common were derived by Michael via Psellus’ Brevis Traditio. Ebbesen
does not exclude the possibility that in other cases Michael drew on Anonymous
Heiberg. But the important point is that the relation of dependency does not go
the other way round, so that there is no need to date Michael earlier than
Anonymous Heiberg.

73
Sten Ebbesen, CLCAG 7.1, 268–85 (V.14 Michael of Ephesus); see also Chapter 19 below.
 The Greek commentators on Aristotle’s Ethics 467

Ebbesen’s study incidentally reveals that Michael’s technique in the Sophistici

Elenchi commentary, like that in his Ethics commentary, is a painstaking
compilation and editing of earlier scholia, motivated by nothing so much as the
desire to elucidate Aristotle.
Tornikès’ testimony situates Michael’s floruit incontestably in the first half of
the twelfth century. Taken together with its sequel, it contains also a strong hint
that Michael, who, unlike Eustratius, certainly was no theologian, may have been
a physician. Praechter refers to the surmise, in a book of literary history, that
Michael followed the medical profession.74 Consistent with this hypothesis,
Praechter concedes, are Michael’s interest in Aristotle’s biological works,
exceptional not only for his time but in the Aristotelian tradition as a whole, his
personal observations in the areas of physiology and psychology, and especially
various remarks on the activity of sight, not derived from Aristotle; although
a passage in his commentary on GA implies that, at the time Michael wrote
this, he was not (yet) a medical man.75 It is of course not unthinkable that
considerations of this kind are at the origin of the hypothesis. In any case,
Michael’s aetiology of his own eye-disease has a professional ring. Tornikès’ next
sentence reads surprisingly:76

Moreover, she [= Anna Comnena] herself, although she did not yet make an
exact investigation [akribôsamenê] of medicine – for she was putting off its
detailed study [analusin] for the time being, and that because she was engaged in
philosophy and preferred philosophy, the medicine of the soul, before that of the
body – nevertheless organised [sundietitheto] the philosophical materials
according to the most scientific canons of the other discipline.

It is hard to understand this sudden jump in Tornikès’ train of thought – there

has been no mention of medicine before – unless the last of the words he
attributes to Michael suggested to him the notion of that profession; and that
again is most unlikely if Michael had no part in it. Michael’s remark, in one of his
well known references to his celebrated but unnamed master, that he died from
peripneumonia, also betrays a medical interest.77

74
Praechter, ‘Michaeli Ephesii in libros . . . ’, 864, n. 2: ‘In the literary history of Nicolas, III , p. 306, I find
the remark whose source I do not know, that Michael was supposedly a doctor by calling.’
75
ibid. 863–4. Cf. the remainder of n. 2, 864: ‘That would fit well with the above explanation. The
passage at p. 215, 30 gives, to be sure, an opposing indication, insofar as it shows that Michael was
not a doctor at a certain time, at least. Michael reports here, ‘one of my companions was a doctor’.
Someone who practised the same occupations would not express it like that. One would have to
expect, ‘. . . a companion was also a doctor by trade’.
76
Georges et Démétrios Tornikès: Lettres et discours, 283.
77
in Parva Nat., CAG 22.1, 142,5. See ‘Michaeli Ephesii in libros . . . ’, 902.
468 Aristotle Transformed

Michael, we have seen, was remarkable among Byzantine scholars for the
scope of his interests. He commented on Aristotelian works which were all but
ignored by other commentators as well as on those which were studied
traditionally. To the former class belong the commentaries on de Partibus
Animalium, de Motu Animalium, de Incessu Animalium,78 the Parva Naturalia79
and de Generatione Animalium80 – all of which are physical treatises – as well as
the only known Byzantine commentary on the Politics, of which but a few pages
survive.81 To the latter class belong the commentaries on the Organon, of which
none but that on the Sophistici Elenchi82 is preserved; on Books 6–14 of the
Metaphysics,83 if it is by Michael; and on Books 5, 9 and 10 of the Nicomachean
Ethics. The commentaries on the Physics and the Topics are lost.
L. Tarán has mounted a powerful challenge to the widely accepted view that
the commentary on Metaphysics 6–14 is by Michael. He argues that it was already
used seven centuries earlier by Syrianus, and that it was written by a forger
(pseudo-Alexander) who wanted it accepted as being the lost second part of
Alexander’s commentary on the Metaphysics.84 On the other side there are three
points to be made. Given Michael’s work habits, it is not unlikely that the
commentary we have is Michael’s reworking of the forger’s commentary
incorporating independent material, for example from Alexander’s Problemata.
Secondly, it is not clear that we should dismiss as spurious the remark at the end
of Michael’s commentary on the Parva Naturalia: ‘I also wrote on the Metaphysics
from precisely (autou) 7 to 14 (see Michael, in Parv. Nat. 149,8–16). It is
admittedly surprising that no commentary on Book 6 is mentioned, and that the
commentary on 7–14 is mentioned as an afterthought, when it would have been
in fact his most important work. But it is possible that Michael had not at the
time finished the Metaphysics commentary. Hence the unexpected ‘precisely’

78
M. Hayduck (ed.), CAG 12.2, Berlin, 1904. The last two works are translated into English in Anthony
Preus, Aristotle and Michael of Ephesus. On the Movement and Progression of Animals, Hildesheim
and New York, 1981.
79
P. Wendland (ed.), CAG 22.1, Berlin, 1891.
80
M. Hayduck (ed.), CAG 14.3 (wrongly attributed to Philoponus).
81
In the form of scholia in a Berlin MS of the Politics, O. Immisch (ed.), Aristoteles, Politica, Teubner,
2nd edition 1929, xvii–xxi and 293–327, translated in Ernest Barker, Social and Political Thought in
Byzantium, Oxford 1957. The scholia reflect, according to Immisch, not a historical or a political,
but a philosophic commentary, with little originality, in the Neoplatonic style. ‘But occasionally
Michael addresses himself . . . to questions of literary criticism; and occasionally, too, he refers . . . to
the conditions of his own time’ (E. Barker, op. cit., 137).
82
Max Wallies (ed.), CAG 2.3 (wrongly attributed to Alexander of Aphrodisias). Perhaps also
fragments of a commentary on the de Interpretatione: see CAG 4.5, suppl., xlvff.
83
M. Hayduck (ed.), CAG 1, 440ff (wrongly attributed to Alexander of Aphrodisias).
84
Leonardo Tarán, ‘Syrianus and Pseudo-Alexander’s Commentary on Metaph. E-N’, in Jürgen
Wiesner (ed.), Aristoteles: Werk und Wirkung, vol. 2: Kommentierung, Ueberlieferung, Nachleben,
Berlin and New York 1987, 215–32.
 The Greek commentators on Aristotle’s Ethics 469

(autou), the silence on Book 6 and the mention of this commentary as an

afterthought. Finally, the ascription of the extant commentary to Alexander is
easy to understand, since that reflects the forger’s intentions, but the ascription
by other sources to Michael is hard to explain, unless there were good grounds
for it.
Michael’s commentaries on the Ethics are according to his own word skholai,
lectures or commentaries, written either with a view to or as a product of his
teaching. In them he explains both the letter and the spirit of Aristotle’s text,
concerns himself with textual criticism, points out the connection between
various passages, spells out arguments that are too succinct for direct
comprehension, advances parallel texts, makes cross-references throughout the
Aristotelian corpus, and refers also to other philosophers, especially to Plato.
In his commentary on Book 5 Michael follows almost step for step the
anonymous scholia, which he often summarises or repeats in a different form,
sometimes even borrowing literally. Michael’s commentary on Book 5, while not
adding anything essential to the anonymous scholia, fills in what they left without
comment, supplies a short introduction but no concluding paragraph, and bears
witness to a better text of the anonymous scholia than the one we actually
possess.
Michael of Ephesus is above all known as an Aristotelian. His Aristotelianism
manifests itself in many ways. He chooses the physical and other treatises of
Aristotle to comment upon; he stays close to the text of Aristotle, which he
explains with reference to its general import (dianoia) as well as to its detailed
form (lexis); he makes a wide use of other works of the Stagirite; and he does not
attempt to make Aristotle fit into a Neoplatonic strait jacket. Michael differs
greatly from Eustratius on this account. Praechter contrasts Michael, the
Aristotelian, with Psellus, the Platonist.
However, Michael’s Aristotelianism is never a militant one. This point has
never been made, so far as I know, but it deserves to be emphasised because the
existing accounts tend to give the wrong impression. That his philosophical
language is coloured by Neoplatonism, although much less so than that of his
contemporaries, is a feature he shares with his age, but nowhere in his commentary
on the Ethics does he show an interest in attacking Plato and the Platonists in the
name of Aristotle and the Peripatetic school.
This is brought out by an analysis of the passages which explicitly mention
Platonic doctrines and expressions. They fall roughly into three classes: (a) where
Plato or Platonists are advanced in support of Aristotle; (b) where a specific
Platonic usage is simply mentioned or an opposition between Platonism and
470 Aristotle Transformed

Aristotelianism simply explained; and finally (c) where actual arguments against
Platonic positions are given. Let us consider each of them separately.
(a) There are two passages in the commentary on Book 985 and five in
Book 1086 where Plato is summoned in support of Aristotle. We may add to this
the single quotation from Plotinus.87 In none of these, however, does Michael
attempt to force Aristotle into a Platonic mould. The doctrines defended are
genuinely Aristotelian.
(b) In many places a Platonic usage or doctrine is simply stated or explained
without any discussion.88 Although in most of these passages Michael opposes
Aristotle to Plato or the Platonists, whenever he does so, he simply states the
opposition without supplying at that point an argument in favour of either of
the opponents.
(c) Only two passages in Michael’s commentary on the Ethics contain an
explicit argument against Plato. The first, found in the commentary on Book 5,89
is a reductio ad absurdum: if the gods possess virtue, they possess justice (in the
Aristotelian sense), which implies in turn that they possess exterior goods.
Michael refers to a similar argument ‘elsewhere’ which leads to the impossible
conclusion that the gods are subject to passions. These arguments are entirely in
line with Aristotle’s concepts of justice and virtue in the Ethica Nicomachea, and
are actually implied in EN 1178b11 and 1178b16, respectively. This is the only
place in Michael’s whole commentary on the Ethics that contains arguments
against Plato which go beyond what is expressed in the text of Aristotle on which

85
in EN 483,3 and 515,30.
86
ibid. 537,13; 579,5 and 30; 598,23; and 616,35. On 537,13, see below under (c).
87
in EN 529,21.
88
They are three in relation to Book 5: (1) in EN 25,28–9, which notes the difference between
Aristotle’s and the Platonists’ use of the word adikein (to treat unjustly): it is a synonym of blaptein
(to harm) for Aristotle, but not for the Platonists; (2) in EN 47,17–18, where the belief that justice
belongs to the gods, here implied by Aristotle’s argument, is ascribed to Plato; (3) in EN 72,5–6,
where Michael, explaining how Aristotle accounts in his own terms for Plato’s doctrine of justice,
explicitly quotes Plato’s definition of justice as the parts of the soul doing their own job.
One in relation to Book 9: in EN 524,17, which simply states the opposition ‘since . . . for Plato
pleasure and pain are not contraries’.
And ten in relation to Book 10: in EN 533,22 and 35, where the Platonists’ use of compound
nouns, formed by prefixing t- or auto- to an adjective, is explained and their identification of the
One with the Good is stated; in EN 578,16, where it is alleged that the Peripatetics, unlike the
Platonists, distinguish between moral and civil virtue; in EN 579,23, where the term nous (intellect),
here used by Aristotle in the Platonic sense, is contrasted with the Platonic expression zôê logikê
(rational life); in EN 585,10 and 13, as well as 601,4, which assert that the really happy man,
according to Plato, does not need anybody nor anything else, since he possesses the source of
happiness within his mind; in EN 619,15 and 19, as well as 620,4, where Plato is mentioned in
connection with Aristotle’s contention that nobody wrote an (adequate) work on politics before he
started on it, and where his judgment, that some of what according to Plato would preserve the state
would in fact be detrimental to it, is proffered as a matter of fact.
89
in EN 66,5–9.
 The Greek commentators on Aristotle’s Ethics 471

he is commenting. Even here, Michael does not proffer an independent

argumentation of his own.
The second place containing an argument against Plato is in EN 542,27ff in
the commentary on Book 10. Michael’s attitude, however, becomes very clear
when we consider the wider context, which includes also in EN 537,13.90 Michael
had here an excellent occasion to use Plato as a target, since both Aristotelian
passages involved (EN 1172b32–5 and 1173a15ff ) are direct attacks on Plato, but
Michael does not take this opportunity before in EN 542,27 and even then does
not go beyond repeating the argument of Aristotle’s text. Let us take the two
passages in turn.
First Michael simply explains Plato’s reversal of Eudoxus’ argument (EN
1172b28), still keeping within the scope of the text commented upon. Then,
however, he treats 1172b32ff as if it were actually an argument of Plato, although
it is directed against the very notion of an absolute good. Nor does he understand
1173a15ff as the criticism of the Philebus which it is. He even misunderstands
Aristotle’s appeal to Plato’s distinction between ‘mixed’ and ‘unmixed’ pleasure.
This shows how little Michael knew Plato. Next, Michael explains at length
Aristotle’s argument that pleasure is not a movement, without so much as
mentioning Plato. Then comes a surprise: ‘Having said what moved Plato to call
pleasure a movement, he (Aristotle) now argues against him and says what else
convinced him (Plato) to call it a movement.’ In the following lines Michael adds
nothing to Aristotle’s argumentation. He explains Plato’s arguments with the
same care as Aristotle’s.
In conclusion, with the qualified exception of 5, 66,5–9,91 Michael’s attitude is
clearly that of a commentator who attempts to explain what Aristotle says.
Platonic doctrines, explicitly mentioned by Aristotle or understood by Michael
as being alluded to by Aristotle, fall within the scope of such an explanation.
Platonic doctrines or expressions, whenever Michael introduces them, serve to
clarify the text. Nowhere does he venture an independent argument either for or
against Plato. He even fails to discern some obvious attacks on Plato in the text
of the Ethics.
The spirit in which Michael comments on Aristotle’s Ethics is on the whole
not unlike that exhibited in Albertus Magnus’ commentaries on Aristotle, except
that Michael does not sharply distinguish the scope of philosophy from that of
theology: he does not even mention theology. Michael is obviously not interested

90
See above.
91
See above.
472 Aristotle Transformed

in relating Aristotle to the dogma of the Church or to the Christian way of life,
despite the impression left by the repeated reference to the friendship between St
Gregory of Nazianzus and St Basil, by the equally repeated example of apostasy
as an instance of what one should not do under any circumstances, and by the
concluding apostrophe to the ‘Lord and Creator, crucified out of an immense
love for men’.

Anna Comnena’s ‘Aristotelians’

Robert Browning’s findings (Chapter 17 above) on Anna Comnena’s influence

have been widely acclaimed by scholars. The one exception is Eva de Vries-van
der Velden.92 She finds the very idea unacceptable that Anna could have inspired
an attempt at constructing a new philosophical system based on Aristotle rather
than on Plato in order to obtain a synthesis of ancient philosophy and Christian
dogma. This would have gone against the whole Byzantine tradition as well as
against Anna’s own attitude regarding Aristotle, expressed in her History and
echoed in the funeral oration. Dr de Vries points out the absence of any
theological concern in Michael’s commentaries and underlines that the rhetorical
style of a funeral oration does not promote exactitude. She believes that Tornikès,
himself ignorant in philosophical matters, only wanted to praise Anna’s general
culture, which included an interest in philosophy. She thinks it possible that
Anna requested some philosophers to procure her commentaries on Aristotle
for her private information only.
We are indebted to Browning for his re-dating of Michael and his hypothesis
on Eustratius, both of which give a role to Anna. The only question is how much
we can infer from the evidence to which he has drawn attention. My own view is
that the revival of Aristotelian scholarship did not go much beyond the activities
of Michael and Eustratius, and that it was confined to exegesis, not extending to
a revival of Aristotelian philosophy. Tornikès does not speak of a philosophical
circle around Anna, and though the study of Aristotle is later emphasised around
1265 in the inaugural lecture of Michael Anchialis as head of the philosophy
school (hupatos tôn philosophôn), the chair had been vacant for some years
before that. It is not possible either to be sure about the extent of Michael’s
dependence on Anna, or to exclude the view of Praechter that Michael was a

92
Eva de Vries-van der Velden, Théodore Métochite, une réévaluation, Amsterdam 1987, 111, n. 10.
 The Greek commentators on Aristotle’s Ethics 473

philosophy professor at the Academy founded by Constantine IX Monomachus,

and of Barker that he wrote most of his commentaries there.93 He could have
started his work as a commentator before Anna contacted him, and, as Preus has
argued,94 he could have continued it after 1138, when she redirected her attention
to work on her Alexiad.
The extent of Michael’s work on the composite Ethics commentary is also
hard to gauge. He certainly wrote the portions on Books 5, 9 and 10 himself. But
he could also have been the person who planned the entire commentary and put
it all together, including the makeshift portion on Book 7.95

The anonymous commentator on Book 7

The anonymous commentary on Book 7 has again the character of a formal

commentary by one man, although the title refers to it as skholia.
One can but subscribe to Schleiermacher’s verdict that it exceeds everything
in poverty. It misinterprets passages that are quite clear in themselves. It lacks
good taste as well as an elementary feeling for (classical) Greek, as is illustrated
by the Latin word karbônes (for anthrakes – coals), and by the present tense of
phagein (to eat).96 The style is clumsy and marred by the endless repetition of êtoi
(or) by means of which explanation is heaped upon explanation. Nor is the
commentator afraid of repeating himself.
The language marks the composition as very late. Rose could only give
Grosseteste’s translation as a terminus ante quem. The commentary was evidently
written with the sole purpose of filling the gap left by the commentaries I have
already discussed. The subject matter of Book 7 (incontinence and pleasure)
could appeal to a man Schleiermacher suspected might be a physician because
of his manifest liking for examples from the medical field.

93
Karl Praechter, review of CAG 22.2 (Michael in PA ), Göttingische gelehrte Anzeigen 22, 1906, 903ff
(reprinted in his Kleine Schriften, H. Dörrie (ed.), 1973); Ernest Barker, Social and Political Thought
in Byzantium, Oxford 1957, 136.
94
Anthony Preus, Aristotle and Michael of Ephesus, On the Movement and Progression of Animals,
Hildesheim and New York 1981, 10, n. 22. Preus, however, envisages a more extensive and vigorous
circle around Anna Comnena.
95
Sten Ebbesen in Chapter 19 below also considers it likely that he was the editor.
96
It also makes the verb agree in number with a neuter plural. Although this is common in sixth-
century authors, it is not found in Michael or Eustratius, any more than in classical Greek.
474 Aristotle Transformed

Aspasius on Book 8

Aspasius was a renowned Peripatetic teacher at the beginning of the second

century.97 His commentaries were still used by Plotinus in his school, as were
those of two other Peripatetics, Adrastus and Alexander, both of Aphrodisias.98
Galen advised a friend who requested a commentary on Aristotle’s Categories
from him to begin with reading those of Adrastus and Aspasius.99 He himself in
his youth had heard a pupil of Aspasius, the Peripatetic, lecture. Galen does not
name this pupil but identifies him as a fellow-citizen, i.e. as originating from
Pergamon.100 Aspasius is not to be identified with his contemporary, the sophist
Aspasius of Byblus, an Atticist rhetorician, nor with his namesake of Tyre, also a
rhetorician, both mentioned in the Suda.101
Aspasius wrote formal commentaries on Aristotle: on the Categories, de
Interpretatione, Physics, de Caelo, de Sensu and Metaphysics, none of which have
survived, and on the Nicomachean Ethics. Of the latter we possess the elucidations
of Books 1 to 4, part of Book 7, and most of Book 8.
Many gaps occur in the extant parts. Moraux attempts to show that the whole
of our present-day text probably goes back to one incomplete and quite damaged
exemplar that must have been discovered and copied in the late thirteenth or
early fourteenth century (pp. 252–3). However, Aspasius’ commentary on Book 8
in the Byzantine collection ends in the same tattered way as it does elsewhere.
Hence, at least the tail-end mutilation dates from the twelfth century or before.
That the Byzantine compiler obviously had no access to the remainder of
Aspasius’ commentary is another indication that the work had already fallen to
pieces by that time. It is true that all the MSS that contain Aspasius on other
Books beside 8 belong to the same tradition and that his commentary on Book
1, where it appears after Eustratius’, is acephalous, i.e. derives from a copy with a
defective beginning, but the latter does not have to be the archetype of that
tradition. Nor do we have to ascribe the existing lacunae to this archetype, for the
text of Book 8 that belongs to it, although differing from the one that occurs in
the Byzantine compilation, is no less complete.

97
On Aspasius, see Paul Moraux, Der Aristotelismus bei den Griechen, vol. 2, 226–93, especially 249–
93: ‘G. Zur Nikomachischen Ethik: a) Ueberlieferung; b) Tendenz und Einordenung; c) Die
kontroversen Bücher; d) Aspasios und sein Vorgänger; e) Philosophischer Standort.’
98
Porph. Vit. Plot. 14,10–14.
99
Galen 19, de Libr. Propr. 42,10–43,1 = Scr. Min. 2.118,17–119,2.
100
Galen 5, de Cogn. An. Morb. 8, 41,17–42,2 = Scr. Min 1.32, 5–7.
101
Pace R. Hanquet, Aspasius, sa vie, son oeuvre, Diss. Louvain 1945, 12ff.
 The Greek commentators on Aristotle’s Ethics 475

There is no reason to doubt that Aspasius’ work covered the ten books of the
Nicomachean Ethics. What remains of it is, in Gauthier’s judgment, the most
valuable legacy left by antiquity for the understanding of Aristotle’s moral
philosophy. Aspasius bears witness to a much older state of the text of the Ethics
than that represented by our manuscripts, the oldest of which dates only from
the tenth century.102 His education and his ties with the Peripatetic tradition
often enabled him to comprehend Aristotle’s intentions accurately.103 Gauthier,
nevertheless, in his own commentary frequently points to traditional
misinterpretations that originated with Aspasius. He reminds us that Aspasius
was strongly influenced by the Stoics and that he was after all separated by five
centuries from his master. The repeated emphasis on being in a natural state
(ekhein kata tên phusin) as the standard for both friendship and virtue, which at
the same time shows them to be rational (kata ton logon), is Stoic. The appreciation
of virtue as the highest good, although tempered by the qualification ‘after
happiness’, is also Stoic.
That Aspasius’ language is coloured by Stoic expression has been noted by
R. Hanquet.104 It does not mean that he deliberately takes a Stoic position.
As Moraux observes (p. 269), orthodox Aristotelians such as Alexander of
Aphrodisias often use typical Stoic terms, without experiencing them as foreign
elements. Stoic terminology had invaded the philosophical language of the time
in such a way that its use is no evidence for a Stoic approach. In all great ethical
questions, Moraux continues, Aspasius takes a position against the Stoa: external
goods contribute to happiness; some passions are compatible with reason; the
main species of passion are only pleasure and pain, to which fear and desire must
be reduced; the virtuous man can suffer pain.
Aspasius’ method is that of paraphrasing Aristotle in a clear, succinct manner.
He keeps well to the point and abstains from introducing literary and historical
information, a practice that characterises the anonymous scholia on Books 2
to 5. But he concerns himself with problems of textual criticism and attempts to
elucidate the meaning of different variants.105
One interesting textual problem about which Aspasius has some information
to offer is that of the so-called common Books 4–6 of the Eudemian Ethics (EE )
which are the same as 5–7 of the Nicomachean (EN ). Hans Gottschalk has

102
Moraux notices, however, that for his commentaries on the Physics and de Sensu Aspasius possessed
manuscripts with corrupt passages, some of which were the result of unscrupulous contemporary
emendations (238–9; 244–5). His readings of the Ethics diverge less from our vulgate (257).
103
R.A. Gauthier, L’Éthique à Nicomaque2, vol. 1, part 1, 100.
104
R. Hanquet, op. cit. in n. 101, 78–81.
105
Moraux gives a list of the passages discussing variants, 257, n. 101.
476 Aristotle Transformed

shown106 that Aspasius found the common books in his text of EN, but thought
they had been transferred from EE, which he imagined to be by Aristotle’s pupil
Eudemus of Rhodes instead of by Aristotle himself, the transfer being designed
to replace some lost books of EN . Kenny has further shown that at first it was EE
rather than EN that was most often quoted and enjoyed the higher esteem.107 But
in Chapter 3 above Gottschalk challenges Kenny’s view that it was Aspasius who
initiated the change and transferred the common books.
Aspasius critically examines the definitions of passion given by Andronicus
of Rhodes and by his pupil Boethus,108 whom he contrasts with the ancient
Peripatetics, none of whom give such a definition.109 Occasional objections to
Aristotle are always resolved. Some of these objections appear to come from an
anti-Aristotelian source: those against the thesis that the absence of good birth,
of good children and of beauty affect happiness; against the interpretation of
irony as a vice, since Socrates possessed this quality; against the opinion that the
high-minded person (megalopsukhos) would be ashamed of being the recipient
of benefaction; and against the characteristic tendency of the same to remember
benefits bestowed but to forget benefits received.110 In his defence of metriopatheia,
Aspasius rejoins the Aristotelian position of his time.
Moraux (pp. 255ff ) wonders whether the relatively pure Aristotelianism of
Aspasius is the result of his deliberately adhering to the orthodoxy of the school
or of his understanding his task as an interpreter to be that of explaining the text
of Aristotle as carefully as possible from the author’s point of view. In several
places in his commentary, Aspasius refers in the third person plural or by the
pronoun houtoi (they) to unnamed individuals who must have been
Aristotelians.111 Nowhere does he expressly present himself as belonging to the
school, not even when he distinguishes the older from the younger Peripatetics.
Moraux is inclined to consider Aspasius as formally an outsider, but with strong
Aristotelian sympathies.

106
H.B. Gottschalk, ‘Aristotelian philosophy in the Roman world’, in Wolfgang Haase (ed.), Aufstieg
und Niedergang der römischen Welt II , 36.2, 1158.
107
Anthony Kenny, The Aristotelian Ethics, Oxford 1978. Cf. also Moraux, op. cit. 257–61.
108
Boethus the Athenian Peripatetic of the first century BC is not to be confused with Boethus the
Stoic in the second century BC .
109
in EN 44,20–4. For Aspasius’ discussion of the definition of passions see also Moraux, 279–88;
H.B. Gottschalk, op. cit. in n. 106, 1156–7; F. Becchi, ‘Aspasio e i peripatetici posteriori. La formula
definitoria della passione’, Prometheus 9, 1983, 83–104; and Pier Luigi Donini, Tre studi
sull’aristotelismo nel II secolo D.C., Torino etc. 1974, 63–125: ‘Capitulo II . Il platonismo medio e
l’interpretazione dell’etica aristotelica’, especially 100–104.
110
Moraux, 267–8.
111
in EN 2,16–3,2; 5,23–30; 10,29–32; 156,14–16.
 The Greek commentators on Aristotle’s Ethics 477

The argumentum a silentio is always tricky. Given that Aspasius was known
in antiquity as a Peripatetic and that his doctrinal leanings are Aristotelian,
the relative distance he appears to take from other Aristotelians may well be
that of a contemporary interpreter against his predecessors in the same
school. Moreover, he may find it a good pedagogical device not to proclaim
himself too openly as a follower of the master whose texts he wants to clarify.
Gottschalk still considers Aspasius as one of the few ‘professed Aristotelians’ of
the first two centuries of the Roman Empire (p. 1080). He points out that
Aspasius’ criticism of Andronicus’ and Boethus’ definition of passion was
inspired by a desire to purge the school teaching of accretions that resulted
from Stoic influence, thus placing Aspasius within ‘the trend towards a purer
form of Aristotelianism’ (p. 1158). What is important, in any case, is that Aspasius
clearly saw it as his task to develop the content and meaning of Aristotle’s text
as carefully as possible from the author’s point of view (p. 269) and, one should
add, that he possessed the necessary professional equipment to accomplish that
task well.
Aspasius’ commentary on Book 8, the one incorporated in the composite
commentary, survived in a mutilated form. It appears to be complete up to 180,8
hekastôi gar, where it breaks off abruptly. The passages covering EN 1159a35–
b19, 1160a9–33 and 1160b21–1161a9 are missing. Moreover, from 180,10
onward, the commentary gradually acquires the character of disconnected
skholia, becoming sparser and sparser toward the end, with the order of the text
disturbed. The last five pages, approximately one-sixth of the commentary, cover
one-third of Book 8.

Eustratius in Albertus Magnus and Bonaventure

I shall finish by returning to Eustratius, to see how his commentary, once

translated into Latin by Grosseteste, was received by Albertus Magnus and
Bonaventure.
Albertus Magnus commented on the Nicomachean Ethics twice; the first time
about 1250–1252 in Cologne, the second time about 1267–1270, possibly in
Strasbourg. There is a great difference between the two commentaries. The first
time, Albert lectured on the newly translated complete text of Aristotle’s Ethics at
the freshly founded studium generale of his order. He chose the form used by the
masters of the arts in Paris: an explanation of the littera, acompanied by
quaestiones. We possess a reportatio (notes taken down by one or more students)
478 Aristotle Transformed

of these lectures. It is likely that it was made by Thomas Aquinas.112 Albert tries
in these lectures to come to grips with a newly discovered text, aiming at exegesis
and explanation, but not at the sort of interpretation which would fit Aristotle’s
theory into a thoroughly Christian ethics, as Eustratius had attempted to do.
Here, for the first time, a sharp distinction between philosophy and theology is
formulated, which was soon to be taken over by Thomas Aquinas. Albert also
uses all the sources he can: among others, the recently translated Greek
commentaries including Eustratius, Michael of Ephesus and the notule of
Grosseteste; Averroes113 and other materials from the Arabic tradition, especially
Avicenna; and Nemesius, de Natura Hominis and John of Damascus, de Fide
Orthodoxa.
The second time, almost twenty years later, when Albert ‘felt himself to
be an old man’, he did not lecture on the Ethics but wrote a paraphrase of it,
‘based on the Avicennan model, aimed at popularizing the texts of Aristotle,
which many found too difficult’.114 This was one of the last in a series of
such commentaries on the works of the Stagirite. Whereas Albert was at pains
to distinguish Aristotle’s philosophical treatment of ethics from a theological
one in his first commentary, here this distinction is taken for granted. He
gives ‘Aristotle’s teaching a purely philosophical context’, sees it ‘in the light
of other rationalist moral philosophy’, that of ‘the Stoics whose head was
Socrates’, in other words, the Greek Platonist school, which he combines with
the Roman Stoics, quoting Plato, Cicero and Seneca all under the same
category.115
Also in his second commentary, Albert takes over a great deal of the material
from Eustratius and the other commentators. For example, he uses Eustratius’
account of the theory of Ideas, stating the argument in the same order but more
condensed, often using the Eustratius/Grosseteste phrases, without attribution,
except occasionally to the ‘Commentator’, who here is Eustratius, not Averroes. If
we compare Albert lib. 1, tract. 5, cap. 12 with Eustratius in EN 40,4–14; 17–34,
we find that Albert has extracted six different arguments from one page of
Eustratius.

112
Albertus Magnus, Opera Omnia, 14, Super Ethica Commentum et Quaestiones, W. Kübel (ed.),
Münster, 2 vols, 1968–1970 and 1987.
113
Averroes did not write a long commentary on the Ethics, only a middle commentary, completed in
1147 and translated by Herman the German, Toledo 1240.
114
Jean Dunbabin, ‘The two commentaries of Albertus Magnus on the Nicomachean Ethics’, Recherches
de Théologie ancienne et médiévale 30, 1963, 232–50.
115
Dunbabin, op. cit., 246.
 The Greek commentators on Aristotle’s Ethics 479

Thomas Aquinas learned about the Grosseteste translation of Eustratius ‘and

others’ from Albert’s Cologne lectures, but seems not to have had a copy of his
own. Of course his interpretation owes a good deal to Albert, who developed
those lectures relying on the Greek commentaries.116
We may possibly also trace the influence of Eustratius ‘and others’ through
Albert to Bonaventure. In the middle of Lectio 6 of Book 1, in the first
commentary, Albert says:

These arguments of Aristotle necessarily work against Plato, if Plato posited one
Idea of all good things as their generic or specific form, as we explained above;
this is how Aristotle understands him. If, on the other hand, Plato meant one
Idea, by virtue of which all things descend as likenesses from one first formal
cause of all good things, as the Commentator says, then it is clear that Aristotle’s
arguments have no force (quod rationes nihil valent).117

The last words are also found in Bonaventure’s famous anti-Aristotelian

Collationes in Hexaemeron 6,2,118 but the Seraphic Doctor does not make the
restrictions which Albert makes. He completely agrees with Eustratius’ criticism
of Aristotle’s attack on the Supreme Good as one Idea:

Whence Aristotle is also the first man to attack these Ideas in the Ethics,
where he says that the supreme Good cannot be an Idea. And his arguments
have no force (quod rationes nihil valent); they have been solved by the
Commentator.119

This is clearly a reference to Eustratius, not Averroes, and indeed to the argument
which we have already discussed. In the thirteenth to fifteenth centuries, the
designation ‘the commentator’, which normally applies to Averroes, is used in
relation to the Ethics for Eustratius. And no doubt it was found reassuring that
the commentator on that work was a Christian.
In his second commentary, Albert follows Eustratius more closely. He gives an
affirmative answer to the question120 whether this is the first good before which
there is nothing, and which is the cause of good in all other things, and adds: ‘but
some Peripatetics have objected with sophistical objections (sophistice) to this’.

116
See R.A. Gauthier, ‘Appendix: St. Thomas et l’Éthique à Nicomaque’, in Thomas Aquinas, Opera
Omnia, vol. 48, Rome 1971, xviii.
117
Lib. 1, lectio 6, 27,40–5.
118
Quaracchi edition, Opera Omnia, Florence, 5,361.
119
Collatio 6.2, 361.
120
Lib. 1, tract. 2, cap. 3.
480 Aristotle Transformed

Eustratius uses the word sophistice at least three times to describe Aristotle’s
criticism of Plato’s doctrine of the Good. Albert then adduces the Stoics
(= Platonists) to argue in favour of a ratio boni which is not a ratio generis, which
would belong to all its subjects, but a ratio principii, under which all good things
fall, but which transcends them.121

121
I am grateful to Professor G. Locke of the University of Leiden and the Catholic University of
Nijmegen, to Dr S. Scott-Fleming of Linacre College, Oxford, and to Mr Robin Rollinger of Utrecht,
who all read considerable portions of this chapter and helped to improve my English. I am
particularly indebted to Professor Sorabji for his many helpful suggestions. The new material in this
chapter will be included in my introduction to CLCAG VII 3 (forthcoming).
 19

Philoponus, ‘Alexander’ and the origins

of medieval logic1
Sten Ebbesen

Some modern scholars, Richard Sorabji prominent among them, are building
a heroon for John Philoponus, the man who courageously rebelled against
Aristotelian physics and pointed out weaknesses in traditional arguments for
the eternity of the world.2 In my mythology he used to be of less than heroic
stature; I used to think of him as a chimera, whose head would happily
chew away at its own tail, for I knew him only as the author of some generally
flaccid commentaries on the Organon3 whose few interesting statements tend
to be mutually inconsistent. But perhaps both views of Philoponus are right.
Professor Sorabji’s heroic inventor of a new physics is the mature scholar
who furthermore is fired by a crusader’s conviction that he is fighting for truth.
My John Chimera is the diffident young assistant professor of logic charged
with teaching a subject he has not quite mastered; he relies on notes taken
during professor Ammonius’ lectures, but they are none too good, for though
Ammonius taught lots of logic classes in his time, he never managed to become
really interested in the discipline. Like most of his contemporaries, he considered
acquaintance with the Organon as only a stepping-stone on the way to real
philosophy.
In antiquity the great era of logic was over by about 200 bc . Its next blossoming
occurred in Western Europe between ad 1100 and 1400. In this chapter I shall

1
This is a slightly revised version of a lecture given in Professor Sorabji’s seminar at the London
Institute of Classical Studies on 22 February 1988. It was not meant for publication and time did not
permit a thorough revision to supply documentation for all claims. See the Bibliography on
pp. 460–1 for further details of the authors cited in the footnotes.
2
See in particular Sorabji (ed.) 1987.
3
On the Categories, Prior Analytics and Posterior Analytics; published in CAG 13.1–3.

481
482 Aristotle Transformed

contend that late ancient Greek scholastic writers on logic, Philoponus among
them, in spite of their own minute status as logicians, not only gave important
impulses to medieval scholastic logic in general, but more specifically contributed
both to the rise of a nominalistic current which was strong in the twelfth and
fourteenth centuries, and to the rise of the realistic current which dominated the
thirteenth century.

When I was as young as Philoponus perhaps was when he wrote his commentaries
on Aristotle, that is in my early twenties, I embarked on a one-man crusade to
rescue the glorious achievements of Byzantine logicians from the oblivion into
which they had fallen. I was then working on an edition of a Latin logic book
from about 1270; some quotations in it suggested use of a rather sophisticated
Greek commentary on Aristotle’s Sophistici Elenchi, probably by the old professor
of Aristotelian philosophy, Alexander of Aphrodisias, who lived round ad 200.
However, no manuscript could be found, whether of the Latin translation or of
the Greek original. So to find out what was going on, I had to gather Latin
fragments and to look for Greek counterparts by searching Byzantine sources
for traces left by their use of Alexander. Once I had got that far in my
considerations, the idea of an expedition to rescue the beautiful virgin, Logica
Byzantina, was born. My reasoning was simple: Everybody who ever said
anything about Byzantine philosophy describes it as essentially theology and if
they mention logic at all, they agree that it was of no value, consisting of bad
rehashes of the ancients. This is exactly the description that used to be given of
Western European logic from the Middle Ages, before the research of this
century showed it to be of a standard not reached again till very recently. So
probably, if someone who does not wear the distorting spectacles of a late
nineteenth-century scholar starts to look at the texts, he will find gold where the
old philologists found dirt.
Simple and persuasive argumentation – but wrong. Old Busse and his
companions in the late nineteenth century had been quite right in their
evaluation of Byzantine logicians.
A fiasco, then, which became only more tragic when my research did not
reveal traces of a lost work by Alexander. Instead, many footprints stopped at
Philoponus’ door. Still, I continued the study of those miserable Greek texts for
more than a decade because it became clear that for all their lack of logical
insight they had proved historically important, helping to start that most glorious
era of logic research that lasted from about 1100 till about 1400.
 Philoponus, ‘Alexander’ and the origins of medieval logic 483

The text I was editing twenty years ago consisted of ‘quaestiones’ on Aristotles’
Sophistici Elenchi,4 composed by an unknown and mediocre master at Paris
about 1270, i.e. about the end of Thomas Aquinas’ life. By then, even mediocre
Parisian masters were much better logicians than Philoponus or the run of
Greek commentators. Nowhere in the ancient world was there anything like the
university of Paris. Yet my Parisian master refers with deference to a Greek
commentator, ‘Alexander’.
In a discussion of the six fallacies that Aristotle says depend on features of
language,5 it carries the weight of an argument that according to Alexander some
of them depend on an actual, some on a potential and some on a purely imaginary
multiplicity of meaning.6 A little research will establish that this division is first
met with in a small essay7 on language-dependent fallacies by the famous
physician Galen in the second century and reappears in a commentary on the
Sophistici Elenchi by Michael of Ephesus from about 1130.8 This Greek item not
only was standard doctrine in Western Europe between 1150 and 1500, it also
created the framework for new theories.
In what sense exactly is ‘flying planes can be dangerous’ an ambiguous
expression? When correctly pronounced it is not ambiguous at all – it is not even
it. Either you have a sentence that points out the dangers threatening mankind
from objects that do not stay on earth; or one that warns you about exercising a
certain action which affects aeroplanes, whether paper-aeroplanes or real ones is
not clear. Using the Galenic classification you say we have a case of potential
duplicity of meaning in the case of ‘flying planes’. Now, this catches an intuition
that we are dealing with something that is close to being real ambiguity – so
close that it cannot be written off as pure imagination, as something that only an
ignorant person could think of as a case of ambiguity. How is this intuition to be
spelled out in scientific language? Well, the name ‘potential ambiguity’ provides
a clue: a potency is actualised when matter is equipped with form. Why not think

4
Incerti Auctores (1977).
5
Arist. SE 4, 165b24–27.
6
For Western sources that attribute this division to Alexander, see Ebbesen 1981a, vol. 2, 395ff.
7
Galenus, de Captionibus in Dictione. Edition in Ebbesen 1981a, vol. 2. Discussion of the work in
Ebbesen 1981a, vol. 1, 79–87 and 236–9.
8
The final edition of Michael’s commentary was edited by Wallies in CAG 2.3, as ‘Alexandri
quod fertur commentarium . . .’; in Ebbesen 1981a and some other publications I refer to it as
‘Ps.-Alexander-1’. An earlier version, which I have baptised ‘Ps.-Alexander-2’, remains unedited,
but there are some excerpts in the preface to CAG 2.3 and in Ebbesen 1981a, vol. 2. A table
showing Michael’s loans from Galen is found in Ebbesen 1981a, vol. 2, 3. Michael’s text about the
three sorts of multiplicity and a couple of Latin echos are printed in parallel columns in Ebbesen
1979, xxxii–xxxiii.
484 Aristotle Transformed

of the single words of the two sentences as the common, undifferentiated matter
of both, which becomes actualised when certain suprasegmental features,
intonation, pausation etc., are added? Why not, indeed? There you have got a
theory of the relation between string of words and sentence, and one that may
also be transferred to the relationship between phonemes and word. The theory
I have just sketched developed in the thirteenth century from a germ that was
just a few lines of a Greek scholium translated into Latin.9
I shall not bother you with the details here, but my research uncovered many
references to and loans from ‘Alexander’ in Latin texts from about 1150 or later.10
It proved impossible to accept the ascription of the Greek commentary to
Alexander of Aphrodisias, whereas it proved to have had an uncanny resemblance
to the commentary that Michael of Ephesus composed about 1130 and that
Maximilian Wallies edited in Commentaria in Aristotelem Graeca 2.3 a century
ago. I was able to chart the genesis of Michael’s commentary,11 and a remarkable
feature of the Latin translation appeared: it seemed to contain some material
from an early, some from a late, edition of Michael’s work, and some material
known to have been in his possession though left out in the final edition.12
Do not imagine, however, that Michael was fastidious. He could not afford to
be, since none of the ancient Elenchi-commentaries had survived for him to
pillage. He therefore vacuumed old books for useful passages that might serve as
or in scholia on the Elenchi, then emptied the bag of the vacuum cleaner and
called the rubbish-heap a commentary. The heap emits a characteristic smell. It
smells of Philoponus and his age, for that was where Michael gathered many a
crumb and formed his whole conception of his subject and his task.13
Let me now turn from the Elenchi to the Posterior Analytics, which also makes
its appearance in the West around 1150. It took some time for it to become a
regular item in the curriculum, and the earliest known Latin commentary on it
is from about 1225, composed by the idiosyncratic bishop of whom Richard
Southern recently painted an unforgettable portrait in his Robert Grosseteste: the
growth of an English mind in medieval Europe – the man who used the incomes

9
For further discussion of Western theories about potential ambiguity, see Ebbesen 1981b.
10
The fragments of ‘Alexander’s’ commentary on Arist. SE were published in Ebbesen 1981a, vol. 2,
331–535. I have found a few more since then, but none that gives important new information.
11
Ebbesen 1981a, vol. 1, 268–85. There is also a summary of my findings in Ebbesen 1979, viii–xii.
12
The evidence for this is scarcely conclusive, but while most Latin fragments match Michael’s first
edition best, one discussed in Ebbesen 1981a, vol. 3, 221 seems to reflect the final edition. A Latin
text discussed in op. cit. 206 appears to echo a scholium from a collection (Comm. II ) that Michael
certainly knew though he did not include this particular item from it in his commentary.
13
Thus Michael’s preface to the Elenchi commentary is closely related to, and possibly derived from
Philoponus’ in his commentary on Analytica Posteriora, CAG 13.3.
 Philoponus, ‘Alexander’ and the origins of medieval logic 485

of England’s richest diocese to keep up his own research collective; the man who
at the end of his life went to Lyons to tell the pope that there were strong
indications that the pope was Antichrist. Southern contrasts Grosseteste’s grasp
of what is going on in the Posterior Analytics with the superficial acquaintance
displayed by the slightly older Alexander Neckham, who ‘looked on it partly
as a collection of logical puzzles disturbing the traditional study of logic, partly
as a haphazard collection of scientific curiosities ripe for moralisation’, and thus
in a heavily moralising chapter about the stars tells his reader that

Stars twinkle, as Aristotle says in his Posterior Analytics, because they are very
distant from the earth. Other stars do not twinkle, because they are near. The
same effect can be seen in a candle which does not twinkle when it is near, but
only when it is far away. Similarly, the minds of the faithful who are far removed
from earthly desires are radiant with the light of grace.14

Southern correctly notes that Neckham does not pay any attention to the form
of the argument, which was Aristotle’s only reason for mentioning the matter,
and shows no interest in the scientific explanation of the phenomenon either.
But he adds,

And yet, in parenthesis, there is one genuine observation – whether Nequam’s or

borrowed we cannot tell, but it is not Aristotle’s – he notices that a candle in the
distance twinkles, but nearby it does not. This small observation brings us a little
nearer to scientific ‘experience’, and to Grosseteste.15

It probably is true that Grosseteste had a better grasp of the Posterior Analytics
than Neckham. And it is true that the candle-example brings us a little nearer to
Grosseteste – but in quite another way than Southern thought. For the candle-
example is borrowed from John Philoponus’ commentary on the Aristotelian
passage,16 and Grosseteste also contains material from Philoponus – in fact, a

14
Alexander Neckham de Rerum Naturis I.6; the translation is taken from Southern 1986, 154–5.
15
Southern 1986, 155.
16
Philoponus in An. Post. (CAG 13.3) 171,13ff. In a footnote on p. 155 Southern comments on similar
casual reference to An. Post. in Neckham de Nat. Rer. II .22: ‘pisces non respirant, ut ex doctrina
Posteriorum Analecticorum liquet.’ Southern says: ‘There is in fact no mention of the respiration of
fish in An. Post., and this is not Aristotle’s doctrine in those works (. . .) where he discusses the problem.
Nequam’s statement, therefore, is an illustration of the distorted way in which the ideas of works not
yet fully available in translation circulated in the early thirteenth century.’ Yes – but the distortion just
represents the common medieval practice of crediting an auctor with saying what his text means
according to the standard commentary. In An. Post. 1.13, 78b15ff Aristotle says that although the fact
that a wall is no living being may be used to show that it does not breathe, this fact does not explain
why it does not breathe, for while ‘everything that breathes is a living being’ is true, ‘every living being
breathes’ is not. Aristotle does not tell his reader what living organisms do not breathe, but Philoponus,
in An. Post. (CAG 13.3) 174,28–9 gives insects and fish as examples of such creatures.
486 Aristotle Transformed

whole page of his commentary is a literal translation of Philoponus.17 The

candle tells a story about a debt to the Greeks, not about English curiosity about
nature. Research by Minio-Paluello, Bernard Dod and myself has unearthed
a considerable number of citations and direct quotations of Philoponus’
commentary on the Posterior Analytics in Latin works from the twelfth to
thirteenth centuries. However, when the Westerners mention their source, they
call him Alexander.18
It is much the same story with the Prior Analytics. It began to be studied only
after the middle of the twelfth century. And again it turns out that the first men
to work on it had access to some Greek explanatory notes, possibly a whole
commentary. And again it turns out that there is a very good match between
Philoponus’ commentary and the echoes of Greek scholia that we find in Latin
works.19
You may feel relieved to learn that one of the books of the post-1150 New
Logic came without Philoponean notes. Apart from one or two scholia by
Alexander of Aphrodisias, Western exegetes depended for their understanding
of the Topics on Boethius, and hence indirectly on Themistius.20
For the Old Logic, i.e. Porphyry’s Isagoge, and Aristotle’s Categories and
de Interpretatione, early Western scholars neither had nor needed Greek
commentaries. Their standard companions to these works were by Boethius, a
contemporary of Philoponus’, who had exploited Greek predecessors – Porphyry
in particular – to the utmost.
Until c. 1150 only the three books of the Old Logic were read, for only they
had Boethian commentaries (we must remember that not one Western scholar
in a thousand knew Greek). The New Logic, i.e. the rest of the Organon, became
accessible through a combination of digging up forgotten Boethian translations

17
Rossi 1981, 142–3 = Philop. in An. Post. (CAG 13.3) 109–10.
18
The most complete collection of fragments of the Latin ‘Alexander’/Philoponus commentary on
An. Post. is the one in Ebbesen 1976, 89ff with corrigenda in Ebbesen 1979, xxxix–xl. A few more
are found in Dod 1970. There are others as well. For Grosseteste’s use of the Greek material, cf. Rossi
1978 and 1981.
19
The exact nature of the Greek material on the Prior Analytics that had been translated into Latin is
still a matter for research. James Shiel, following a suggestion by L. Minio-Paluello, in a number of
papers (1982, 1984a, 1984b) has defended the thesis that all the West had was the set of scholia
found in MS Firenze BNC Conv. Soppr. J.6.34 (published in Aristoteles Latinus III ) and that they
are due to Boethius. In Ebbesen 1981c reasons are given for thinking there may have been more
material than offered by the Florence MS . I also think that the identity of the translator cannot be
determined with certainty before a more thorough search for traces of the Greek material in the
Latin tradition has been carried out. As things are, the argument for Boethius is pretty strong. For
the whole question, cf. Ebbesen 1987a, 290–1 n. 15 (= p. 376 of this volume).
20
For the translated scholium or scholia on the Topics, see Ebbesen 1976, 118–20 and 1979, xli–xlii.
For the Western tradition of exegesis of the Topics, see Green-Pedersen 1984.
 Philoponus, ‘Alexander’ and the origins of medieval logic 487

of Aristotle, making new ones, and providing the necessary guidance for the
reader by translating Greek commentaries. Even though the masters of France
were often great logicians, they needed help to find their way in the maze of
Logica Nova. Thanks to the long Greek tradition of exegesis, a mediocre Greek
could help a brilliant Frenchman.
The man who did the translations was James of Venice – Jacobus Veneticus
Graecus.21 We do not know where he lived at the time he did the work, nor
whence he got the Greek manuscripts needed. But we do know that he was in
Constantinople at least once. And we know one environment there where he
could find all the material needed. It was a circle of scholars whom princess
Anna Comnena had gathered in the first decades of the twelfth century to study
Aristotle. Michael of Ephesus was one of them. I strongly suspect Michael let his
friend James use his manuscripts, including some that were the working copies
for his own commentaries.22
I mentioned earlier that Michael vacuumed old manuscripts to find notes for
his Elenchi commentary. Indeed his whole method of work consisted in gathering
whatever ancient materials he could lay hands on, putting them together,
mending them and supplementing them, so as to produce something that
could claim to be a companion to a whole work by Aristotle. He put together
commentaries on the Metaphysics and the Ethics in this way too.23 Why? The
reason is as simple as it is terrible: there was no living philosophical tradition in
Byzantium to draw on. No wonder the empire of Constantine lasted for only one
millennium!
Ammonius and his pupils from the sixth century were the men a Byzantine
scholar would primarily consult if he wanted to do some work on the Organon.
Those men who made it possible for a small part of the ancient philosophical
tradition to cross the border into the Christian age would be the authors or

21
The fundamental study is Minio-Paluello 1952. Additional information about James’s work as a
translator may be found in the volumes of Aristoteles Latinus. For a handy list of translations of
Aristotelian works into Latin, see Dod 1982.
22
This would explain the curious relationship between ‘Alexander’s’ comm. in SE and Michael’s that I
mentioned above. Moreover, it might explain how ‘Alexander’s’ commentary came to contain
extracts from scholia on the Ethics, as some Latin texts suggest it did (see Ebbesen 1981a, vol. 3,
155–7); Michael himself at one time worked on the project of annotating the Ethics. The idea that
James might have known Michael was first proposed by Browning 1962, 8 (p. 401 of this volume)
in the paper that for the first time established that Michael lived in the twelfth century and worked
under the patronage of Anna Comnena.
23
This statement may need modification, but I consider it probable that the Byzantine commentary
on Arist. EN that Robert Grosseteste translated was put together by Michael, the latest identifiable
contributer to the work (cf. Mercken 1973). Michael certainly supplemented Alexander’s
commentary on Metaphysics 1–5 with scholia on 6–13, and he may at the same time have revised
Alexander’s work (cf. Ebbesen 1981a, vol. 1,201).
488 Aristotle Transformed

sources of much that James of Venice could find in Michael’s study. By that route
they acquired a rôle in the formation of Western scholasticism.
But the ground had been prepared by Boethius. This older contemporary of
Philoponus was the man who first of all made a part of Aristotle’s Organon
accessible in faithful Latin translation and accompanied with adaptations of
Greek commentaries. He gave the West such a basic course in ancient Aristotelian
Scholasticism that when the educational boom arrived in the twelfth century it
was natural to boost the progress of knowledge by acquiring more Aristotle with
Greek commentaries rather than ask for more Plato. Just as Boethius made the
West receptive to Philoponus and Michael, so the common origin of Western
and Islamic philosophy in ancient Greek scholasticism made it possible for the
West to acknowledge Avicenna and Averroes as people with a relevant message.

What, then, did the Greek scholastics do to the development of Western logic?
First, via Boethius, they equipped the West with a belief in the superiority of
Aristotelian logic over any imaginable rival.
Secondly, via Boethius they taught the West to take it easy if it looked as if the
ontology behind the logic was at loggerheads with the one implied by platonising
theologians. If Boethius had not taught the West that a logic for the language we
use about the sensible world may recognise no paradigmatic universals, and yet
be reconcilable with a strong form of realism which considers more than just the
sensible world, the dominance of Augustinian theology in early medieval times
might have made it impossible for Aristotelian logic to become established as a
respected discipline.
Thirdly, via Boethius’ logical monographs, they kindled an interest – untypical
of themselves – in the art of argumentation, and also in logical analysis of
questions in which no serious ontological principle seemed to be at stake.24 It is
not difficult to see that the ontological status of propositional contents is relevant
to the truth of the principle that there is only one eternal thing, God himself, and
also relevant to the question whether the patriarchs of the Old Covenant believed
in the same articles of faith as Christians do. Though these were not favourite
themes in late ancient times as they became in twelfth-century France, it is a sort
of problem that is kindred in spirit to those that Simplicius, for example, would
address about the ontological status of accidents or the like. But the Latin interest
in how arguments work, whether there is some sort of warrant for non-syllogistic

24
The best introduction to Latin medieval logic is still Pinborg 1972. Cf. also Kretzmann et al. 1982.
 Philoponus, ‘Alexander’ and the origins of medieval logic 489

inferences of the type ‘Socrates is a man, so he is an animal’ or whether the

principle that anything follows from an impossibility is valid – that is something
that can only be understood with reference to the fact that Boethius’ logical
monographs, which are not representative of a major interest of the ancients,
made up a major part of the set texts in early medieval schools.
Fourthly, they gave the West the first encouragement to embark on the wider
reading of Aristotle from c. 1150 which resulted in the ‘classical’ scholasticism of
the late thirteenth century with its important contributions to epistemology,
ontology, psychology and several other -ologies plus general linguistics.
It is remarkable that the Greek scholia/commentaries on the Prior and
Posterior Analytics and Elenchi seem to have had very scant diffusion in the
thirteenth century, and the very same Grosseteste who copies a page from
Philoponus does not to my knowledge repeat his predecessor’s explanation of
anything philosophically important. Though much research remains to be done
to see what influence the Greek commentaries on Ars Nova exerted, the general
picture is that they had a tremendous effect in starting a redirection of Western
logic, but soon lost readers because they were unsuited (too jejune) for use once
the train with Western scholars had acquired some speed down the new track.
Little details, like the story about the twinkling candle, remained in margins of
Aristotle’s text as a reminder of the time when the Greek scholia were the only
guide to the text.

Among Boethius’ logical monographs25 the one on hypothetical syllogisms and

the one about topics were particularly dear to the early Middle Ages. Boethius’
Topica (or de Differentiis Topicis) shows how all arguments depend for their
acceptability on certain axioms about genera, species, opposites and the like –
the so-called topical differences. Topical axioms – or ‘maxims’, as Boethius calls
them – may be like ‘of whatever a species is predicated, of that the genus may be
predicated’; the topical difference ‘species’ and the associated maxim makes ‘since
Socrates is a swine, Socrates is a beast’ an excellent argument. There is a long and
complicated history behind this piece of theory which is perhaps best understood
in a Platonic light to the effect that ‘genus’, ‘species’ and the like are ideas, which
may be unfolded in arguments. The first step in the unwrapping is the formulation
of the maxims which reveal the mutual properties of these quite formal ideas.
The next step consists in the formulation of some of the infinitely many

25
List of Boethius’ works in Ebbesen 1987a, 286 = this volume p. 373.
490 Aristotle Transformed

arguments of which the maxims are the ultimate generalisations or forms – with
a little good will and not making too much of the difference between conditional
and declarative propositions you can see ‘Socrates is a swine, therefore he is a
beast’ as an instance of the axiom ‘of whatever a species is predicated, of that the
genus is predicated’.26 But Boethius does not say that his topical differences are
Platonic ideas. He claims that they have a certain power which they may transfer
to arguments, but he does not say what they are. The medievals tried to find an
answer.

One most remarkable feature of early medieval philosophy in the eleventh and
early twelfth centuries is the occurrence of very strong forms of nominalism,
with the explicit claim that predication is a linguistic matter – a predicate is a
word, not some sort of thing27– and that there is nothing universal except for
words, not even universal concepts. Roscelin, Abélard’s teacher and the object of
Anselm’s wrath, seems to have been a truculent nominalist. Mitigated, and more
viable, forms of the position also occur. Thus Abélard’s variety which introduces
a special sort of non-thing, the status ‘to be white’, ‘to be a man’ etc., to account for
the applicability of one word to a multitude of things. Later varieties tend to
become conceptualistic – the primary universal is a concept, via its universality
its companion word is universal. This latter form is particularly popular in the
late Middle Ages, from the time of Ockham and Buridan. But it is a continuation
of the earlier forms in the sense that a main reason for introducing it was an
acceptance of a logical tradition with twelfth-century roots which seemed to
require an extensional semantics that only some sort of nominalism could
provide. This logical tradition had been pushed somewhat in the background in
the years 1200–1315, but had not quite disappeared.28 While the metaphysical
battles raged, logic hibernated, and got out again when realism in metaphysics
had left all metaphysicians severely battered.
Now, thirteenth-century mainstream philosophy was heavily Aristotle-
centred and owed a lot to the late ancient blend of Aristotelian and Platonic
thought – though sometimes by devious routes, as via Avicenna or Averroes.

26
For Boethian theory and its background, see Ebbesen 1981a, vol. 1, 106–26; Green-Pedersen 1984,
part 1.
27
Thus, e.g., the anonymous and unedited late eleventh-century commentary on Boethius’ Diff. Top.
called ‘Primum oportet’ (see Green-Pedersen 1984) says: ‘in quo destruuntur veteres qui
praedicatum et subiectum dicunt res esse, non voces’ – ‘and thus the traditional view is refuted
according to which predicate and subject are things, not words.’
28
cf. Ebbesen 1985.
 Philoponus, ‘Alexander’ and the origins of medieval logic 491

Like so much late ancient philosophy it first produces an inextricable tangle of

the real and the intelligible, then claims that there is a pure reality prior to the
tangle, but screened from the gaze of mortals.
People with a robust sense of reality like to maintain that before the rise of
thinking beings there was a real world, and one that would have been there even
if there had not now been humans to think of it; and similarly that there are
unknown realities now and after the extinction of intelligent life in the universe
whose being there is independent of whether they are ever discovered by man or
any other living organism; moreover, that things do not change their character
for being studied or not, and that the things that we can study would neither be
nor appear different to us if we were supplied with the extra powers needed to
see those that we actually cannot see.
In the late thirteenth-century type of philosophy which is often styled
‘modism’ all objects grasped by thought are in some way real in the way they are
thought – Socrates has singular, independent, actual, concrete reality; man has
universal, concrete reality; humanity has universal, abstract reality. There are as
many sorts of real things as we have angles from which to view them. Things are
irreducibly intentional. Yet this exuberant ontology is reduced by introducing
super-things, common natures. What is Socrates but human nature thought of
in concretion and individually? What is man but human nature thought of in
concretion and universally? What is humanity but human nature thought of in
abstraction?29
The undifferentiated natures make up a kind of mind-independent world, but
a strange one, consisting of non-existing things, each having less-than-numeric
identity, as Duns Scotus said. Things which upon inspection by a created mind at
least will or would manifest themselves as abstract or concrete, universal or
particular according to the perspective chosen. Like eternity it is a world that by
necessity disappears once you start to examine it – it is by definition inaccessible
to examination in its pureness.
As Ockham saw, whatever strange constructs theology may need to deal with
the Trinity’s peculiar way of being one, there is a simple way to avoid that sort of
twilight world in philosophy by letting universality be a property of words and
concepts only, and letting things come in one variety only: as numerically
distinct. You may still feel uncomfortable about things that nobody sees, but if
you want to make room for them in your ontology, you can do so without

29
This presentation of modism is grossly over-simplified. Some idea of the actual complexity may be
obtained from Ebbesen 1988. Some of the best studies of modistic logic and theory of science are
found in Pinborg 1984.
492 Aristotle Transformed

equipping them with the exotic property of being invisible when unmodified, a
property that seems to belong to principles of things rather than to real things
and thus, to speak in medieval terms, obscures the distinction between natures
and God, between creature and creator. And, surely, the world is created and
depends on God’s will for whatever sort of being it possesses.
It is often held that there is a connection between fourteenth-century
nominalism and a realisation that the notion of an omnipotent God, creator of a
contingent universe, each particular thing in which is also contingent with
respect to the whole, is difficult to reconcile with the notion of a universe, the
singular objects or facts of which either reduce to being God himself because
they are just principles seen in a certain light; or are all accessible to intellectual
inspection thanks to their having also a universal mode of being apart from
whatever other modes they may have.
But if this can with some plausibility be invoked as a major motive force
behind fourteenth-century nominalism, I see no evidence that its predecessor
about 1100 was similarly motivated. Nominalism looks implausible in a time
when philosophers were raised on Boethius, one of those Neoplatonic Aristotelians
of whose philosophy thirteenth-century modism was both historically and
systematically a continuation. How could the early nominalism arise?
A first hint at an explanation is given if we turn our attention to ninth-century
Byzantium, where Photius exhibits some extraordinarily nominalistic leanings
in an essay occasioned by a reading of Porphyry’s Isagoge with some late ancient
commentary.30 In contemporary France, you find not only strong forms of
realism, as with Scotus Eriugena, but also a strong attack on such realism and the
way it trades on the everyday sense of figurative expressions used in such
formulas as ‘species flow from genera’, ‘the more general engenders the less
general’ etc. The attack occurs in Ratramnus’ De anima ad Odonem; some of
Ratramnus’ heaviest ammunition was supplied by Boethius’ commentary on the
Isagoge. Eleventh- and twelfth-century defences of nominalism often adduce
supporting evidence from the Categories, the standard commentary on which
was Boethius’.
Both to the ancients and to the early medievals, Isagoge and Categories belong
together. They were read together and they were generally supposed to express
the same doctrine, only in slightly different areas. Aristotle’s book obviously was
the more important one, so their interpretation of the Categories was bound to
determine what they would say about universals when commenting on Porphyry.

30
Photius Amphilochia 77.
 Philoponus, ‘Alexander’ and the origins of medieval logic 493

To see how late ancient exegesis of Porphyry and Aristotle could inspire
nominalism, we should remember that the way the Neoplatonists made it
possible for themselves to wear both a Platonic and an Aristotelian mask was by
interpreting the Categories in a nominalistic-conceptualistic way, summarised in
the statement that the work is about words qua significative of things and the
slogan ‘words signify things via concepts’, plus the limitation that the theory of
categories does not apply when we turn to higher levels than the sensible world.
It will be convenient to remember that the discussion whether the subject of the
Categories is words or things takes up much space in ancient commentaries.
We may also remember that Boethius in his commentary on the Categories
follows Porphyry’s on the same work in all essentials. He, in turn, was not only
the first strong advocate of the word-qua-significative view, but also one of the
last philosophers with some sympathetic interest in Stoic thought, though he
may have known the school only at second hand. The Stoics were scarcely
genuine nominalists; they did, however, reject universal things, substituting for
universals some quasi-things that words signify, accepting as real only the
particulars that words denote.
I believe that Stoic anti-realism via its influence on Porphyry and his on
Boethius is part of the explanation why a nominalism could arise in the early
Middle Ages. It should be kept in mind, though, that the medievals got the Stoic
influence in a form which had obscured both the beauty and some of the
difficulties of the original theory. Thus there is something fishy about Stoic
particulars which is not quite so obvious in medieval nominalism. After all,
Stoics accept just one substance, and so like all monists get into trouble when
trying to find values of the x in their ‘(x) Fx’ formulas. Bertrand Russell, having
embraced a mathematical and hence, in his own words, incurably Platonic, logic,
throughout his long life continued the search for an italicised a, a genuine
individual to satisfy his propositional functions. The medievals did not quite
escape the problem, inter alia because they had inherited Porphyry’s definition
of an individual as a unique set of properties. The problem was brought to
consciousness in the fourteenth century, when it was shown that even Ockham-
style nominalists actually operated with nothing or everything – non-descript
thises – as carriers of properties or names.31 But at least the problem could be
more easily concealed in a tradition where, thanks to Aristotle, people were not
quite strangers to the notion of an entity whose formal component is self-
satisfying. A primary substance, in short.

31
For the late medieval recognition of the problem, see Ebbesen and Pinborg 1982, 120.
494 Aristotle Transformed

However, this is not quite enough to explain the characteristics of early

medieval nominalism. We should also not forget the theory of imposition of
words, which came to the Middle Ages from Porphyry via Boethius. In short
description it says that man has two different sorts of vocabulary: a primary one
created by naming sensible things; and a secondary one created by naming the
names that belong in the object-language. ‘Man’ and ‘rotund’ are words of the first
type, ‘noun’ and ‘verb’ of the second type.32
Boethius presents the theory of the two impositions in the introduction to his
commentary on the Categories, but he never uses it for anything except in one
place, where he shows that it can be used to solve a sophism which threatens the
transitivity of predication. The sophism is ‘Socrates is a man, man is a species,
ergo Socrates is a species’. Boethius explains that the transitivity breaks down in
the second premise, because there is no predication ‘in quid’: when we say ‘man
is a species’, the function of ‘species’ is to indicate that ‘man’ is predicable of
individuals only, and thus ‘species’ is ‘in a way, a name of names’.33
It was not Boethius’ idea to use the ‘name-of-names’ solution in this place, for
it is also mentioned by Dexippus, that wonderfully unfaithful pupil of Iamblichus,
always ready to deliver an unexpected kick at ‘those who fancy the Pythagorean
assumptions’.34 In this case, though, he may owe the not particularly Pythagorean
piece of doctrine to his teacher; or else his and Boethius’ common source was
Porphyry.
Anyhow, this little scholium – some five lines all told – was to be of tremendous
importance.

So here is a recipe for a potent brew, which may release significant amounts of
nominalism and hence should be labelled ‘dangerous’ and kept away from
children and innocent people in general:

Take
(1) Porphyry’s own claim that his treatise on universals is an introduction to
Aristotle’s Categories.
Add
(2) Boethius’ reproduction of Porphyry’s explanation that the Categories is a
book about words
32
For more details, see the extract from Ebbesen 1981a, vol. 1, 141ff (reprinted above, pp. 146ff ).
33
For Boethius on imposition and names of names, see Ebbesen 1987a, 287–99 (reprinted in this
volume, pp. 374–83). The Dexippus passage referred to below is in his comm. in Cat. (CAG 4.2) 26.
34
Dexippus in Cat. 17.1ff.
 Philoponus, ‘Alexander’ and the origins of medieval logic 495

and
(3) Boethius’ reproduction of Porphyry’s story about the two impositions
plus
(4) Boethius’ version of the Greek scholium about ‘man is a species’.

Stir the whole thing and you start to realise that many of the statements in Ars
Vetus and Boethius’ commentaries make best sense if you take ‘species’ as a name
of words.
Now add
(6) the Aristotelian claim of priority for primary substances, and Boethius’
emphatic endorsement of this (again echoing Porphyry).
Finally add
(7) Boethius’ doctrine of topics, which cries for an answer to the question,
what ontological status do the topical differences have?

Stir the mixture once more, and the nominalistic vapours start pouring out of
the cauldron. For now you suddenly see that of course the topical differences are
words of the second imposition, that is names of those words which are names
of individual things. All of logic will work without the assumption of universal
entities except for the linguistic ones. A lot of authoritative statements become
understandable. You need not devise strange explanations of why the auctores
sometimes say that a universal ‘is predicated’, sometimes that it ‘is said’ or ‘uttered’.
If any universal, hence any predicate, is a word, of course ‘is said’ or ‘uttered’ is a
good predicate of predicates.
If this last point sounds familiar, it’s all right. Similar arguments are related by
Dexippus and other ancient commentators. But for all I know, there was no true
Aristotelian nominalism in ancient times, at most a conceptualism with concepts
abstracted from sets of similar objects.
In the late eleventh century, if not before, some Westerners took the crucial
step from exegesis to philosophical claim, and said: ‘When Isagoge, Categories,
Topics, all make sense with a nominalistic interpretation, nominalism is the
truth.’35
The first, apparently rather crude versions came under heavy fire and were
succeeded by Abélard’s more sophisticated variety. However, this too disappeared

35
Most sources for early nominalism remain unedited. There are strong nominalist tendencies in
Garlandus Compotista and Anon. Comm. in Boeth. Diff. Top. ‘Primum oportet’ (both c. 1050/1100).
An account of (presumably early) nominalists’ arguments is given in Ps.-Rhabanus’ commentary on
Porphyry, in anonymous fragment in MS clm 29520/2 and other sources as well. I owe my
information about the early Latin Porphyry commentaries to Y. Iwakuma who is preparing an
edition.
496 Aristotle Transformed

before the end of the twelfth century. What did not disappear so fast was a
generally extensionalistic interpretation of predication, where at least the subject
term had to range over a non-empty set of individuals for a proposition to be
true.36 Also the view that the Aristotelian categories are primarily a classification
of words survived. Gilbert the Porretan saw things that way and noticed that the
logical behaviour of words differs according as they are used to speak about
things whose structure derives from nature – animals, plants etc., – things that
result from human decisions such as social positions and artefacts, or, finally,
things that belong in the sphere of rational and linguistic activity – words,
arguments etc. Hence he resolutely tripled the categories, reserving one set of ten
categories for the natural vocabulary, one for the ethical, and one for the rational-
linguistic.37
When the new wave of Aristotelian texts arrived about 1150, the Elenchi and
the Greek commentary on it were the first to achieve widespread attention. It
caused some confusion that the treatment of fallacies was different from the
traditional Western one, but at first the new material, which included for instance
a version of the Stoic ‘not someone’ (outis) sophisma – an argument against
crude Platonic realism – at first this new material just helped refine existing
logic, including the so-called theory of supposition or theory of what a term
stands for in given contexts.

However, a virus was spreading. The availability of Philoponus’ commentary

gave some people the courage to study the Posterior Analytics, and they were
infected with yearning for a theory of knowledge and science, such as that book
seemed to promise devoted readers, a theory vastly superior to anything the
West had previously known.38 Those infected with the disease would turn their
backs on the world of particulars, and many were persuaded that the theory of
which they expected so much bliss required acceptance of some sort of real
universals to be the immutable objects of knowledge and verifiers of sentences
about things with no present representatives. Avicenna persuaded almost
everybody that this was what should be done, and though few gave in to rampant
Platonism, many accepted some variant of the Avicennan distinction between

36
cf. Ebbesen 1981d.
37
See Ebbesen et al. 1983; Jacobi 1987.
38
The twelfth-century texts published in Ebbesen 1976 and 1979 are witnesses to the fascination that
An. Post. with Philoponus’ commentary could exert.
 Philoponus, ‘Alexander’ and the origins of medieval logic 497

existential and essential being, only the latter of which is required for objects of
knowledge.
The thirteenth-century turn towards realism is epitomised in the way the
subject of the Categories was redefined. It was now commonly said to be (ens)
dicibile (incomplexum) ordinabile in genere ‘that which is expressible, and simple,
in its hierarchical ordering’,39 i.e. realities qua signifiable by words. Porphyry’s
definition ‘words qua significative of things’ had been stood on its head. The way
was open for common natures to make their entry.
And now I have closed the circle. We are back in modistic times, waiting for
an English mind, waiting for Ockham. Medieval logicians surpassed the ancients.
But they were deeply influenced by the men from Porphyry to Philoponus; both
in nominalistic periods and in realistic ones. The ancient scholastics were
chimeras with an Aristotelian body barely managing to keep the Platonic head
and the Stoic tail from either running in opposite directions or engaging in a
deadly fight with each other. The medievals saw the monster from both ends.

Bibliography

Browning, Robert. 1962; ‘An unpublished funeral oration on Anna Comnena’,

Proceedings of the Cambridge Philological Society 188 (n.s. 8) 1–12. Repr. in Browning
1977 and in this volume as Chapter 17.
Browning, Robert. 1–12: Studies on Byzantine History, Literature and Education.
Variorum: London.
Dod, Bernard G. 1970: The Study of Aristotle’s Posterior Analytics in the Twelfth and
Thirteenth Centuries, Unpublished B.Litt. thesis, Oxford.
Dod, Bernard G. 1982: ‘Aristoteles Latinus’, in N. Kretzmann, A. Kenny, J. Pinborg (eds),
The Cambridge History of Later Medieval Philosophy, 46–79.
Ebbesen, Sten. 1976: ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and
Boethius. Ancient Scholasticism and twelfth-century Western Europe’, CIMAGL 16,
1–128.
Ebbesen, Sten. 1979: ‘Anonymi Aurelianensis I Commentarium in Sophisticos
Elenchos’, Cahiers de l’Institut du moyen âge grec et latin (CIMAGL ), University of
Copenhagen, 34.
Ebbesen, Sten. 1981a: Commentators and Commentaries on Aristotle’s Sophistici Elenchi.
A Study of Post-Aristotelian Ancient and Medieval Writings on Fallacies, vols 1–3 =
Corpus Latinum Commentariorum in Aristotelem Graecorum (CLCAG ) 7.1–3, Leiden.

39
So already in John Pagus’ commentary on the Categories (MS Padova BU 1589), presumably dating
from c. 1230.
498 Aristotle Transformed

Ebbesen, Sten. 1981b: ‘Suprasegmental phonemes in ancient and mediaeval logic’, in

H.A.G. Braakhuis et al. (eds), English Logic and Semantics from the End of the Twelfth
Century to the Time of Ockham and Burleigh. Acts of the 4th European Symposium on
Mediaeval Logic and Semantics = Artistarium, Supplementa I, Nijmegen, 331–59.
Ebbesen, Sten. 1981c: ‘Analyzing syllogisms or Anonymus Aurelianensis III – the
(presumably) earliest extant Latin commentary on the Prior Analytics, and its Greek
model’, CIMAGL 37, 1–20.
Ebbesen, Sten. 1981d: ‘The present king of France wears hypothetical shoes with
categorical laces. Twelfth-century writers on well-formedness’, Medioevo 7, 91–113.
Ebbesen, Sten and Pinborg, Jan. 1982: ‘Thott 5810 4, or De ente rationis, De definitione
accidentis, De probatione terminorum’, in A. Maierù (ed.), English Logic in Italy in
the 14th and 15th Centuries. Acts of the 5th European Symposium on Medieval Logic
and Semantics = History of Logic I, Naples, 111–46.
Ebbesen, Sten, K.M. Fredborg and L.O. Nielsen (eds). 1983: ‘Compendium logicae
Porretanum ex codice Oxoniensi Collegii Corporis Christi 250: a manual of
Porretan doctrine by a pupil of Gilbert’s’, CIMAGL 46.
Ebbesen, Sten. 1985: ‘OXYNAT: a theory about the origins of British logic’, in
P.O. Lewry (ed.), The Rise of British Logic, Acts of the Sixth European Symposium
on Medieval Logic and Semantics = Papers in Medieval Studies 7, Toronto, 1–17.
Ebbesen, Sten. 1987a: ‘Boethius as an Aristotelian scholar’, in J. Wiesner (ed.), Aristoteles:
Werk und Wirkung 2, Berlin and NY, 286–311. Repr. in this volume as Chapter 16.
Ebbesen, Sten. 1987b: ‘The way fallacies were treated in scholastic logic’, CIMAGL 55,
107–34.
Ebbesen, Sten. 1988: ‘Concrete accidental terms: late thirteenth-century debates about
problems relating to such terms as “album” ’, in N. Kretzmann (ed.), Meaning and
Inference in Medieval Philosophy, Dordrecht, 104–74.
Garlandus Compotista. 1959: Dialectica, L.M. de Rijk (ed.), Wijsgerige teksten en studies
3, Assen.
Green-Pedersen, Niels Jørgen. 1984: The Tradition of the Topics in the Middle Ages,
Munich.
Incerti Auctores. 1977: Quaestiones super Sophisticos Elenchos, ed. S. Ebbesen. Corpus
Philosophorum Danicorum Medii Aevi VII , Copenhagen.
Jacobi, Klaus. 1987: ‘Kategorien der Sittenlehre. Gedanken zur Sprache der Moral in
einem Logik-Kompendium des 12. Jahrhunderts’, in Beckmann, J.P. et al. (eds),
Philosophie im Mittelalter. Entwicklungslinien und Paradigmen, Hamburg, 103–24.
Kretzmann, N., A. Kenny, J. Pinborg (eds). 1982: The Cambridge History of Later
Medieval Philosophy, Cambridge.
Mercken, H.P.F. 1973: The Greek Commentaries on the Nicomachean Ethics of Aristotle
I = Corpus Latinum Commentariorum in Aristotelem Graecorum 6.1, Leiden.
Minio-Paluello, Lorenzo. 1952: ‘Iacobus Veneticus Grecus: canonist and translator of
Aristotle’, Traditio 8, 265–304. Repr. in Opuscula, The Latin Aristotle, Amsterdam
1972.
 Philoponus, ‘Alexander’ and the origins of medieval logic 499

Photius. 1986: Epistulae et Amphilochia 5, L.G. Westerink (ed.), Leipzig.

Pinborg, Jan. 1972: Logik und Semantik im Mittelalter, Stuttgart and Bad Cannstatt.
Pinborg, Jan. 1984: Medieval Semantics. Selected Studies on Medieval Logic and
Grammar, London.
Ratramne de Corbie. 1952. Liber de anima ad Odonem Bellovacensem, D.C. Lambot
(ed.), Analecta Mediaevalia Namurcensia 2, Namur and Lille.
Rossi, Pietro. 1978: ‘Tracce della versione latina di un commento greco ai Secondi
Analitici nel Commentarius in Posteriorum Analyticorum libros di Roberto
Grossatesta’, Rivista di filosofia neo-scolastica 70,433–9.
Rossi, Pietro (ed.). 1981: Robertus Grosseteste, Commentarius in Posteriorum
Analyticorum libros = Testi e studi per il ‘Corpus Philosophorum Medii Aevi’ II ,
Firenze.
Shiel, James. 1982: ‘A recent discovery: Boethius’ notes on the Prior Analytics’, Vivarium
20,128–41.
Shiel, James. 1984a: ‘A set of Greek reference signs in the Florentine ms. of Boethius’
translation of the Prior Analytics’, Scriptorium 38,327–42.
Shiel, James. 1984b: ‘Aristoteles Latinus III : Scholiorum in Analytica Priora
supplementa’, Bulletin de philosophie médiévale édité par la SIEPM 26, 119–26.
Sorabji, Richard (ed.). 1987: Philoponus and the Rejection of Aristotelian Science,
London.
Southern, Richard W. 1986. Robert Grosseteste: the growth of an English mind in medieval
Europe, Oxford.
500
 20

Aristotle’s doctrine of abstraction

in the commentators
Ian Mueller

1. Universals

Discussions of mathematical ontology are frequently blurred by the assertion

that mathematics or science generally is about universals. Consider the following
geometric proposition:

(1) If ABC and DEF are two triangles with equal sides AB, DE and BC, EF, and
equal angles ABC, DEF, then the sides AC, DF are also equal.

In one sense we may say that this proposition and any other general proposition
is ‘about’ universals: it enunciates a consequence of being a triangle. But there is
another, prima facie more straightforward sense of ‘about’ in which (1) is not
about being a triangle, but about triangles, not about universals, but about
mathematical objects, as I shall call them.1 This sense of ‘to be about’ can perhaps
be paraphrased as ‘to presuppose as existing’.2 Although there is a tendency for
individuals to hold analogous views about universals and mathematical objects,

1
Most of the ancients, including Aristotle, speak of mathematics and mathematical objects generally,
but carry on their discussion in terms of geometry in ways not always easily generalisable to include
arithmetic. It is convenient to follow them by using the word ‘mathematics’ and ‘mathematical
objects’, but much of what is said will be most easily understood in terms of geometry.
2
The notion of ‘presupposition of existence’ I have in mind is clearly formulated by W.V.O. Quine in
his frequently reprinted paper ‘On what there is’, in From a Logical Point of View, Cambridge Mass.
1953, 1–19. I, however, would like to use this notion more intuitively than he does without
presupposing the representation of a theory or proposition in the first order predicate calculus or
insisting on Quine’s monolithic conception of existence based on the use of bound variables in a
first-order theory. From my point of view it would be coherent to say that a theory expressed in the
vocabulary of Aristotle’s syllogistic was about universals, and even to go on to say that the universals
are not real entities, e.g. not substances. Worrying about these subtler issues would obscure my
main points.
I should also mention that in translations from the Greek concerning this topic I render peri as
‘about’ without any intention of imposing an interpretation on the passage translated.

501
502 Aristotle Transformed

there is no necessary connection between the two issues. One might, for example,
be a realist about universals and yet think that the triangles mentioned in (1) are
mental constructions or fictions, or be a nominalist and yet think that there are
real triangles (defined, say, by points in space).
I shall take for granted that mathematics is about universals in the first sense
I have mentioned, or, as I shall say, mathematics concerns universals, since I do
not think there would be serious disagreement on this topic among the ancient
commentators and their associates. However, like Aristotle and Plato themselves,
and like many modern commentators, they are not always clear about the
distinction between mathematical universals and mathematical objects. I do not,
of course, mean that they do not make the distinction clearly in some contexts,
but only that at crucial points they violate it, invoking truths about universals to
justify what seem to be claims about mathematical objects or vice versa. Since I
am primarily interested in the treatment of mathematical objects, I shall focus
on discussions of them, but it is impossible to avoid introducing material on
universals, since they intrude so frequently into these discussions.

2. Aristotle’s account of mathematical objects

It seems to me that there are four main interpretations of Aristotle’s account of

mathematical objects. According to one interpretation, there are no such objects
for him since he believes geometry to be about universals. I shall not discuss this
alternative, which although thoroughly in line with the doctrine of science of the
Analytics, hardly fits with the things Aristotle says about mathematical objects in
his major discussions of them in Physics 2.2 and Metaphysics 13.3. I shall simply
sketch the remaining alternatives and the major difficulties they face.

Alternative 1. Mathematical objects are physical objects which the mathematician

studies by leaving out of account their mathematically irrelevant properties. On
this interpretation, which seems to be gaining in popularity3 (if one can speak
about popularity in such rarefied matters), the only difference between
mathematics and, say, physics is the properties studied and left out of account by
students in the two fields: the mathematician worries about the shape of the
earth, but not its weight; the physicist studies the weight of the earth, but not the
distribution of crops over its surface. This interpretation is, I believe, foreign to

3
The view is put forward by Jonathan Lear in ‘Aristotle’s philosophy of mathematics,’ Philosophical
Review 91, 1982, 161–92.
 Aristotle’s doctrine of abstraction in the commentators 503

most Neoplatonists, who take for granted that physical things do not satisfy the
definitions of mathematics.4

Alternative 2. Mathematical objects are embodied in pure extension underlying

physical objects; the geometer’s abstraction of non-geometric properties enables
him to apprehend these things which satisfy the mathematician’s definitions.
This interpretation, which I have espoused,5 has the disadvantage of assigning to
Aristotle a theory about which one might expect him to have been more explicit
if he held it.

Alternative 3. Mathematical objects exist only in the mind of the mathematician

who reasons about triangles, angles, etc., which he conceives separately from
matter. This interpretation, which I think one can say is that of at least the
majority of the commentators, has not found much explicit support in modern
scholarship6 because of the generally accepted view that Aristotle’s conception of
knowledge is too realistic to allow for such a mentalistic account of mathematics.

My major reason for going through these alternatives is to make clear in advance
that the commentators’ interpretation of Aristotle is certainly disputable and
probably wrong. However, I am much less interested in the question of the
correctness of their interpretation than in the question of its origins and
development.

3. Ammonius and Philoponus

At 55,23ff of his commentary on the de Anima Philoponus distinguishes three

factors in natural things, form, matter, and the cause of the presence of form in
matter, and five methodoi for dealing with them, physics, particular tekhnai
(technologies), dialectic, mathematics, and first philosophy. The first two of these
methodoi deal with all three factors, the last three with form only, first philosophy
with forms completely separate from matter, dialectic with the formal causes of
things in matter, but not as cause. Here is Philoponus on mathematics:

4
I am inclined to think Aristotle also takes this position, but he is not at all explicit about it.
5
‘Aristotle on geometric objects’, in Jonathan Barnes, Malcolm Schofield, Richard Sorabji (eds),
Articles on Aristotle, vol. 3, London 1979, 96–107.
6
By ‘explicit support’ I mean only detailed interpretations and defences. Many general accounts of
Aristotle include brief descriptions of his philosophy of mathematics which I would classify under
alternative 3. See, for example, Abraham Edel, Aristotle and his Philosophy, Chapel Hill 1982, 106:
‘Mathematics in [Aristotle’s] view deals with the features of things that can be separated in thought
(that is, abstracted) and held separate.’
504 Aristotle Transformed

The mathematician also deals with forms which are inseparable from matter,
not with all, but only with those which can be separated by thought (epinoia).
These are the so-called common sensibles, such as magnitudes and figures.
For it is not possible to separate the form of flesh or bone and things of this
kind from matter even in epinoia. For if one has conceived softness and moistness
and ruddiness and whatever else gives form to flesh, one will have also
simultaneously conceived the appropriate matter, so that if one abstracts the
matter one also abstracts these things. Therefore, the mathematician gives
definitions of the per se (essential) forms capable of abstraction; without taking
matter into account, he gives these definitions in and of themselves. Consequently
in a definition he does not mention the cause either. For if he defined the cause,
he would certainly also have taken the matter into account. Since, then, he
doesn’t discuss matter, he doesn’t mention the cause either. For example: What
is a triangle? A figure contained by three straight lines. What is a circle? A
figure contained by one line. In these definitions matter is not mentioned and
therefore neither is a cause through which a specific form is in a specific matter.
Consequently, the mathematician will give [only] the causes for the features
which necessarily belong to figures (ta sumptômata ta kath’ hauta sumbainonta
tois skhêmasi); for example, he will say why the triangle has angles equal to two
right angles. (in DA 57,28–58,6)

Philoponus is here quite clearly dependent on Ammonius, who expresses very

much the same view in his commentary on Porphyry’s Isagoge (11,22–13,7).
Ammonius makes the following correlation between the parts of philosophy
and the entities they study:

theology: things entirely separate from matter in hupostasis (existence) and in

the epinoia of them;
physics: things entirely inseparate from matter both in hupostasis and in the
epinoia of them;
mathematics: things separate in one respect (kata ti), not separate in another.

And he explains the character of mathematics as follows:

For a circle or a square and such things cannot exist (hupostênai) by themselves
separate from some matter, and therefore they are inseparable from matter; but
when we see a wooden or copper or stone circle we get an impression of the
form of the circle itself in our dianoia (discursive thought), and we maintain
such impressions by themselves apart from matter (just as wax receives the
impression of a signet without also receiving any of the matter); and therefore it
is separable from the matter in the sense that it is separated in epinoia. (in Isag.
11,31–12,6)
 Aristotle’s doctrine of abstraction in the commentators 505

Ammonius introduces another trichotomy related to but not identical with the
distinction among the parts of philosophy in his commentary on Porphyry’s
Isagoge (39,14ff ), taking up a question Porphyry (1,9–14) had set aside
as requiring too deep and lengthy a discussion, namely the question of the
ontological status of genera and species (eidê). By the end of Ammonius’
explanation the term eidos seems to have taken on its more general sense of
form:

Consider a signet ring with a certain image (ektupôma), e.g., of Achilles,

and several wax tablets, and let the signet be impressed on the tablets. Someone
later seeing the tablets will know that they are all derived from one image; let
him be said to have the replica (tupos) which is the image in the dianoia. The
image (sphragis) in the signet is said to be prior to the many, that in the tablets is
said to be in the many, and that in the dianoia of the replicator is said to be
derived from (epi) the many and posterior to them. Let the same thing be
conceived in the case of genera and species. . . . There is an eidos in the Demiurge
[Plato’s Creator] like the tupos in the signet, and it is said to be prior to the many
and separate from matter. And there is also the eidos of human in each human,
like the images in the tablets, and it is said to be in the many and inseparable
from matter. And when we see that all individual humans have the form of
human, as we reach the posterior [image] by seeing the tablets, we replicate this
eidos in our dianoia; and this is said to be derived from or after the many and
posterior to them. Such eidê are separable from bodies, since they exist not in
body, but in soul, but they are not absolutely separable; for they cannot be
apprehended per se as Plato supposes the eidê prior to the many to be. (in Isag.
41,13–20, 42,6–16)

I will call the doctrine of mathematical objects expressed by Philoponus and

Ammonius abstractionism because commentators expounding it frequently
refer to mathematical objects as abstractions (ta ex aphaireseôs). My impression
is that Ammonius and Philoponus accept abstractionism, i.e. think it is a
correct account of mathematics. As is well-known, the form of a commentary
does not always make it possible to infer acceptance from failure to criticise,
but I believe the importance of this fact can be exaggerated. Simplicius certainly
manages to convey his own views on most subjects, and Philoponus does so
often enough to provide good grounds for thinking he accepts ideas he doesn’t
criticise.7

7
For a discussion of this point from a different perspective see H.J. Blumenthal, Chapter 13 above.
506 Aristotle Transformed

4. Alexander

It seems reasonably certain that the doctrine of abstractionism can be traced

back to Alexander of Aphrodisias, who made it authoritative as an interpretation
of Aristotle and authoritative for later philosophers who made heavy use of
what they took to be Aristotelian ideas in their own philosophising. The
evidence in the surviving texts for Alexander’s views is not as clear as one
would like, but I think the general tenor of it is unmistakable. Perhaps the
most explicit statement of abstractionism comes in the commentary on the
Metaphysics (52,15–19) where Alexander says that mathematical objects do not
exist per se, but by epinoia: ‘for matter and motion are separated from enmattered
things, with respect to which and with which mathematical objects have their
existence, and mathematical objects are left . . .’8 There are several other passages
in which Alexander associates mathematical objects with epinoia. In his
commentary on the de Sensu (111,17–19) he explains that he means by a
mathematical body something that doesn’t exist in its own right, but is
apprehended by epinoia separate from sensible properties. He appears to assert
much the same view at 52,13–16 of in Metaph., and the view is ascribed to him
by Philoponus (in GC 77, 8–15).9
Perhaps the strongest (and strangest) evidence that Alexander adopted an
abstractionist conception of mathematics comes in Simplicius’ discussion of
Physics 4.1, 208b22–5, a part of Aristotle’s list of the considerations which lead to
the belief that there is such a thing as place. The first is the fact of antimetastasis
(things taking each other’s place), the second that of natural place. In discussing
the second Aristotle insists that differences of natural place are not just differences
in position (thesis) relative to us. He then says:

This is shown also by the objects of mathematics: without being in any place
(topos), they none the less have a right and left corresponding to their position
(thesis) in relation to us, suggesting that things which are only so described
through their position do not by nature have a right and left (hôs ta monon
legomena dia thesin ouk ekhonta phusei toutôn hekaston).

8
See also in Metaph. 201,4–11.
9
Another text in the same vein is a report of Philoponus (in An. Post. 181,11ff ), who describes
Alexander as saying that ‘geometry uses only forms without substrata.’ The parallel statement in
Themistius’ presentation of the same passage says more explicitly that geometric forms exist in
physical bodies and are separated by the mathematician: ‘Geometry uses only the form of straight
line, which is not per se but always in some substance; for the straight is in either air or stone or
wood or something else. But the geometer examines the straight not as in one these things, but per
se’ (in Phys. 29,20–3).
 Aristotle’s doctrine of abstraction in the commentators 507

This is presumably the text read by Simplicius and adopted by Ross.10 It would
seem that Aristotle wants to affirm the reality of natural place by pointing out the
contrast between the natural or absolute place of real things and the relative
place of geometric objects: when a geometer speaks of drawing a straight line
between two points, he takes no account of direction; he does not say that one
point is above or below or behind or to the right of the other.
Simplicius tells us at in Phys. 526,16–18 that Alexander emended the text
starting from hôs to hôste monon noeisthai autôn tên thesin (so that their position
is merely conceived in the mind), claiming that Aristotle’s point is to make clear
that their position in relation to us (hê pros hêmas thesis) is a position in thought
(hê kat’ epinoian), and explaining, ‘if mathematical objects do not have11 their
principle per se, they could not have their position (thesis) per se’. Simplicius
adds that Alexander’s purpose is to make clear that Aristotle’s remark about
the place of mathematical objects is not an independent consideration but
merely a way of reinforcing the belief in natural place. However, Simplicius also
says that Alexander at one time believed the remark to be an independent
consideration, and tells us that it would then be paraphrased thus:‘if mathematical
objects, although not in place naturally, nevertheless occupy different places
with respect to us, it is evident that place is something; for things which are
by thesis (= convention) are derived from things which are by nature, as imagined
things are derived from sensible ones.’ Presumably, Simplicius’ paraphrase is
derived from Alexander, and is based on the text read by Simplicius, the second
sentence of the paraphrase being a rendering of the final clause of the passage.
It is hard to know how much to make of the comparison with the relation
between imagined and sensible things, but it is the only trace of a conception of
mathematical objects as being brought into existence mentally in the first
reading. However, Alexander’s second reading is not only mentalistic, it involves
a radical alteration of the text with, as far as we can tell, no manuscript authority.
Simplicius rightly rejects the alteration, but he manages to hold on to a mentalistic
reading.12

10
Ross, however, places the comma after thesin, a reading he thinks to be ‘equally likely’ as the ‘quite
likely’ one I have adopted. Simplicius divides his discussion of the passage where I have placed the
comma.
11
Reading ekhei for esti at 526,20; cf. 526,21.
12
One other passage suggesting that Alexander is a source for the abstractionist interpretation
is in Metaph. 199,19–22 where he says that, unlike mathematical objects, the objects of physics
are not abstractions because ‘it is impossible to separate by epinoia substance from physical
body’.
508 Aristotle Transformed

Alexander may also have been a major contributor to the tendency to run
together mathematical objects and universals.13 In his de Anima Alexander
distinguishes between per se forms, which are identical with nous (intellect) when
they are known by it, and enmattered forms which nous grasps by separating
them from their matter; from this process only a partial identification results
because ‘the thing known has its being in some matter’ (87,24–8; 88,10–16).
Alexander connects this doctrine with mathematics in the following passage:

In the case of enmattered forms there is no nous of them when they are not known
since their hupostasis (existence) is in being known. For universals and what is
common [in their case] have their huparxis (being) in enmattered particulars;
they become common and universal when they are known separate from matter,
and nous exists when they are known. If they were not to be known, they would no
longer be, so that these things are destroyed when they are separated from knowing
nous, since their being is in being known. And abstractions, such as the objects of
mathematics, are similar to these things. (Mant. 90,2–10)

There is, of course, no way to determine the precise degree of similarity which
Alexander has in mind. However, it seems to me likely that he thinks that
abstractions are separated from matter and exist only when they are known, but
I am not at all sure whether he means to say that abstractions are universals. The
doctrine that enmattered forms exist separately as universals only when they are
known makes reasonable sense as an interpretation of Aristotle, but if abstractions
were also universals of this sort for Aristotle, it is not immediately clear what
difference Aristotle would be bringing out by using the word ‘abstractions’ for
mathematical objects. On the other hand, if for Alexander the difference between
abstractions and separated forms lies only in the contrast between being
particular and being universal, one would expect him to focus more on the step
of generalisation in the process of separating forms. That is to say, one would
expect a clear marking of the stages enmattered form, separated particular form,
separated universal form, rather than a running together of the last two steps.14

13
For a discussion of some difficult passages bearing on Alexander’s treatment of universals see
Martin M. Tweedale, ‘Alexander of Aphrodisias’ views on universals’, Phronesis 29, 1984, 279–303.
14
For a discussion of the running together of these steps see Paul Moraux, Alexandre d’Aphrodise:
Exégète de la Noétique d’Aristote, Bibliothèque de la Faculté de Philosophie et Lettres de l’Université
de Liège 99, Liège and Paris 1942, 69–71.
Alexander does distinguish universals and the objects of a science at in Metaph. 199,35–9, when
he writes, ‘Every science and tekhnê is about the universal, which is eternal and is not sensible. For
even if one applies sciences and tekhnai to sensibles, they are not about sensibles. But although they
are about universals and such things, they deal with (pragmateuesthai peri) sensibles, not as
sensibles and particulars, but as such and such and sharing in the common.’ (I read metekhonta for
metekhei in line 39.)
 Aristotle’s doctrine of abstraction in the commentators 509

5. Syrianus

Aristotle’s theory of mathematical objects was clearly anti-Platonic in motivation,

and prima facie abstractionism would seem difficult to reconcile with the central
role assigned to mathematics in philosophical education in the Republic. Here is
how Philoponus attempts to reconcile them:

Mathematical objects . . . have an immaterial ousia (essence) even if they don’t

have an immaterial huparxis (being), so that becoming accustomed to act in
relation to these things and advancing in an immaterial way, one will reach
things which are entirely separate from matter, I mean divine things. Therefore
Plotinus says, ‘Let the young be led through mathematics to get them used to
immaterial nature (pros sunethismon tês asômatou phuseôs).’ (in DA 3,8–13)

Philoponus here quotes part of a sentence from Ennead 1.3.3; Plotinus actually
recommends the study of mathematics as a preliminary to dialectic to get people
used to incorporeal thought and belief (pros sunethismon katanoêseôs kai pisteôs
asômatou), and there is, I think, no clear evidence that he would understand
mathematical objects in the way Philoponus and Ammonius do. At least as early
as Ammonius, the phrase quoted by Philoponus became a kind of slogan,
repeated again and again.15 However, it seems to me that the abstractionist
conception of mathematics really provides an inadequate account of the role of
mathematics in philosophical ascent. This was certainly Syrianus’ view, and I
now want to discuss some of the things he says about abstractionism in his
commentary on the Metaphysics.
At 3.2, 996a12 of the Metaphysics Aristotle introduces as one of the puzzles
(aporiai) to be discussed the question whether mathematical objects are substances
and, if so, whether they are separate from or inherent in sensibles. Commenting on
this passage, Syrianus (12,28ff ) affirms their substantiality, and continues,

15
I have found versions of it in the following places. Iamblichus Comm. Math. 55,19 (without naming
Plotinus), Ammonius in Isag. 12,25–7, Proclus in Eucl. 20,21–2, Philoponus in Cat. 6,15–16, in DA
3,12–13, and in Nic. 1 and 27, Simplicius in Phys. 14,4 (without naming Plotinus), Asclepius in
Metaph. 98,7–8, 104,13–15 (both without naming Plotinus), and 151,4–6, Olympiodorus Proleg.
9,27–10,2, David Proleg. 59,17–19, and [Galen] de Part. Phil. 7,19–20.
It is interesting to note that Alexander gives essentially the same defence of the study of
mathematics as that offered by Philoponus in the name of Plotinus, although he uses an Aristotelian
rather than a Neoplatonic conceptual framework: ‘For geometry does not just theorise about
sensibles but about things arising from sense and intelligibles . . . Furthermore, geometry accustoms
young people to theorise about lines and planes and solids on their own (idiai), and none of these
things are sensible. For being able to separate from one another by logos things which differ from
one another in substance but cannot be separate from one another in hupostasis or huparxis is most
necessary for philosophy. For this is the way one apprehends the principles of natural things, matter
and form. For these two things are inseparable in hupostasis’ (Alexander in An. Pr. 4,3–14).
510 Aristotle Transformed

For one can see figure and number and natural surface and its limits in the
sensible activities of nature; and these things also exist in our imagination and
opinion, whether they are taken by abstraction from sensibles, as Aristotle
thinks, or whether they are completed by us from the substantial forms of the
soul. These imagined and opined things participate in being, but they are not
substances; they should be referred to quantity or quality or another category.
But the substantial logoi of the soul which contain these things are substances.
But if one were to consider the paradigms of these things in nous or among
intelligibles, he would see number and figure and magnitude itself counted
among the most primary substances. (in Metaph. 12,29–13,3)

Syrianus here invokes a standard Neoplatonic hierarchy:

Nous – forms
Soul – logoi of the forms
Physical world – embodiments of the logoi in matter

However, below soul and probably above the physical world Syrianus inserts
imagination16 and mathematical objects corresponding to it. Here he admits as
options the possibility that these objects come to be by abstraction from sensibles
or as a kind of completion of things in the soul. Elsewhere he leaves no doubt
that he favours the latter alternative, which he associates with Plato and the
Pythagoreans. I have in mind particularly two other passages in which Syrianus
comments on Aristotle’s criticism of realistic conceptions of mathematical
objects and puts forward his own account.
In the first, Syrianus is dealing with a general passage (Metaph. 13.2, 1077a14–
20) in which Aristotle dismisses the attempt to treat mathematical objects
as separate entities. Syrianus (91,20) introduces the distinction between the
universal which is cause of the sensible and the universal which is generated
posterior to it, and says that there are similarly two kinds of magnitude, the
one in dianoetic logos, which is cause of the existence of the imagined form,

16
I will not discuss opinion here because I am not sure how much can be made of Syrianus’ mention
of it. I know of only one analogue. In an intriguing passage (in Eucl. 94,19ff ) in his discussion of
Euclid’s definition of point Proclus gives his answer to the problem of how the geometer can
imagine a partless point since the imagination receives things in shapes and part by part
(morphotikôs kai meristôs). He then turns to the interpretation of the allegedly Pythagorean
definition of the point as a unit with position, and explains the difference between the point and the
unit indicated by this definition: ‘When they speak of a unit having position, I think they are
indicating that the unit and number (I mean monadic [as opposed to eidetic] number) has its
hupostasis in doxa (opinion); and therefore in each soul each number, e.g. 5 and 7, is one and not
many and free of figure or adventitious form. But the point comes forth (proteinetai) in imagination
and comes to be in a kind of place . . .’ (in Eucl. 95,26–96,7). Obviously the suggestion in this
paragraph is that doxa is to arithmetic as imagination is to geometry, but unfortunately Proclus
does not develop the comparison, and we possess no discussion of arithmetic analogous to in Eucl.
 Aristotle’s doctrine of abstraction in the commentators 511

the other by abstraction from the sensible. The absence of a reference to the
universal (or magnitude) in sensibles is presumably to be explained by its
irrelevance to the topic of mathematics. Syrianus goes on to deny that abstracted
things are the concern of geometry on the grounds that we do not perceive
anything with the accuracy, character, and independence of matter exhibited by
the objects which the geometer discusses.

Geometry deals with imagined things insofar as they coexist with the substantial
logoi of the dianoia, which also provides the cause of demonstration; or rather
geometry wants to deal with the partless logoi of the soul, but is too weak to use
noîseis not involving imagination and so spreads out the logoi into imagined and
extended figures and magnitudes, and studies the logoi in them. (in Metaph.
91,29–34)

Syrianus continues by comparing the geometer’s use of imagination to his use of

diagrams; just as one would not say that the geometer is proving things about his
diagrams, so one should not say he is proving things about the imagined figures.
The strongest evidence for this is that proof is of the universal, but every imagined
thing exists as an individual.
In the other passage (95,29ff ), Syrianus takes on those who try to avoid his
claim that abstractionism is incompatible with the accuracy of geometric figures
by saying that ‘we add what is lacking and make them more accurate’. Syrianus
asks where we get our ability to do this unless there are pre-existing logoi in our
soul. He adds, presumably having in mind the abstractionist doctrine that
mathematics does study sensibles in a way,17 that adding things will make the
abstractions less rather than more accurate. Syrianus admits that we can derive
images of things from sensibles, but again insists that our making them more
accurate depends on the pre-existent logoi. And it is the sense of approximation
to these logoi which makes mathematics an appropriate vehicle of philosophical
ascent. Syrianus also concedes that we are reminded of the Forms by sensibles,
but distinguishes this from asserting that the dianoia introduces a Form derived
from them. Concern with sensibles only blinds the soul, and such blinding is
certainly the result of dealing with abstractions.18 He continues by arguing that
abstractionism makes the relation of mathematics to knowledge of the forms
like the relationship between the rhetoric teacher’s artificial speeches and real

17
In Metaph. 13.3 Aristotle uses the slogan ‘sensibles not as sensibles’ to characterise the things
studied by mathematics.
18
In this connection Syrianus attacks Alexander for saying that a person moulds intelligibles by
himself (noêta atta heautêi anaplasasa).
512 Aristotle Transformed

legal arguments: there are, indeed, similarities, but the fundamental contrast
between the real and the artificial remains. Such a view cannot explain the ascent
of our soul nor why mathematics is comparable to a bridge leading us from
sensibles to intelligibles, nor why it isn’t better to deal with sensible magnitude in
the way in which it is rather than in a way in which it neither is nor can be.
The fullest development of Syrianus’ position on geometric reasoning is given
by his pupil Proclus in his commentary on book 1 of Euclid’s Elements (notably
at 48,1–57,8, which includes a strong attack on abstractionism).19 According
to this view, the geometer deals with forms or logoi in the dianoia, but he does
so by projecting images of them into his imagination. These images are perfect
embodiments of geometric definitions because they are derived from the forms,
as they could not be if they were somehow put together out of sense perceptions.
Equally, geometry can play its Platonic role of bridge or ladder leading from the
sensible to the intelligible world because it is the study of these projections from
the higher world.
There is a clear sense in which abstractionism is the opposite or converse of the
doctrine of Syrianus and Proclus, which I will call projectionism. In abstractionism
the objects of mathematics are derived from sense perceptions of such features of
things as their shapes, but these perceived features are somehow mentally
separated from their material substratum, this separation from matter being what
makes mathematics a suitable bridge to the intelligible world. However, there do
not seem to be any important differences in the metaphysical or epistemological
underpinnings of abstractionism and projectionism. Proponents of both accept
the idea that there are forms and logoi which exist prior to sensibles and forms or
mathematical objects which are derived from sensibles. This is brought out clearly
in Philoponus’ characterisation of the first philosopher which immediately follows
the passage I quoted at the beginning of this chapter:

But the first philosopher will discuss forms which are entirely separate from
matter. For there are in our soul logoi of the enmattered forms, and these also
exist transcendentally (exêirêmenoi) in demiurgic nous. And of the things in our
soul those in the imagination are extended, those in the rational part of the soul
are partless and unextended. And the geometer discusses the extended forms in
the imagination; for he uses imagination as a drawing board, acting by way of
division (meristôs) and measuring and dividing intervals. (in DA 58,6–13)

19
I discuss Proclus’ views in my paper ‘Mathematics and philosophy in Proclus’ commentary on Book
1 of Euclid’s Elements’, in Jean Pépin and H.D. Saffrey (eds), Proclus, Lecteur et Interprète des Anciens,
Paris 1987, 305–18. What I say about Proclus in this chapter is based on the more extended
discussion in that paper.
 Aristotle’s doctrine of abstraction in the commentators 513

However, the shared vocabulary and metaphysical apparatus should not lead us
to suppose a shared conception of mathematical objects. Nor should we think
that for Philoponus the imagination is the projection screen of Syrianus and
Proclus. For elsewhere Philoponus says of the imagination that it is included in
and has its source in perception (in DA 78,24–6), and ‘takes tupoi (imprints) of
sensibles from aisthêsis (sense perception) and weaves them together inside
itself ’ (in DA 5,38–6,1).
Philoponus and Ammonius do not, I think, have anything to say on the question
of the accuracy of abstractions, an issue on which Syrianus focuses much of his
attack. Their silence may be due to the inability to respond to the criticism, but I
suspect that other factors are involved. There survive slight traces of attempts to
argue that sensible things do satisfy mathematical definitions, e.g. that the division
between a shaded and a lighted area is a perceptible breadthless length,20 but the
evidence does not suggest that this position was developed fully or widely accepted.
Another, presumably more widely accepted position, which might even be taken
for granted, would be that separation from matter by itself yields perfection, i.e.
that matter is the source of the failure of sensible objects to satisfy mathematical
definitions. A final important factor is the failure to distinguish the conception of
abstraction as leading to the apprehension of particular shapes separate from their
physical embodiment and as leading to the apprehension of universals.

6. Simplicius21

Although the word ‘imagination’ occurs in my last quotation from Philoponus,

he, Ammonius and Alexander more frequently use the word epinoia in

20
Philoponus in Cat. 85,8–86,1. Analogous material is found in Proclus in Eucl. 100,5–19 (cf. 114,20–
5) and [Heron] Deff. 16,5–11 (cf. 20,18–22). In the light of Syrianus’ characterisation of Alexander’s
view of mathematical objects (see n. 18), it is perhaps interesting to note that Philoponus
characterises his considerations as showing that the notion that there is such a thing as breadthless
length is not an anaplasma (moulding) of our dianoia but something in the nature of things.
21
In my discussion of Simplicius I use only the de Anima commentary traditionally ascribed to him.
The authorship of this commentary has been placed in doubt by F. Bossier and C. Steel in ‘Priscianus
Lydus en de “In de anima” van Pseudo(?)-Simplicius’, Tijdschrift voor Filosofie 34, 1972, 761–822. For
arguments minimising the doctrinal significance of their assignment of the commentary to
Priscianus see Ilsetraut Hadot, Le problème du néoplatonisme Alexandrin: Hiéroclès et Simplicius,
Paris 1978, 193–202 and Chapter 12 above, and for sceptical arguments concerning the reassignment
see H.J. Blumenthal, ‘The psychology of (?) Simplicius’ commentary on the de Anima,’ in
H.J. Blumenthal and A.C. Lloyd (eds), Soul and the Structure of Being in Late Neoplatonism,
Liverpool 1982, 73–5 with the discussion on 94–5. Since I make no major use of the undisputed
commentaries of Simplicius, nothing I say turns on the authorship of the de Anima commentary,
and it seems simplest to use Simplicius’ name in discussing it.
514 Aristotle Transformed

connection with mathematics, a word suggesting that mathematical objects are

purely mental.22 Imagination plays a much more prominent role in Simplicius’
discussions of abstraction in his de Anima commentary, discussions which
offer other interests as well. Commenting on a passage (3.4, 429b18) in which
Aristotle invokes the distinction between to euthei einai (the form of straightness)
and to euthu (the straight line) and uses the word ‘abstractions’ (ta en aphairesei),
Simplicius says,

Aristotle usually believes and says that mathematical objects are abstractions, in
this differing from the Pythagoreans who investigate the logoi which are
projective of the mathematical tupoi in the imagination and are substantial in
the corporeal life of the soul. However, Aristotle says that the mathêmata (objects
of mathematics)23 are in the imagined tupoi themselves, concerning which the
mathematicians in truth carry on all their business. The Pythagoreans, looking
to the living logoi which are projective of the tupoi, suppose mathêmata to be
substances. However, Aristotle denies that the tupoi are substances, and says they
arise from abstraction in that the soul receives them because of their imitation
of sensible things. Therefore also, the soul does not project them by its own
activity. (in DA 233,7–17)

Here the contrast between Aristotelian and Pythagorean conceptions of

mathematics is made explicit, but the primary contrast does not relate to the
distinction between abstractions and projections, but to that between concern
with images believed to be derived from sense perception and concern with the
logoi which are projective of images not believed to be so derived. Simplicius
goes on to argue that the accuracy of mathematical objects shows they are
projected from within, although he concedes that such projection is impossible
without apprehension of sensibles. However, his main concern is to explicate
the distinction between to euthu (the imagined straight line) and to euthei einai
(the form of straightness).
Simplicius might appear to be siding straightforwardly with Syrianus and
Proclus here, but another passage (276,28ff ) in which he proposes to explain
Aristotle’s notion of abstraction suggests a more complicated situation. Simplicius
first recalls the way Aristotle distinguishes between mathematics and physics in
Book 2 of the Physics:

22
Another word frequently used in connection with abstractionist discussions of mathematics is
logos, the sense being that mathematical objects are separated by an act of definition.
23
Simplicius here follows a standard practice among later authors of using mathêmata as synonymous
with mathêmatika, as Philoponus announces he himself does at in DA 3,8–9. Compare Plato’s
specification of the highest mathêma as the Idea of the Good. (Republic 6, 505A. I thank Hayden
Ausland for this reference and for a correction in one of my translations.)
 Aristotle’s doctrine of abstraction in the commentators 515

The mathematician who investigates the kath’ hauta sumbebêkota (essential

accidents) such as figures, magnitudes, motions, numbers and consonant ratios,
needs perception itself for his apprehension. The physicist investigates both the
shapes and their consequences which attach appropriately to substances,
[making use of] causes of both the forms and of the compound substances
themselves as substances. Therefore the subject of physics is a kind of philosophy
since it deals with substance, and knows things which have causes on the basis of
their causes. However, mathematics is not [a kind of philosophy], but, as Plato
says, it resembles a bridge. For, as the same philosopher declares, there is a
beginning for it which the mathematician does not know and therefore does not
know about the middle things or the conclusion from them,24 because he leaves
out the causes by not investigating substances. For the absolutely authoritative
cause is substance or a cause stronger than this. (in DA 276,28–277,1)

Here Simplicius assigns the role of bridge to a mathematics which he concedes

to be at best a minimal form of knowledge. However, he then goes on to contrast
the mathematics he has just described with what the Pythagoreans call
mathematics:

I am now speaking of mathematics not in the manner of the Pythagoreans

who investigate the living logoi which are productive by projection of extensions
and figures, whether in natural bodies or imaginations (for that is philosophy),
but of the whole treatment itself of the projected things, e.g. figures and
magnitudes and motions, which Aristotle usually calls abstractions. (in DA
277,1–6)

Thus for Simplicius the contrast between mathematics as viewed by Aristotle

and mathematics as viewed by the Pythagoreans is the contrast between
mathematics and a kind of philosophy. Simplicius gives nothing but a general
characterisation of Pythagorean philosophy, but I think it is certain that he has
in mind the kind of mystical and symbolic interpretation of mathematical
concepts which we associate with Pythagoreanism and Neopythagoreanism and
which we find in such Neoplatonists as Syrianus and Iamblichus. Simplicius next
returns to the topic of mathematics, stressing the imaginary and non-substantial
character of mathematical objects, and giving examples of mathematical
statements to substantiate his claim: all the radii of any circle are equal; all the
circles about a single pole move together; every even-times even number is
divisible down to the unit; and every 2:1 ratio is composed of 4:3 and 3:2. He

24
arkhê . . . autêi . . . ha mê oide, mesa de kai teleutê ex hôn mê oide, obviously a remembrance of Plato
Republic 7, 533C3–4, but Plato nowhere compares mathematics to a bridge.
516 Aristotle Transformed

then repeats his claim about the insubstantial character of the objects of ordinary
mathematics, but quite suddenly denies the adequacy of the account he has
assigned to Aristotle both because imagined things transcend sensibles in
accuracy and because the activity involved in mathematical reasoning is
incompatible with mere passive reception through sense. At this point it would
appear that Simplicius has contradicted himself, first accepting and then rejecting
Aristotle’s account of mathematical objects as abstractions. But in his summary
he makes it clear that he accepts Aristotle’s account for ordinary mathematics,
but rejects it for another form of reasoning, which presumably is, or includes,
Pythagorean mathematics:

It has been said then what abstractions are, namely mathematical objects, and
not the causes of things like substances, but extended things and sumbebêkota
(accidents). The reason for the name is that they are not investigated as they
exist, i.e. with substance, but without it by those the majority of people usually
call mathematicians. Perception and imagination are the modes of apprehension
co-ordinated with these sensible and imagined sumbebêkota. For these modes of
apprehension make judgments about sumbebêkota, but not about substance; for
logos and nous apprehend substance. . . . The tupoi which come to the imagination
from sensibles are not substances. Such things are not only . . . investigated by the
co-ordinated modes of apprehension, but also by our nous, sometimes
interwoven with imagination and sensation and acting only concerning them
without apprehension of their substances. (Some people call this mathematical
nous, deriving the name from the person the many customarily call a
mathematician.) But sometimes nous does not attend to these things but to their
substantial causes, the formal and the composite, and subsequently grasps the
sumbebêkota which are caused in terms of their causes, this stronger nous
apprehending substances. (277,30–278,6)

I am not sure I understand Simplicius’ position fully and I certainly do not see
how it could be worked out with complete consistency. However, it appears to
me that he wants to treat abstractionism as a correct account of ordinary
mathematics, to which he also wants to assign the Platonic role of bridge or
ladder. To this extent he seems to agree with Philoponus and Ammonius,
although he seems to place more emphasis on the role of imagination in this
mathematics than they do. Syrianus denies that a mathematics based on
abstraction can play this role, and espouses a projectionist doctrine according to
which real mathematics deals with logoi by projecting them into the imagination.
Simplicius admits the validity of this subject, but makes it into a form of
philosophy as opposed to (ordinary) mathematics.
 Aristotle’s doctrine of abstraction in the commentators 517

It is not entirely clear whether Syrianus thought of projectionism as applying

to both ordinary and Pythagorean mathematics or only to Pythagorean
mathematics. It seems undeniable that Proclus intended it to apply to both since
he defends projectionism in a commentary on ordinary geometry, although he
also invokes a number of Pythagorean ideas in the commentary; it is also quite
certain that he considered mathematics so described to be the Platonic bridge.
But I am inclined to think that Syrianus thought of projectionism rather as an
account only of Pythagorean mathematics of the kind labelled philosophy by
Simplicius. Certainly in the Metaphysics commentary he shows much more
interest in this sort of thing than in ordinary mathematics. And in one passage
(101,22–103,12) he dismisses Aristotle’s account of mathematics in favour of
the genuine opinions of the Pythagoreans, the subjects of which he proceeds
to list. The list is a fairly close approximation to a table of contents of Iamblichus
de Communi Mathematica Scientia, and at the end of it Syrianus refers the
person eager to learn more to the work of Nicomachus and Iamblichus,
the latter of whom is said to provide constructions and proofs and more
intelligible apprehensions (noerôterai epibolai). No modern, or at least none with
a background in analytic philosophy, would describe the contents of Comm.
Math. in terms of constructions and proofs, although the word ‘intelligible’
might be appropriate as long as the Neoplatonic meaning were borne in mind.
The important point is that Comm. Math. is a work of Pythagorean philosophy
in Simplicius’ sense. Indeed, like Simplicius, Iamblichus insists on the difference
between Pythagorean and ordinary mathematics:

Pythagorean mathematics is not like the mathematics pursued by the many. For
the latter is largely technical (tekhnikos) and does not have a single goal, or aim
at the beautiful and good, but Pythagorean mathematics is pre-eminently
theoretical; it leads its theôrêmata toward one end, adapting all its assertions to
the beautiful and good, and using them to conduce to being. (Comm. Math.
91,3–11)

The similarities between Comm. Math. and the first prologue of Proclus’
Euclid commentary have often been noted, but I believe it can be shown that
the two texts diverge precisely in Proclus’ dropping of many of the most
Pythagorean features of Comm. Math.25 There are clear suggestions of

25
For details of the argument see my paper ‘Iamblichus and Proclus’ Euclid commentary’, Hermes 115,
1987, 334–48.
518 Aristotle Transformed

projectionism in Comm. Math. (see especially 33,19–34,18), but whether or

not Iamblichus is the originator of the doctrine, he seems to be our earliest
source, and I would guess he was an important source for the combination of
projectionism and glorification of Pythagorean mathematics we find in Syrianus.
I would claim that Proclus transforms projectionism into a philosophy of
ordinary mathematics, although he by no means abandons completely his
teacher Syrianus’ view of the profundity of the Pythagoreans. Simplicius, on
the other hand, seems to restrict projectionism to Pythagorean mathematics,
while conceding the validity of abstractionism as an account of ordinary
mathematics.

7. Porphyry

In this last main section I want to argue that Porphyry was an abstractionist.
His being one would help explain the acceptance of the doctrine by later
Neoplatonists. However, the evidence in his case is rather sparse. One piece
is provided by Proclus, who remarks explicitly on the divergence of his
projectionism from Porphyry’s account of geometry (in Eucl. 56,23–57,4),
although he does not tell us what Porphyry’s account was. Moreover, Proclus’
acknowledged borrowings from Porphyry in the Euclid commentary all relate
to quite ordinary geometric arguments or terminological points, having
nothing to do with Pythagorean mathematics. Porphyry’s distance from the
Pythagoreanising of Iamblichus and Syrianus is perhaps most strongly confirmed
by his commentary on Ptolemy’s Harmonics, in which he completely accepts
Ptolemy’s criticisms of the arbitrary procedures of the Pythagoreans. Finally
in the same commentary there is an interesting passage (13,15–14,28) in
which Porphyry comments on Ptolemy’s statement that the criteria in harmonics
are hearing and logos, hearing for matter and what is passive, logos for the form
and cause.
In his comment Porphyry introduces the notion of ‘psychic matter’, pointing
out that all the underpinnings are immediate matter for the superstructure
(panta ta hupobebêkota prosekhôs hulai tôn epanabebêkotôn). He illustrates
this relation by contrasting external and superior nous with the nous which
it uses as matter in its own activity. The correspondence with Iamblichus,
Syrianus, and Proclus is here quite clear, but Porphyry goes on to give a highly
Aristotelian account of judgment in which the soul is said to ‘tear off ’ (apospan)
forms from matter. He describes an epistemological sequence consisting of
 Aristotle’s doctrine of abstraction in the commentators 519

aisthêsis (perception), antilêpsis (apprehension), and then doxastic hupolêpsis

(conception), in which what has been received is written down for the soul
through logos, as if on a writing tablet. After these things comes imagination,
which does not copy sense impressions exactly, but introduces accuracy into the
images26 and presents them to the soul as ennoiai (concepts), which, when they
are made firm, produce epistêmê (scientific understanding), which in turn ‘lights
up’ nous (intellect). This account of the activating of nous remains ‘Platonic’
because the activation is a stimulation to recall rather than an implanting of
something new,27 but the mechanism is purely Aristotelian in substance.
Particularly important for the philosophy of mathematics is the fact that
imagination produces ‘accurate’ images. Porphyry does not mention mathematics
and so does not say that the images of imagination satisfy the conditions of
mathematical definitions, but it does not seem unreasonable to read this notion
into what he does say.

8. Summary

In this paper I have argued for the following account of the treatment of
abstractionism in later antiquity:
(1) Alexander established abstractionism as an interpretation of Aristotle’s
philosophy of mathematics. Alexander’s account was accepted, with insignificant
variations, as an interpretation by all subsequent philosophers.
(2) The doctrine of abstractionism was accepted as a true account of
ordinary mathematics by Porphyry, Ammonius, and Philoponus, who saw
ordinary mathematics as a Platonic bridge from the sensible to the intelligible
world.
(3) Iamblichus put forward the doctrine of projectionism as an account of
Pythagorean mathematics, which he glorified at the expense of ordinary

26
The text here is difficult: tritê d’esti . . . phantasia ouk arkoumenê tôi tês prosagoreuseôs eidei kai
tôi tês anagraphês, all’ honper tropon hoi tous katapleontas eikonizontes ê kata tous tois sumbolois
parakolonthountas tên akribeian tês homoiotêtos eklogizonta; houtô kai hautê tou pragmatos
hapasan tên morphên eklogizomenê, hopotan touton ton tropon akribôsêi, tote apetheto en têi psukhêi
to eidos. (13,29–14,3, using a textual correction of Bengt Alexanderson, Textual Remarks on
Ptolemy’s Harmonica and Porphyry’s Commentary, Studia Graeca et Latina Gothoburgensa 27,
Göteborg 1969, 20.)
27
See on this point W. Theiler’s review of Düring’s editions of Ptolemy’s Harmonics and Porphyry’s
commentary, Göttingische gelehrte Anzeigen 198, 1936, 203.
520 Aristotle Transformed

mathematics; he was followed by Syrianus, but Proclus transformed projectionism

into an account of ordinary mathematics to which he restored its Platonic role.
(4) Simplicius accepted the ‘Porphyrean’ position on ordinary mathematics,
but elevated Pythagorean mathematics, of which he thinks projectionism
provides the correct account, to the level of philosophy.28

28
This paper was written while the author held a fellowship from the American National
Endowment for the Humanities. It was revised as a result of discussion after presentations at
the Institute of Classical Studies in London and the West Coast Colloquium on Ancient
Philosophy in Berkeley.
 Note on the frontispiece
‘Aristotle and Alexander of Aphrodisias’ by Ulocrino
Donald R. Morrison

This charming small Renaissance plaquette is also rather mysterious.1 The only
precise fact we know is that it was produced by the same hand as certain other
plaquettes, some of which are signed with the name ‘Ulocrino’. Who Ulocrino
was, and exactly why, when and where he carved this plaquette, no one knows.
An ingenious suggestion concerning Ulocrino’s identity was made a hundred
years ago by Molinier in his comprehensive catalogue of plaquettes.2 Noting that
the name ‘Ulocrino’ has no meaning in Italian, he proposed that the word is a
hybrid word-play, formed from the Greek oulos and the Latin crinis, meaning
‘curly-haired’, the Italian for which is riccio, and the Latin crispus. Word-games of
this sort were much loved in the Renaissance, and it happens that there was a
celebrated sculptor active in Padua at the time, Andrea Briosco, who was called
‘Riccio’ because of his curly hair, and who himself signed a medal ‘Crispus’.
Therefore Molinier proposed to identify the mysterious Ulocrino with the well-
known sculptor Andrea Riccio. This sort of story ought to be true; but Molinier
himself noted that Ulocrino’s style is somewhat different from Riccio’s, and the
identification has been firmly rejected by J. Pope-Henessey on that ground.3
So we are left with nothing but a name and a small corpus of plaquettes. This
is enough to give us a context: for on grounds of style and subject-matter the
plaquettes can be safely placed in Padua or Venice and dated between, say, 1485
and 1530. And much was happening in Padua and Venice during this period
which helps explain the production of a plaquette devoted to Aristotle and

1
A rectangular bronze plaquette with brown patina. Dimensions approx. 72 × 55mm. Examples are
found in the Dahlem Museum, Berlin; the Victoria and Albert Museum, London; the Museo Correr,
Venice; and the National Gallery of Art, Washington, DC .
2
Molinier 1886, p. 176.
3
Pope-Henessey 1964, p. 77.

521
522 Note on the frontispiece

Alexander of Aphrodisias. The iconography is remarkable, because the subject-

matter of most Italian plaquettes from the period is religious, or else mythological
or allegorical. Such subjects as a portrait of St Jerome or a scene of Apollo and
Marsyas are easy to find – Ulocrino himself did both – but the portrait of an
historical person from antiquity is unusual.
However, at Padua and Venice during this period interest in Aristotle and in
his greatest Greek commentator, Alexander of Aphrodisias, was strong and
growing. The first Italian printed edition of Aristotle’s complete works (in Latin
translation) was produced at Venice in 1483. The Latin translation of Alexander’s
de Anima was first printed in Venice in 1495, and by 1520 it had been reprinted
five times.4 After years of lecturing on the subject, in 1516 the Paduan philosopher
Pietro Pomponazzi published his most famous work, de Immortalitate Animi, in
which he explicitly defended Alexander’s views. Pomponazzi’s book provoked
furious opposition, and he published two further works in its defence. As for
Alexander’s commentaries on Aristotle, they were gradually published, in Greek,
in Venice between 1513 and 1536.
The translator of Alexander’s de Anima, Girolamo Donati, was an early patron
of Andrea Riccio. Pope-Henessey suggests that this connection between the
publication of Alexander’s de Anima in 1495 and the circle of Paduan sculptors
makes it tempting to link our plaquette with this event.5 The thought is tempting
– and for the intellectual historian, tantalising. Since Alexander’s de Anima
contains the controversial doctrines on the soul which gave Paduan Alexandrism
its distinctive character, to connect the plaquette with the first printing of this
work would tie it with a major event in intellectual history. However that may be,
a more cautious and secure conclusion is that Ulocrino’s plaquette ‘Aristotle and
Alexander of Aphrodisias’ arose out of a growing fascination with Aristotle and
his commentator Alexander in Venice and Padua during the period 1485–1530.
Ulocrino’s charming plaquette can stand for us today as a symbol of this movement.
The prototype of the scene in ‘Aristotle and Alexander of Aphrodisias’ is the
magnificent hand-illuminated frontispiece to volume 1 of the Pierpont Morgan
Library’s copy of the 1483 Latin Aristotle, probably by Girolamo da Cremona.6

4
Edward Cranz, ‘Alexander of Aphrodisias’, in Kristeller 1960, pp. 184ff.
5
Pope-Henessey 1964, pp. 77–8.
6
A colour reproduction of that frontispiece is most easily found in Alexander 1977, plate 17, with
bibiliography, pp. 31–2. Pope-Henessey reports (1964, p. 77) that the closest analogies to this scene
occur in the illustrations to Zoppino’s Vite de philosophi moralissimi (a translation of Diogenes
Laertius), Venice 1521. Unfortunately, I have not yet been able to see these illustrations. Therefore
the conclusions drawn in the remainder of this note must be conditioned by a warning that the
question whether Ulocrino based his plaquette directly upon the Morgan Aristotle, or rather upon
one or more intermediates, remains open.
 Note on the frontispiece 523

Many features of the scene on the frontispiece, which shows Aristotle speaking to
Averroes, are echoed in Ulocrino’s plaquette. In both scenes, a seated Aristotle
speaks to a turbaned philosopher in the setting of a country landscape. In both, his
right arm is raised and extended while his left arm is held lower, roughly parallel to
his left leg. Both scenes include a large hill on the horizon between the two figures,
and both show buildings – human habitation – on the hill and also lower down to
its right. In both scenes a smaller hill in the middle ground curves downward from
the right edge of the image to disappear behind Aristotle’s left shoulder.
Some of the differences between the two images are the result of Ulocrino’s
successful efforts to adapt the horizontal format of the scene in the Morgan
frontispiece to the vertical format of his plaquette. The tree along the right-hand
side of the plaquette (which replaces a cave), the greater proximity of the two
figures, and Alexander’s standing posture (Averroes is seated) are all to be
explained in this way. One difference may reflect Girolamo’s greater subtlety (he
is undeniably the greater artist); his villas, barely visible in the background, are
replaced by Ulocrino’s more intrusive fortified buildings.
One change Ulocrino made in his prototype is puzzling. What is the meaning
of the hare in the right foreground? Fertility? Aristotle’s biological interests?
Other differences can be explained as reflecting Ulocrino’s changed
conception of the scene. Girolamo’s figures are old, and have the wisdom and
nobility of age. Ulocrino’s figures are nearly as subtly portrayed (bronze relief
does not permit the detail of painting), but they have a more rugged and youthful
aspect. This greater ruggedness is due to the broader, less delicate shapes of both
hats; both men’s shorter, less pointed beards, and the plain bronze colour of
the beards, replacing their snowy whiteness in Girolamo’s image. Girolamo’s
Aristotle is the wise old man who wrote Metaphysics 12; Ulocrino’s Aristotle is
the younger dialectician and scientist who wrote the Topics and studied fishes
off Lesbos.
Another significant change Ulocrino made is in the commentator’s activity.
Averroes is seated in front of an open book, with an inked quill in his right hand.
He is writing down the words or ideas which he hears from Aristotle, and
the book in front of him is his commentary. The scene recalls portrayals of
the inspiration of the Evangelists. By contrast, Ulocrino’s Alexander has no quill.
He holds a book and concentrates his gaze on the page while Aristotle speaks. It
is possible that Alexander here is holding his commentary, and that he has
returned to Aristotle in a worried state to ask whether in a certain passage he had
actually got it right. But by far the more likely explanation is that the book
Alexander holds is Aristotle’s text itself, and he is staring at a puzzling phrase
524 Note on the frontispiece

or sentence while Aristotle explains its meaning. Ulocrino’s image emphasises

the commentator’s puzzlement and the difficulty of Aristotle’s text. By contrast,
the focus on inspiration in Girolamo’s image makes the process of writing the
commentary seem effortless. Thus we may say that Girolamo’s scene portrays the
myth projected by a naive reader: ‘This commentary reads as if its author had
been under the direct inspiration of the Master.’ Ulocrino changes this to present
the ultimate fantasy of a commentator: ‘If only I could have the Author in front
of me and get him to explain this passage himself.’ The spirit behind Ulocrino’s
plaquette is more secular, and reveals a greater sensitivity to the experience of
writing a commentary.7

Bibliography

Alexander, J.J.G. Italian Renaissance Illuminations, New York 1977.

Bange, E.F. Die italienischen Bronzen der Renaissance und des Barock, zweiter Teil: Reliefs
und Plaketten, Berlin and Leipzig 1922.
Bode, W. von. Berliner Museen: Amtliche Berichte 29,1907–8.
Cott, National Gallery of Art, Washington (DC ). Renaissance Bronzes, Statuettes, Reliefs
and Plaquettes, Medals and Coins from the Kress Collection, Washington, DC 1951.
Jacobsen, E. ‘Plaketten im Museo Correr zu Venedig’, Repertorium für Kunstwissenschaft
16,1893, pp. 54–75.
Kristeller, P. Mediaeval and Renaissance Latin Translations and Commentaries I, 1960.
Lazari, Notizia della Raccolta Correr di Venezia, 1859.
Maclagan, E. Victoria and Albert Museum: Catalogue of Italian Plaquettes, London 1924.
Migeon, Les Arts 80,1908.
Molinier, E. Les Plaquettes: Catalogue raisonné, 2 vols, Paris 1886.
Planiscig, L. Andrea Riccio, Vienna 1927.
Panvini Rosati, F. Italienische Medaillen und Plaketten von der Frührenaissance bis zum
Ende des Barock, Cologne 1966.
Panvini Rosati, F. Medaglie e Placchette italiane dal Renascimento al XVIII secolo, Rome
1968.
Pope-Henessey, J. ‘The Italian Plaquette’, Proceedings of the British Academy, 1964,
pp. 63–76.
Pope-Henessey, J. Renaissance Bronzes from the Samuel H. Kress Collection, London 1965.
Ricci, The Gustav Dreyfus Collection: reliefs and plaquettes, Oxford 1931.

7
Further bibliography: Lazari 1859, p. 198, n. 1050 (n. 50); Jacobsen 1893, p. 62; Migeon 1908, p. 27;
Bange 1922, p. 58, n. 427, plate 40; Bode 1907, col. 252; Maclagan 1924, p. 27; Planiscig 1927, p. 316
(fig. 358); Ricci 1931, p. 125, n. 160; Cott 1951, p. 150; Pope-Henessey 1965, p. 72, n. 243, fig. 348;
Panvini Rosati 1966, p. 98, n. 12; Panvini Rosati 1968, pp. 68–9, n. 12, fig. 12.
 Select Bibliography of Secondary Literature

The following bibliography on the commentators was prepared for the 1990 edition of
this book by John Ellis, drawing on the work of Titos Christodoulou and many others. It
is reprinted here as guide to the literature up to 1990.
In 2004 John Sellars prepared a guide to work on the commentators from 1990 to
2004, conceived as a supplement to the bibliography here. It was published as ‘The
Aristotelian Commentators: A Bibliographical Guide’, in P. Adamson, H. Baltussen,
M. Stone (eds), Philosophy, Science, and Exegesis in Greek, Arabic, and Latin
Commentaries, Bulletin of the Institute of Classical Studies Suppl. 83.1, 2004, 239–68.
A bibliography of material since 2004 is available online at <www.ancient
commentators.org.uk/bibliographies.html>. This was initially prepared in 2013 by
Annie Hewitt, and later expanded and reorganised in 2015 by Malcolm Nicolson,
drawing on work by Katherine O’Reilly.

RE = Pauly-Wissowa (eds), Real-Encyclopädie der classischen Altertumswissenschaft,

Stuttgart.

General

Blumenthal, H.J., ‘Plotinus in Later Platonism’, in H.J. Blumenthal & R.A. Markus (eds),
Neoplatonism and Early Christian Thought. Essays in Honour of A.H. Armstrong,
London 1981, 212–22.
Blumenthal, H.J., ‘Neoplatonic elements in the de Anima commentaries’, Chapter 13
above, repr. with revisions from Phronesis 21, 1976, 64–87.
Brandis, C.A., ‘Über Aristoteles Rhetorik und die griechischen Ausleger derselben’,
Philologus 4, 1849, 34ff.
Brink, K.O., ‘Peripatos’, RE Suppl. 7, 1940, cols 938ff.
Burnett, C., Commentaries and Glosses to Aristotelian Logical Texts: the Syriac, Arabic
and European tradition, forthcoming.
Diels, H. (ed.), Commentaria in Aristotelem Graeca (CAG ), Berlin 1892–1909, see the
introductions.
Festugière, A.J., ‘Modes de composition des commentaires de Proclus’, Museum
Helveticum 20, 1963, 77–100; repr. in his Études de philosophie grecque, Paris 1971,
551–74.

525
526 Select Bibliography of Secondary Literature

Festugière, A.-J., ‘L’ordre de lecture des dialogues de Platon aux Ve-VIe siècles’, Museum
Helveticum 26, 1969, 281–96; repr. in his Études de philosophie grecque, Paris 1971,
535–50.
Glucker, J., Antiochus and the Late Academy, Hypomnemata 56, Göttingen 1978.
Hadot, I., ‘Les Introductions aux commentaires exégétiques chez les auteurs
néoplatoniciens et les auteurs chrétiens’, to be included in her French translation of
Simplicius in Cat., vol. 1; also in M. Tardieu, Les règles de l’interpretation, Paris 1987.
Hadot, I., ‘La division néoplatonicienne des écrits d’Aristote’, in J. Wiesner (ed.),
Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987,
249–85.
Hadot, P., ‘Les divisions des parties de la philosophie dans l’antiquité’, Museum
Helveticum 36, 1979, 201–23; unpublished English translation by Victor Caston.
Hadot, P., ‘The harmony of Plotinus and Aristotle according to Porphyry’, Chapter 6
above, translated from Atti del Convegno Internazionale sul tema Plotino e il
Neoplatonismo in Oriente e Occidente (Roma 5–9 ottobre 1970), Accademia dei
Lincei, Rome 1974, 31–47.
Lamberz, E., ‘Proklos und die Form des philosophischen Kommentars’, in J. Pépin &
H.D. Saffrey (eds), Proclus. Lecteur et interprète des Anciens, Paris 1987.
Lynch, J.P., Aristotle’s School. A Study of a Greek Educational Institution, Berkeley-Los
Angeles 1972.
Mercati, G., ‘Fra i commentatori greci di Aristotele’, in Mélanges d’archéologie et d’histoire
35, 1915, 191–219; repr. in Opere Minori III, Rome 1937, 458–67.
Moraux, P., ‘La critique d’authenticité chez les commentateurs grecs d’Aristote’, in
Mélanges Mansel 1, Ankara 1974, 265–88.
Moraux, P., D’Aristote à Bessarion: trois études sur l’histoire et la transmission de
l’aristotélisme grec, Laval, Quebec 1970.
Müller, C.W., ‘Die neuplatonischen Aristoteleskommentatoren über die Ursachen der
Pseudepigraphie’, Rheinisches Museum 112, 1969, 120–6.
Plezia, M., De Commentariis Isagogicis, Cracow 1947.
Praechter, K., Review of Commentaria in Aristotelem Graeca, Chapter 2 above, repr.
from ‘Die griechischen Aristoteleskommentatoren’, Byzantinische Zeitschrift 18, 1909,
516–38.
Praechter, K., ‘Richtungen und Schulen im Neuplatonismus’, Genethliakon für Carl
Robert, Berlin 1910, 105–55; repr. in Kleine Schriften, H. Dörrie (ed.), Collectanea 7,
Hildesheim-New York 1973, 165–216.
Richard, M., ‘Apo phônês’ (in French – unpublished English translation by Victor
Caston), Byzantion 20, 1950, 191–222.
Sorabji, R., ‘The ancient commentators on Aristotle’, Chapter 1 above.
Sorabji, R. (ed.), The Ancient Commentators on Aristotle, a series of English translations
with introductions and indexes, London-Ithaca N.Y. 1987–.
Tannery, P., ‘Sur la période finale de la philosophie grecque’, Revue philosophique 42,
1896, 226–87.
 Select Bibliography of Secondary Literature 527

Tarán, L., ‘Amicus Plato sed magis amica veritas. From Plato and Aristotle to Cervantes’,
Antike und Abendland 30, 1984, 93–124.
Todd, R.B., ‘Galenic medical ideas in the Greek Aristotelian Commentators’, Symbolae
Osloenses 52, 1977, 117–34.
Usener, H., Review of CAG, Göttingische gelehrte Anzeigen, 1892, 1001–22.
Westerink, L.G., ‘Academic practice about 500 A.D. in Alexandria’, unpublished paper
delivered to Dumbarton Oaks conference.
Westerink, L.G., ‘Ein astrologisches Kolleg aus dem Jahre 564’, Byzantinische Zeitschrift
64, 1971, 6–21; repr. in his Texts and Studies in Neoplatonism and Byzantine
Literature, Amsterdam 1980, 279–94.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962;
pp. ix–xxxii repr. with the revisions prepared for the new Budé edition in Chapter 14
above.

Individual commentators

Adrastus
Gottschalk, H.B, ‘The earliest Aristotelian commentators’, Chapter 3 above, repr. with
revisions from W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2,
Berlin 1987, 1079–174.
Gottschalk, H.B., Review of P. Moraux, Der Aristotelismus, vol. 2, in Liverpool Classical
Monthly 10.8, 1985, 122–8.
Immisch, O., ‘Attikoi exegetai’, Philologus 63, 1904, 31–40.
Kenny, A., The Aristotelian Ethics, Oxford 1978.
Mercken, H.P.F., ‘The Greek commentaries on Aristotle’s Ethics’, Chapter 18 above,
revised from ‘Introduction’, Corpus Latinum Commentariorum in Aristotelem
Graecorum 6.1.
Moraux, P., Der Aristotelismus bei den Griechen von Andronikos bis Alexander von
Aphrodisias, vol. 2, Der Aristotelismus im I. und II . Jh. n. Chr., Peripatoi 6, Berlin-
New York 1984, 294–332.

Alexander of Aphrodisias
Berti, E., ‘Alessandro di Afrodisia’, in Enciclopedia di Filosofia 1.2, Florence 1967, cols
171–2.
Gercke, A., ‘Alexander von Aphrodisias’, RE 1, cols 1453–5.
Lloyd, A.C., ‘Alexander of Aphrodisias’, in P. Edwards (ed.), Encyclopedia of Philosophy,
New York-London 1967, 73.
Merlan, P., ‘Alexander of Aphrodisias’, in C.S. Gillispie (ed.), Dictionary of Scientific
Biography, New York 1970, 117–20.
528 Select Bibliography of Secondary Literature

Mueller, Ian, ‘Stoic and Peripatetic logic’, Archiv für Geschichte der Philosophie 51, 1969,
173–87.
Nardi, B., ‘Alessandro di Afrodisia’, Enciclopedia Cattolica, vol. 1, Vatican City 1968,
778–80.
Sharples, R.W., ‘The school of Alexander?’, Chapter 4 above.
Sharples, R.W., ‘Alexander of Aphrodisias, scholasticism and innovation’, in W. Haase
(ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2, Berlin 1987, 1176–243.
Steinschneider, M., ‘Zu Alexander von Aphrodisias’, Magazin für die Wissenschaft des
Judentums 14, 1887, 190–5.

*
Bazán, B.C., ‘L’authenticité du “de Intellectu” attribué à Alexandre d’Aphrodise’, Revue
philosophique de Louvain 71, 1973, 468–87.
Blumenthal, H.J., ‘Alexander of Aphrodisias in the later Greek commentaries on
Aristotle’s de Anima’, in J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux
gewidmet, vol. 2, Berlin-New York 1987, 90–106.
Blumenthal, H.J., ‘Plotinus Ennead IV.3.20–21 and its sources – Alexander, Aristotle and
others’, Archiv für Geschichte der Philosophie 50, 1968, 254–61.
Coutant, V.C., Alexander of Aphrodisias, Commentary on Book IV of Aristotle’s
Meteorologica, diss., Columbia 1936; reviewed by B. Einarson in Classical Weekly 31,
1937, 33.
Cranz, F.E., ‘Alexander Aphrodisiensis’, in P.O. Kristeller & F.E. Cranz (eds), Catalogus
Translationum et Commentariorum, Medieval and Renaissance Latin Translations
and Commentaries, Washington D.C., vol. 1, 1960, 77–135; 2, 1971, 411–22.
Cranz, F.E., ‘The prefaces to the Greek editions and Latin translations of Alexander of
Aphrodisias, 1450–1575’, Proceedings of the American Philosophical Society 102, 1958,
510–46.
Dietrich, A., ‘Die arabische Version einer unbekannten Schrift des Alexander von
Aphrodisias über die Differentia specifica’, Nachrichten der Akademie der
Wissenschaften in Göttingen, 1964, 85–148.
Donini, P.L., Tre studi sull’Aristotelismo nel II secolo D. C., Historica, Politica,
Philosophica 7, Turin 1974.
Donini, P.L., ‘L’anima e gli elementi nel “de Anima” di Alessandro di Afrodisia’, Atti della
Accademia delle Scienze di Torino, Classe di Scienze Morali, Storiche e Filologiche
105, 1970–71, 61–107.
Donini, P.L., ‘Il de Anima di Alessandro di Afrodisia e Michele Efesio’, Rivista di Filologia
e di Instruzione Classica 96, 1968, 316–23.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Ess, J. van, ‘Über einige neue fragmente des Alexander von Aphrodisias in arabischer
Übersetzung’, Der Islam 42, 1966, 148–68.
 Select Bibliography of Secondary Literature 529

Finnegan, J., ‘Al-Farabi et le “Peri Nou” d’Alexandre d’Aphrodise’, Mélanges L. Massignon,

1957, 133–52.
Freudenthal, J., ‘Die durch Averroes erhaltenen Fragmente Alexanders zur Metaphysik
des Aristoteles untersucht und übersetzt’, Abhandlung der (k.) preussischen Akademie
der Wissenschaften, Berlin 1884.
Gätje, H., ‘Die Schrift des Alexander von Aphrodisias über das Sahen’ (with G. Losseren),
Studien zur Überlieferung der aristotelischen Psychologie im Islam, Annales
Universitatis Saraviensis 11, Heidelberg 1971, 140–72.
Gätje, H., ‘Die arabische Übersetzung der Schrift des Alexander von Afrodisias über die
Farbe’, Nachrichten der Academie der Wissenschaften in Göttingen, 1967, 341–82.
Gätje, H., ‘Zur arabischen Überlieferung des Alexander von Aphrodisias’, Zeitschrift der
deutschen morgenländischen Gesellschaft 116, 1966, 255–78.
Genequand, C., ‘L’objet de la métaphysique selon Alexandre d’Aphrodise’, Museum
Helveticum 36, 1979, 48–57.
Hager, F.P., ‘Die Aristotelesinterpretation des Alexander von Aphrodisias und die
Aristoteleskritik Plotins bezüglich der Lehre vom Geist’, Archiv für Geschichte der
Philosophie 46, 1964, 174–87.
Hamelin, O., La théorie de l’intellect d’après Aristote et ses commentateurs, posthumous
ed. E. Barbotin, Paris 1953.
Lloyd, A.C., ‘The significance of Alexander’s theory of universals’, in Proceedings of the
World Congress on Aristotle, vol. 1, Athens 1981, 155–9.
Madigan, A.S.J., ‘Alexander of Aphrodisias, the book of Ethical Problems’, in W. Haase
(ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2, Berlin 1987, 1260–79.
Merlan, P., ‘Zwei Untersuchungen zu Alexander von Aphrodisias’, Philologus 113, 1969,
85–91.
Mignucci, M., ‘Puzzles about identity, Aristotle and his Greek commentators’, in J.
Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 1, Berlin
1985, 57–97.
Moraux, P., Le commentaire d’Alexandre d’Aphrodise aux Secondes Analytiques d’Aristote,
Peripatoi 13, Berlin 1979. Review by E. Berti, Archiv für Geschichte der Philosophie 63,
1981, 84–6.
Moraux, P., Alexandre d’Aphrodise: exégète de la noétique d’Aristote, Liège-Paris 1942.
Pinès, S., ‘Omne quod movetur necesse est ab aliquo moveri, a refutation of Galen by
Alexander of Aphrodisias and the theory of motion’, Isis 52, 1961, 21–54.
Rescher, N. & Marmura, M., Alexander of Aphrodisias. The Refutation by Alexander of
Aphrodisias of Galen’s Treatise of the Theory of Motion, Islamabad (Islamic Research
Institute), 1965.
Rist, J.M., ‘On tracking Alexander of Aphrodisias’, Archiv für Geschichte der Philosophie
48, 1966, 82–90.
Ruland, H.-J., ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias
über das Wachstum (Quaestio I 5)’, Nachrichten der Akademie der Wissenschaften in
Göttingen, 1981, 51–74.
530 Select Bibliography of Secondary Literature

Ruland, H.-J., ‘Zwei arabische Fassungen der Abhandlung des Alexander von
Aphrodisias über die universalia (Quaestio I 11a)’, Nachrichten der Akademie der
Wissenschaften in Göttingen, 1979, 243–74.
Ruland, H.-J., ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias
über die Sinneswahrnehmung’, Nachrichten der Akademie der Wissenschaften in
Göttingen, 1978, 159–225.
Ruland, H.-J., Die arabische Fassungen von zwei Schriften des Alexander von
Aphrodisias: Über die Vorsehung und über das liberum arbitrium, diss., Saarbrücken
1976.
Schroeder, F.M., ‘Light and the active intellect in Alexander and Plotinus’, Hermes 112,
1984, 239–48.
Schroeder, F.M., ‘The potential or material intellect and the authorship of the de
Intellectu: a reply to B.C. Bazán’, Symbolae Osloenses 57, 1982, 115–25.
Schroeder, F.M., ‘The analogy of the active intellect to light in the de Anima of
Alexander of Aphrodisias’, Hermes 109, 1981, 215–25.
Sharples, R.W., ‘Ambiguity and opposition, Alexander of Aphrodisias on “Ethical
Problems” ’, Bulletin of the Institute of Classical Studies 32, 1985, 109–16.
Sharples, R.W., ‘The Unmoved Mover and the motion of the heavens in Alexander of
Aphrodisias’, Apeiron 17, 1983, 62–6.
Sharples, R.W., Alexander of Aphrodisias on Fate, London 1983.
Sorabji, R., Matter, Space and Motion (ch. 8), London-Ithaca, N.Y. 1988.
Strohmaier, G., ‘al-Iskandar al-Afrudisi’, in G. van Donzel et al. (eds), Encyclopedia of
Islam2, vol. 4, Leiden 1978, 129f.
Théry, G., Autour du décret de 1210: II. Alexandre d’Aphrodise: apercu sur l’influence de
sa noétique, Bibliothèque thomiste 7, Kain 1926.
Thillet, P., ‘Matérialisme et théorie de l’âme et de l’intellect chez Alexandre d’Aphrodise’,
Revue philosophique de la France et de l’étranger 171, 1981, 5–24.
Thurot, C., ‘Alexandre d’Aphrodisias, Commentaire sur le traité d’Aristote “De
Sensu et Sensibili” ’, Notices et extraits des MSS de la bibliothèque nationale 25,
1875, no. 2.
Todd, R.B., ‘The de Intellectu attributed to Alexander of Aphrodisias. Translation,
Introduction and Commentary’, in R.B. Todd & F.M. Shroeder (eds), The Greek
Aristotelian Commentators on the Active Intellect, Toronto, forthcoming.
Todd, R.B., Alexander of Aphrodisias on Stoic Physics, Leiden 1976.
Todd, R.B., ‘Lexicographical notes on Alexander of Aphrodisias’ philosophical
terminology’, Glotta 52, 1974, 207–15.
Tweedale, M.M., ‘Alexander of Aphrodisias’ views on universals’, Phronesis 29, 1984,
279–303.
Verbeke, G, ‘Aristotle’s Metaphysics as viewed by the ancient Greek commentators’, in
D.J. O’Meara (ed.), Studies in Aristotle, Washington, D.C. 1981, 107–27.
Wallies, M., ‘Die griechischen Ausleger der aristotelesichen Topik’, Wissenschaftliche
Beilage zum Programm des Sophien–Gymnasiums zu Berlin 65, 1891, 1–27.
 Select Bibliography of Secondary Literature 531

Ammonius son of Hermeias

Anton, J.P., ‘Ancient interpretations of Aristotle’s doctrine of homonyma’, Journal of the
History of Philosophy 7, 1969, 1–18.
Isaak, I., Le Peri hermeneias en Occident de Boèce à Saint Thomas. Histoire littéraire d’un
traité d’Aristote, Bibliothèque thomiste 29, Paris 1953; includes the first 10 chapters
of William of Moerbeke’s translation, pp. 160–9.
Merlan, P., ‘Ammonius Hermiae, Zacharias Scholasticus and Boethius’, Greek, Roman
and Byzantine Studies 9, 1968, 143–203.
Noica, C., Ammonius, Stephanus. Comentarii la Tratatul Despre Interpretare al lui Aristotel,
însotite de Textul Comentat, Traducere, Curînt înainte, note si Comentariu, Bucarest 1971.
Noica, C., Porfir, Dexip, Ammonius. Comentarii la Categoriile lui Aristotel, însotite de
Textul Comentat Traducere, Curînt înainte si Note, Bucarest 1968.
Saffrey, H.D., ‘Le chrétien Jean Philopon et la survivance de l’école d’Alexandrie au VIe
siècle’, Revue des études grecques 67, 1954, 396–410.
Sheppard, Anne, ‘Proclus’ philosophical method of exegesis, the use of Aristotle and the
Stoics in the Commentary on the Cratylus’, in J. Pépin & P. Saffrey (eds), Proclus.
Lecteur et interprète des Anciens, Paris 1987, 137–51.
Sorabji, R., ‘Infinite power impressed, the transformation of Aristotle’s physics and
theology’, Chapter 9 above (fuller version in Matter, Space and Motion, London-
Ithaca N.Y. 1988).
Sorabji, R., Time, Creation and the Continuum, London-Ithaca N.Y. 1983, ch. 16.
Verrycken, K., ‘The metaphysics of Ammonius son of Hermeias’, Chapter 10 above.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962,
pp. ix–xxxii repr. with the revisions prepared for the new Budé edition as Chapter 14
above.

Andronicus
Düring, I., Aristotle in the Ancient Biographical Tradition, Göteborg 1957.
Düring, I., ‘Notes on the history of the transmission of Aristotle’s writings’, Göteborg
Högskulas Arsskrift 56, 1950, 3, 35–70.
Gottschalk, H.B., ‘Aristotelian philosophy in the Roman world from the time of Cicero
to the end of the second century A.D.’, in W. Haase (ed.), Aufstieg und Niedergang der
römischen Welt, II , 36.2, Berlin 1987, 1079–174; part. repr. with revisions as
Chapter 3 above.
Littig, Fr., Andronikos von Rhodos (3 vols), Erlangen 1, 1890; 2, 1894; 3, 1985.
Moraux, P., Der Aristotelismus bei den Griechen von Andronikos bis Alexander von
Aphrodisias, vol. 1, Die Renaissance des Aristotelismus im I. Jh. v. Chr., Section 2:
‘Andronikos von Rhodos’, Peripatoi 5, Berlin 1973, 97–141.
Plezia, M., De Andronici Rhodii Studiis Aristotelicis, Polska Academia Uniectnosci,
Archiwum Filologiczne, no. 20, Cracow 1946.
532 Select Bibliography of Secondary Literature

Shute, R., On the History of the Process by which the Aristotelian Writings Arrived at their
Present Form, Oxford 1888.

Asclepius
Kremer, K., Der Metaphysikbegriff in den Aristoteles-Kommentaren der Ammonius-
Schule, Beiträge zur Geschichte der Philosophie und Theologie des Mittelalters 39, 1,
Münster 1961.
Madigan, A., ‘Syrianus and Asclepius on forms and intermediates in Plato and Aristotle’,
Journal of the History of Philosophy 24, 1986, 49–71.
Tarán, L., ‘Asclepius of Tralles’ commentary to Nicomachus’ Introduction to Arithmetic’,
Transactions of the American Philosophical Society, n.s. 59, 1969, 8.
Verrycken, K., ‘The metaphysics of Ammonius son of Hermeias’, Chapter 10 above.
Westerink, L.G., ‘Deux commentaires sur Nicomaque, Asclépius et Jean Philopon’, Revue
des études grecques 77, 1964, 526–35; repr. in his Texts and Studies in Neoplatonism
and Byzantine Literature, Amsterdam 1980, 101–10.

Aspasius
Gercke, A., ‘Aspasios’, RE 2, 1896, cols 1722–3.
Uberweg, F., ‘Aspasius’, Grundriss der Geschichte der Philosophie, vol. 1, 562, 177.
Zahlfleisch, J., ‘Aspasio’, Enciclopedia di Filosofia, vol. 1, col. 507.

*
Becchi, F., ‘Aspasio e i peripatetici posteriori, la formula definitiva della passione’,
Prometheus 9, 1983, 83–104.
Gauthier, R.-A., ‘Introduction’, in R.A. Gauthier & J.Y. Jolif (eds), L’Éthique à Nicomaque.
Introduction, traduction et commentaire, 2nd ed., vol. 1, part 1, Paris 1970, 100–5,
121.
Gottschalk, H.B., ‘Aristotelian philosophy in the Roman world from the time of Cicero
to the end of the second century A.D.’, in W. Haase (ed.), Aufstieg und Niedergang der
römischen Welt, II , 36.2, Berlin 1987, 1079–174, part repr. with revisions as Chapter 3
above.
Hanquet, R., Aspasius, sa vie, son oeuvre, sa pensée, diss., Louvain 1945.
Kenny, A., The Aristotelian Ethics, Oxford 1978.
Mercken, H.P.F., ‘The Greek commentaries on Aristotle’s Ethics’, Chapter 18 above, repr.
with revisions from Introduction to Corpus Latinum Commentariorum in
Aristotelem Graecorum 6.1.
Moraux, P., Der Aristotelismus bei den Griechen von Andronikos bis Alexander von
Aphrodisias, vol. 2, Der Aristotelismus im I. und II. Jh. n. Chr., Peripatoi 6, Berlin-New
York 1984, esp. 226–93.
 Select Bibliography of Secondary Literature 533

Rose, V., ‘Über die griechischen Kommentare zur Ethik des Aristoteles’, Hermes 5, 1871,
61–113.
Schleiermacher, F., ‘Über die griechischen Scholien zur Nicomachischen Ethik des
Aristoteles’, Sammtliche Werke, section 3, vol. 2, Berlin 1838, 309–26.

Boethius
Bieler, L. (ed.), ‘Praefatio’ and ‘Bibliographia’, in Boethii Philosophiae Consolatio, ed. 2a,
Turnholt 1989.
Cappuyns, M., ‘Boèce’, DAGE 9, 1939.
Chadwick, H., Boethius. The Consolations of Music, Logic, Theology, and Philosophy,
Oxford 1981.
Fuhrmann, M. & Gruber, J. (eds), Boethius, Wege der Forschung 483, Darmstadt 1984.
Liebeschütz, H., ‘Boethius and the legacy of antiquity’, in A.H. Armstrong (ed.), The
Cambridge History of Later Greek and Early Medieval Philosophy, Cambridge 1967,
538–64.
Minio-Paluello, L., ‘Boethius’, Dictionary of Scientific Biography 2, 1970, 228–36.
Minio-Paluello, L., ‘Boethius’, article in Encyclopaedia Britannica, 1968.
Obertello, L., Severino Boezio, 2 vols, Genoa 1974.
Shiel, J., ‘Boethius’ commentaries on Aristotle’, Chapter 15 above, revised from Medieval
and Renaissance Studies 4, 1958, 217–44.

*
Barnes, J., ‘Boethius and the study of logic’, in M. Gibson (ed.), Boethius. His Life,
Thought and Influence, Oxford 1981, 73–89.
Bidez, J., ‘Boèce et Porphyre’, Comptes rendus de l’académie des inscriptions et des belles
lettres, 1922, 345–50.
Courcelle, P., Late Latin Writers and their Greek Sources, Cambridge, Mass. 1969;
translated from Les lettres grecques en Occident, de Macrobe à Cassiodore, 2nd ed.,
Paris 1948.
Dürr, K., The Propositional Logic of Boethius, Studies in Logic and the Foundations of
Mathematics, Amsterdam 1951.
Ebbesen, S., ‘Boethius as an Aristotelian commentator’, Chapter 16 above; repr. from
J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2,
Berlin-New York 1987, 286–311.
Ebbesen, S., ‘Analyzing syllogisms, or Anonymus Aurelianensis III – the (presumably)
earliest extant Latin commentary on the Prior Analytics, and its Greek model’,
Cahiers de l’Institut du moyen-âge grec et latin 37, Copenhagen 1981, 1–20.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
534 Select Bibliography of Secondary Literature

Ebbesen, S., ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and Boethius:

ancient scholasticism and twelfth-century Western Europe. With an edition of the
Tractatus de Paralogismis’, Cahiers de l’Institut du moyen-âge grec et latin 16,
Copenhagen 1976, 1–128.
Gersh, S., Middle Platonism and Neoplatonism. The Latin Tradition, 2 vols, Notre Dame
1986, ch. 9.
Gibson, M. (ed), Boethius. His Life, Thought and Influence, Oxford 1981.
Hadot, P., ‘Un fragment du commentaire perdu de Boèce sur les Catégories d’Aristote
dans le codex Bernensis 363’, Archives d’histoire doctrinale et littéraire du moyen-âge
26, 1959, 11–27.
Isaak, I., Le Peri hermeneias en Occident de Boèce à Saint Thomas. Histoire littéraire d’un
traité d’Aristote, Bibliothèque thomiste 29, Paris 1953; includes the first 10 chapters
of William of Moerbeke’s translation, pp. 160–9.
Kretzmann, N., ‘Boethius and the truth about tomorrow’s sea battle’, in L.M. de Rijk &
H.A.G. Braakhuis (eds), Logos and Pragma. Essays in honour of Prof. G. Nuchelmans,
Artistarium suppl. 3, Nijmegen 1987, 63–97.
Kretzmann, N., ‘Nos ipsi principia sumus: Boethius and the basis of contingency’, in
T. Rudavsky (ed.), Divine Omniscience and Omnipotence in Medieval Philosophy:
Islamic, Jewish and Christian perspectives, Synthese Historical Library 25, Dordrecht
1985.
Magee, J., Boethius on Signification and Mind, forthcoming, Leiden 1989.
Minio-Paluello, L., Opuscula, the Latin Aristotle, Amsterdam 1972.
Minio-Paluello, L., ‘Praefationes’ in Aristoteles Latinus 1.1–5 and (B. Dod, ed.) 1.6–7,
Paris 1966; includes Boethius’ translations of Isagoge, Cat., Int., An. Pr., Top., SE.
Minio-Paluello, L., ‘Les traductions et les commentaires aristotéliciennes de Boèce’,
Studia Patristica II , Texte und Untersuchungen zur Geschichte der altchristlichen
Literatur 64, 1957, 358–65.
Minio-Paluello, L., ‘A Latin commentary (?translated by Boethius) on the Prior
Analytics and its Greek sources’, Journal of Hellenic Studies 77, 1957, part 1, 93–102.
Nuchelmans, G., Theories of the Proposition. Ancient and Medieval Conceptions of the
Bearers of Truth and Falsity, Amsterdam 1973.
Obertello, L. (ed.), Atti di Congresso Internazionale di Studi Boeziani, Roma 1981.
Prior, A.N., ‘The logic of negative terms in Boethius’, Franciscan Studies 13, 1953, 1–16.
Rijk, L.M. de, ‘On Boethius’ notion of being: a chapter of Boethian semantics’, in N.
Kretzmann (ed.), Meaning and Inference in Medieval Philosophy. Studies in Memory
of Jan Pinborg, Synthese Historical Library 32, Dordrecht 1988, 1–29.
Rijk, L.M. de, ‘Boèce logicien et philosophe, ses positions et sa métaphysique de l’être’, in
L. Obertello (ed.), Congresso Internazionale di Studi Boeziani – Atti, Roma 1981,
141–56.
Rijk, L.M. de, Logica Modernorum, Assen, vol. 1, 1962; vol. 2, 1967.
Rijk, L.M. de, ‘On the chronology of Boethius’ works on logic’, Vivarium 2, 1964, 1–49,
122–62.
 Select Bibliography of Secondary Literature 535

Shiel, J., ‘Religio Grammatici’, Matrix 10, 1990.

Shiel, J., ‘The Greek copy of Porphyrios’ Isagoge used by Boethius’, in J. Wiesner (ed.),
Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987,
312–40.
Shiel, J., ‘A recent discovery, Boethius’ notes on the Prior Analytics’, Vivarium 20, 1982,
128–41.
Shiel, J., ‘Boethius and Eudemus’, Vivarium 12, 1974, 14–17.
Shiel, J., ‘Boethius’, article in Encyclopaedia Britannica, 1974.
Shiel, J., ‘Boethius and Andronicus of Rhodes’, Vigiliae Christianae 11, 1957, 179–85.
Solmsen, F., ‘Boethius and the history of the Organon’, American Journal of Philology 31,
1944, 69–74.
Stump, E., ‘Boethius’ works on the Topics’, Vivarium 12, 1974, 77–93.

Damascius
Chaignet, A., Damascius le Diadoque, Paris 1898.
Heitz, E., Der Philosoph Damascius, Freiburg i. Br. 1884.
Kroll, W., ‘Damaskios’, RE 4, 2, 1901, cols 2039–42.
Strömberg, R., ‘Damascius, his personality and importance’, Eranos 44, 1946, 175–92.
Westerink, L. & Combès, J., ‘Introduction’ in Damascius, Traité des premiers principes,
vol. 1, Budé edition, Paris 1986, xxxvi–xli.
Westerink, L.G., The Greek Commentaries on Plato’s Phaedo, vol. 2, Damascius,
Amsterdam 1977.
Westerink, L.G., ‘Damascius, commentateur de Platon’, in Le néoplatonisme, Actes du
Colloque de Royaumont 9–13 Juin, 1969, Paris 1971, 253–60.

David
Aversatjan, S., ‘David l’Invincible et sa doctrine philosophique’, Revue des études
arméniennes 15, 1981, 33–43.
Benakis, L., ‘David der Armenier in der Werken der byzantinischen Kommentatoren des
Aristoteles’, in David the Invincible. The Great Philosopher of Ancient Armenia,
Yerevan 1983, 558–70. Also in Armenian, in Patma – Banasirakan Handen (Yerevan)
1, 1981, 46–55.
Dörrie, H., ‘David’, RE 4, 1901, cols 2232–3.
Kendall, B. & Thomson, R.W., Definitions and Divisions of Philosophy by David the
Invincible Philosopher, Armenian Texts and Studies 5, California 1983.
Khostikian, M., David der Philosoph, Bern 1907.
Neumann, C., ‘Mémoire sur la vie et des ouvrages de David, philosophe arménien du Ve
siècle de notre ère’, Nouveau journal asiatique 3, 1829, 49–86, 97–157; repr. Paris
1921.
536 Select Bibliography of Secondary Literature

Praechter, K., David Prolegomena (CAG 18, 2.5. 34,6ff.), Hermes 1911, 316–17.
Praechter, K., review of Prolegomena and in Isagogem (CAG ), Göttingische gelehrte
Anzeigen 170, Göttingen 1908.
Rose, V., Leben des Heiligen David von Salonika, griechisch nach der einzigen bisher
aufgefundenen Handschrift herausgeg. Mit Beziehung auf David Invictus Philosophus,
Berlin 1887.
Sanjian, A.K. (ed.), David Anhaghtc, the Invincible Philosopher, Atlanta, Georgia
1986.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962; new
edition in preparation; pp. ix-xxxii repr. with the revisions prepared for the new
Budé edition in Chapter 14 above.

Dexippus
Anton, J.P., ‘Ancient interpretations of Aristotle’s doctrine of homonyma’, Journal of the
History of Philosophy 7, 1969, 1–18.
Boeft, J. den, ‘Desippo’, Enciclopedia di Filosofia, vol. 2, cols 391–2.
Busse, A., ‘Der Historiker und der Philosoph Dexippus’, Hermes 23, 1888, 402–9.
Kroll, W., ‘Dexippos’, in RE 5, 1913, cols 293–4.

*
Aubenque, P., ‘Plotin et Dexippe, exégètes des Catégories d’Aristote’, in C. Rutten &
A. Motte (eds), Aristotelica. Mélanges offerts à Marcel de Corte, Brussels-Liège 1985,
7–40.
Hadot, P., ‘The harmony of Plotinus and Aristotle according to Porphyry’, Chapter 6
above, translated from Atti del Convegno Internazionale sul tema Plotino e il
Neoplatonismo in Oriente e Occidente (Roma 5–9 ottobre 1970), Accademia
Nazionale dei Lincei, Rome 1974, 31–47.
Henry, P., ‘Trois apories orales de Plotin sur les Catégories’, Zetesis, Mélanges de Strycker,
Anvers 1973, 234–65.
Noica, C., Porfir, Dexip, Ammonius. Comentarii la Categoriile lui Aristotel, însotite de
Textul Comentat Traducere, Curînt înainte si Note, Bucarest 1968.
Usener, H., Review of L. Spengel (CAG edition of Dexippus), Litterar. Centralblatt, 1860,
124–5.

Elias
Anton, J.P., ‘Ancient interpretations of Aristotle’s doctrine of homonyma’, Journal of the
History of Philosophy 7, 1969, 1–18.
Blumenthal, H.J, ‘Pseudo-Elias and the Isagoge commentaries again’, Rheinisches
Museum 124, 1981, 188–92.
 Select Bibliography of Secondary Literature 537

Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of

Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinorum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Marcovich, M., Pseudo-Elias on Heraclitus, American Journal of Philology 96, 1975,
31–4.
Westerink, L.G. (ed.), ‘Elias on the Prior Analytics’, Mnemosyne 14, 1961, 126–39; repr. in
his Texts and Studies in Neoplatonism and Byzantine Literature, Amsterdam 1980,
59–72.
Westerink, L.G., ‘Elias und Plotin’, Byzantinische Zeitschrift 57, 1964, 26–32; repr. in
his Texts and Studies in Neoplatonism and Byzantine Literature, Amsterdam 1980,
93–9.
Westerink, L.G., Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,
Amsterdam 1967.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962; new
edition in preparation; pp. x–xxxii repr. with the revisions prepared for the new Budé
edition in Chapter 14 above.

Eustratius
Dräseke, J., ‘Zu Eustratius von Nikaea’, Byzantinische Zeitschrift 5, 1896, 319–36.
Martini, E., ‘Eustratius, Metropolit von Nicaea’, in RE 6, 1, 1907, col. 1490–1.

*
Gauthier, R.-A., ‘Introduction’, in R.A. Gauthier & J.Y. Jolif, L’Éthique à Nicomaque.
Introduction, traduction et commentaire, vol. 1, part 1, 2nd ed., Paris 1970, 100–5,
121.
Giocarinis, K., ‘Eustratius of Nicaea’s defence of the doctrine of Ideas’, Franciscan Studies
24, 1964, 159–204.
Ioannou, P., ‘Die Definition des Seins bei Eustratius von Nicaea. Die Universalienlehre
in der byzantinischen Theologie im XI Jahrhundert’, Byzantinische Zeitschrift 47,
1954, 258–368.
Lloyd, A.C., ‘The Aristotelianism of Eustratius of Nicaea’, in J. Wiesner (ed.), Aristoteles:
Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987, 341–51.
Mercken, H.P.F., ‘The Greek commentaries on Aristotle’s Ethics’, Chapter 18 above, repr.
with revisions from ‘Introduction’, Corpus Latinum Commentariorum in
Aristotelem Graecorum 6.1.
Rose, V., ‘Über die griechischen Kommentare zur Ethik des Aristoteles’, Hermes 5, 1871,
61–113.
Rossi, P., ‘Tracce della versione latina di un commento greco ai Secondi Analitici nel
Commentarius in Posteriorium Analyticorum di Grossatesta’, Rivista di Filosofia
Neoscolastica, 1978, 433–9.
538 Select Bibliography of Secondary Literature

Iamblichus
Bidez, J., ‘Le philosophe Jamblique et son école’, Revue des études grecques 323, 1919,
29–40.
Dalsgaard-Larsen, B., Jamblique de Chalcis: exégète et philosophe, 2 vols, Århus 1972.
Dillon, J., ‘Iamblichus of Chalcis (c. 240–325 A.D.)’, in W. Haase (ed.), Aufstieg und
Niedergang der römischen Welt, II , 36.2, Berlin 1987, 862–909.
Kroll, W., ‘Iamblichus’, RE 10, 1, 1914, cols 649–51.

*
Blumenthal, H.J., ‘Did Iamblichus write a commentary on the de Anima?’, Hermes 102,
1974, 540–56.
Dalsgaard-Larsen, B., ‘La place de Jamblique dans la philosophie antique tardive’, in
H. Dörrie (ed.), De Jamblique à Proclus, Entretiens sur l’antiquité classique 21,
Vandoeuvres-Genève 1975, 1–34.
Hadot, P., ‘The harmony of Plotinus and Aristotle according to Porphyry’, Chapter 6
above, translated from Atti del Convegno Internazionale sul tema Plotino e il
Neoplatonismo in Oriente e Occidente (Roma 5–9 ottobre 1970), Accademia dei
Lincei, Rome 1974, 31–47.
Hoffmann, P., ‘Jamblique exégète du pythagoricien Archytas: trois originalités d’une
doctrine du temps’, Les études philosophiques 1980, 307–23.
Whittaker, J., ‘The historical background of Proclus’ doctrine of the “authupostata” ’, in
H. Dörrie (ed.), De Jamblique à Proclus, Entretiens sur l’antiquité classique 21,
Vandoeuvres-Genève 1975, 193–230.

Michael of Ephesus
Browning, R., ‘An unpublished funeral oration on Anna Comnena’, Chapter 17 above,
repr. with revisions from Proceedings of the Cambridge Philological Society 188, 1962,
1–12.
Donini, P.L., ‘Il de Anima di Alessandro di Afrodisia e Michele Efesio’, Rivista di Filologia
e di Instruzione Classica 96, 1968, 316–23.
Ebbesen, S., ‘Philoponus, “Alexander” and the origins of medieval logic’, Chapter 19 above.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Ebbesen, S., ‘Boethius, Jacobus Veneticus, Michael Ephesius and “Alexander” ’, Cahiers de
l’Institut du moyen-âge grec et latin 34, Copenhagen 1979, 5–13.
Giordano, D., ‘Michele di Efeso’, Enciclopedia di Filosofia vol. 4, col. 624.
Mercken, H.P.F., ‘The Greek commentaries on Aristotle’s Ethics’, Chapter 18 above, repr.
with revisions from ‘Introduction’, Corpus Latinum Commentariorum in
Aristotelem Graecorum 6.1.
 Select Bibliography of Secondary Literature 539

Praechter, K., ‘Michael von Ephesos und Psellos’, Byzantinische Zeitschrift 31, 1931, 1–12.
Praechter, K., review of Hayduck 1904, CAG 22.2 (Michael Ephesii In Libros De Partibus
Animalium Commentaria), Göttingische gelehrte Anzeigen 168, 1906, 861–907.
Preus, A., ‘Michael of Ephesus on Aristotle IA and MA’, in Proceedings of the 1978 World
Congress on Aristotle, vol. 2, Athens 1981, 21–30.
Preus, A., Aristotle and Michael of Ephesus on the Movement and Progression of Animals,
Hildesheim-New York 1981.

Olympiodorus
Anton, J.P., ‘Ancient interpretations of Aristotle’s doctrine of homonyma’, Journal of the
History of Philosophy 7, 1969, 1–18.
Beutler, R., ‘Olympiodorus (13)’, RE 18, 2, 1949, cols 207–27.
Blumenthal, H.J., review of Westerink, The Greek Commentaries on Plato’s Phaedo,
Mnemosyne 33, 1980, 411–14.
Norvin, W., Olympiodoros fra Alexandria og hans Commentar til Platons Phaidon,
Copenhagen 1915.
Praechter, K., ‘Olympiodor und Kedren’, Byzantinische Zeitschrift 12, 1903, 224–30.
Vancourt, R., Les derniers commentateurs alexandrins d’Aristote: l’école d’Olympiodore,
Étienne d’Alexandrie, Lille 1941.
Westerink, L.G., The Greek Commentaries on Plato’s Phaedo, vol. 1, Olympiodorus,
Amsterdam 1973.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962; new
edition in preparation; pp. x–xxxii repr. with the revisions prepared for the new Budé
edition in Chapter 14 above.

Philoponus
Böhm, W., Johannes Philoponos. Grammatikos von Alexandrien, Munich-Paderborn-
Vienna 1967.
Gudeman, A. & Kroll, W., ‘Ioannes (Nr. 21, Ioannes Philoponus)’, in RE 9, 2, 1916, cols
1764–95.
Saffrey, H.D., ‘Le chrétien Jean Philopon et la survivance de l’école d’Alexandrie au VIe
siècle’, Revue des études grecques 67, 1954, 396–410.
Sorabji, R., ‘John Philoponus’, in R. Sorabji (ed.), Philoponus and the Rejection of
Aristotelian Science, London-Ithaca N.Y. 1987, 1–40.
Verrycken, K., ‘The development of Philoponus’ thought and its chronology’, Chapter 11
above.

*
Blumenthal, H.J., ‘Philoponus – John: Alexandrian Platonist’, Hermes 114, 1986, 314–35.
540 Select Bibliography of Secondary Literature

Blumenthal, H.J., ‘Body and soul in Philoponus’, Monist 69, 1986, 370–82.
Blumenthal, H.J., ‘John Philoponus and Stephanus of Alexandria: two Neoplatonist
Christian commentators on Aristotle?’, in D.J. O’Meara (ed.), Neoplatonism and
Christian Thought, Norfolk 1982, 54–66; 244–6.
Davidson, H.A., Proofs for Eternity, Creation and the Existence of God in Medieval
Islamic and Jewish Philosophy, New York-Oxford 1987.
Davidson, H.A., ‘John Philoponus as a source of Medieval Islamic and Jewish proofs of
creation’, Journal of the American Oriental Society 89, 1969, 357–91.
Ebbesen, S., ‘Philoponus, “Alexander” and the origins of medieval logic’, Chapter 19
above.
Endress, G., The Works of Yahyâ ibn ‘Adi, Wiesbaden 1977.
Évrard, É., ‘Jean Philopon, la ténèbre originelle et la création du monde’, in C. Rutten &
A. Motte (eds), Aristotelica, Mélanges offerts à Marcel de Corte, Brussels-Liège 1985,
177–88.
Évrard, É., ‘Jean Philopon, son commentaire sur Nicomaque et ses rapports avec
Ammonius’, Revue des études grecques 78, 1965, 592–8.
Évrard, É., L’école d’Olympiodore et la composition du commentaire à la Physique de Jean
Philopon, diss., Liège 1957.
Évrard, É., ‘Les convictions religieuses de Jean Philopon et la date de son commentaire
aux Météorologiques’, Bulletin de l’academie royale de Belgique, Classe des lettres et
sciences politiques et morales 5 ser. 39, 1953, 299–357.
Grant, E., ‘Aristotle, Philoponus, Avempace and Galileo’s Pisan dynamics’, Centaurus 11,
1965, 79–95.
Kraemer, J.L., ‘A lost passage from Philoponus’ Contra Aristotelem in Arabic translation
(Simpl. de Caelo, I, 3, 270b5–11)’, Journal of the American Oriental Society 85, 1965,
318–27.
Mahdi, M., ‘Alfarabi against Philoponus’, Journal of Near Eastern Studies 26, 1967,
233–60.
McGuire, J.E., ‘Philoponus on Physics ii 1, phusis, hormê emphutos and the motion of
simple bodies’, Ancient Philosophy 5, 1985, 241–67.
Moody, E.A., ‘Galileo and Avempace, the dynamics of the leaning tower experiment’,
Journal of the History of Ideas 13, 1951, 163–93; 375–422.
Pinès, S., ‘Un précurseur bagdadien de la théorie de l’impetus’, Isis 44, 1953, 246–51; repr.
in Pinès, S., Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2,
Leiden-Jerusalem 1986, 418–22.
Pinès, S., ‘An Arabic summary of a lost work of John Philoponus’, Israel Oriental Studies
2, 1972, 320–52; repr. in S. Pinès, Studies in Arabic Versions of Greek Texts and in
Medieval Science, vol. 2, Leiden-Jerusalem 1986, 294–326.
Pinès, S., ‘La dynamique d’Ibn Bajja’, in Mélanges Alexandre Koyré 1, Paris, 1964,
442–68.
Pinès, S., ‘Études sur Awhad al-Zamân Abu’ l-Barakât al-Baghdâdî’, Revue des études
juives n.s. 3, 1938, 3–64 and n.s. 4, 1938, 1–33.
 Select Bibliography of Secondary Literature 541

Pinès, S., ‘Les précurseurs musulmans de la théorie de l’impetus’, Archeion 21, 1938,
298–306.
Riet, A. van, ‘Fragments de l’original grec du de Intellectu de Philopon dans une
compilation de Sophonias’, Revue philosophique de Louvain 63, 1965, 5–40.
Rossi, P., ‘Tracce della versione latina di un commento greco ai Secondi Analitici nel
Commentarius in Posteriorium Analyticorum di Grossatesta’, Rivista di Filosofia
Neoscolastica 1978, 433–9.
Sambursky, S. & Pinès, S., The Concept of Time in Late Neoplatonism. Texts with
Translation, Introduction and Notes, Jerusalem 1981; repr. Leiden 1987.
Sambursky, S., The Concept of Place in Late Neoplatonism. Texts with Translation,
Introduction and Notes, Jerusalem 1982.
Sambursky, S., ‘Philoponus, John’, in P. Edwards (ed.), The Encyclopaedia of Philosophy
vol. 6, New York 1967, 156–7.
Sambursky, S., The Physical World of Late Antiquity, London 1962.
Schmitt, C.B., Gianfrancesco Pico della Mirandola (1469–1533) and his Critique of
Aristotle, The Hague 1967.
Sorabji, R., Matter, Space and Motion (chs 2, 14, 15), London-Ithaca N.Y. 1988.
Sorabji., R. (ed.), Philoponus and the Rejection of Aristotelian Science, London-Ithaca
N.Y. 1987.
Sorabji, R., Time, Creation and the Continuum, London-Ithaca N.Y. 1983 (ch. 14).
Steinschneider, M., ‘Johannes Philoponus bei den Arabern’, in ‘Al-Farabi-Alpharabius,
des arabischen Philosophen Leben und Schriften’, Mémoires de l’academie impériale
des sciences de St. Petersbourg, series 7, 13, 4, 1869, 152–76; 220–4; 250–2.
Tannery, P., ‘Philoponus’, Philosophie ancienne, Mémoires scientifiques publiées par
J.L. Heiberg 7, Paris 1925, 318–20.
Todd, R.B., ‘Philosophy and medicine in John Philoponus’ commentary on Aristotle’s de
Anima’, Dumbarton Oaks Papers, 38, 1984, 103–20.
Todd, R.B., ‘Some concepts in physical theory in John Philoponus’ Aristotelian
commentaries’, Archiv für Begriffsgeschichte 24, 1980, 151–70.
Verbeke, G., ‘Levels of human thinking in Philoponus’, in C. Laga, J.A. Munitz,
L. van Rompay (eds), After Chalcedon. Studies in Theology and Church History
offered to Professor Albert van Roey for his seventieth birthday, Louvain 1986,
451–70.
Verbeke, G., ‘Some later Neoplatonic views on divine creation and the eternity of the
world’, in D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Norfolk 1982,
45–53 and 241–4.
Verbeke, G., Jean Philopon. Commentaire sur le de Anima d’Aristote, traduction de
Guillaume de Moerbeke, Corpus Latinum Commentariorum in Aristotelem
Graecorum 3, Louvain-Paris 1966.
Westerink, L.G., ‘Deux commentaires sur Nicomaque, Asclépius et Jean Philopon’, Revue
des études grecques 77, 1964, 526–35; repr. in his Texts and Studies in Neoplatonism
and Byzantine Literature, Amsterdam 1980, 101–10.
542 Select Bibliography of Secondary Literature

Wieland, W. ‘Zur Raumtheorie des Johannes Philoponus’, in E. Fries (ed.), Festschrift für
J. Klein, Göttingen 1967, 114–35.
Wieland, W., ‘Die Ewigkeit der Welt (der Streit zwischen Joannes Philoponus und
Simplicius)’, in D. Heinrich, W. Schulz, K.H. Volkmann-Schluck (eds), Die Gegenwart
der Griechen im neueren Denken, Festschrift für H.-G. Gadamer zum 60. Geburtstag,
Tübingen 1960, 291–316.
Wildberg, C., John Philoponus’ Criticism of Aristotle’s Theory of Aether, Berlin 1988.
Wildberg, C., Philoponus, Against Aristotle on the Eternity of the World, London-Ithaca
N.Y. 1987.
Wohlwill, E., ‘Ein Vorgänger Galileis im 6. Jahrhundert’, Verhandlungen der Gesellschaft
deutscher Naturforscher und Ärzte 77, part 2, 1905, 80–2; repr. in Physicalische
Zeitschrift 7, 1906, 23–32.
Wolff, M., Geschichte der Impetustheorie. Untersuchungen zum Ursprung der klassischen
Mechanik, Frankfurt 1978.
Wolff, M., Fallgesetz und Massebegriff. Zwei wissenschaftshistorische Untersuchungen zur
Kosmologie des Johannes Philoponus, Quellen und Studien zur Geschichte der
Philosophie 21, Berlin 1971.

Porphyry
Beutler, R., ‘Porphyrios’, RE 32, 1, 1953, cols 175–313.
Bidez, J., Vie de Porphyre, Gent-Leipzig, 1913.
Ebbesen, S., ‘Porphyry’s legacy to logic: a reconstruction’, Chapter 7 above, repr. from
S. Ebbesen, Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A
Study of Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981; 7.1, 133–70.
Evangeliou, C.C., Aristotle’s Categories and Porphyry, Philosophia Antiqua 48, Leiden 1988.
Evangeliou, C., ‘Aristotle’s doctrine of predicables and Porphyry’s Isagoge’, Journal of the
History of Philosophy 23, 1985, 15–34.
Finnegan, J., ‘Avicenna’s refutation of Porphyrius’, in Avicenna Commemoration Volume,
Calcutta 1956, 187–203.
Hadot, P., ‘The harmony of Plotinus and Aristotle according to Porphyry’, Chapter 6
above, translated from Atti del Convegno Internazionale sul tema Plotino e il
Neoplatonismo in Oriente e Occidente (Roma 5–9 ottobre 1970), Accademia
Nazionale dei Lincei, Rome 1974, 31–47.
Hadot, P., Porphyre et Victorinus, Paris 1968.
Lloyd, A.C., ‘Neoplatonic logic and Aristotelian logic’, Phronesis 1, 1955–6, 58–72, 146–60.
Moraux, P., ‘Porphyre, commentateur de la Physique d’Aristote’, in C. Rutten & A. Motte
(eds), Aristotelica. Mélanges offerts à Marcel de Corte, Brussels-Liège 1985, 227–39.
Noica, C., Porfir, Dexip, Ammonius. Comentarii la Categoriile lui Aristotel, însotite de
Textul Comentat Traducere, Curînt înainte si Note, Bucarest 1968.
Romano, F., Porfirio e la Fisica Aristotelica, Catania 1985.
 Select Bibliography of Secondary Literature 543

Romano, F., Porfirio di Tiro. Filosofia e Cultura nel III secolo D.C., Catania 1979.
Smith, A., ‘Porphyrian studies since 1913’, in W. Haase (ed.), Aufstieg und Niedergang der
römischen Welt, II , 36.2, Berlin 1987, 717–73.
Smith, A., Porphyry’s Place in Neoplatonic Tradition. A Study in Post-Plotinian
Neoplatonism, The Hague, 1974.

Proclus
Beutler, R., ‘Proklos’, RE 23, 1, 1957, cols 186–247.
Blumenthal, H.J., ‘Plutarch’s exposition of the de Anima and the psychology of Proclus’,
in H. Dörrie (ed.), De Jamblique à Proclus, Entretiens sur l’antiquité classique 21,
Vandoeuvres-Genève 1975, 123–47.
Festugière, A.-J, ‘L’ordre de lecture des dialogues de Platon aux Ve/VIe siècles’, Museum
Helveticum 26, 1969, 281–96; repr. in his Études de philosophie grecque, Paris 1971,
535–50.
Hadot, I., ‘Les introductions aux commentaires exégétiques chez les auteurs néoplatoniciens
et les auteurs chrétiens’, to be included in her French translation of Simplicius In
Categorias, vol. 1; also in M. Tardieu, Les règles de l’interpretation, Paris 1987.
Hadot, P., ‘Les divisions des parties de la philosophie dans l’antiquité’, Museum
Helveticum 36, 1979, 201–23; English translation by Victor Caston, unpublished.
Sheppard, Anne, ‘Proclus’ philosophical method of exegesis: the use of Aristotle and the
Stoics in the Commentary on the Cratylus’, in J. Pépin & H.D. Saffrey (eds), Proclus.
Lecteur et interprète des Anciens, Paris 1987, 137–51.
Sorabji, R., ‘Infinite power impressed, the transformation of Aristotle’s physics and
theology’, Chapter 9 above (fuller version in his Matter, Space and Motion, London-
Ithaca N.Y. 1988, ch. 15).
Steel, Carlos, ‘Proclus et Aristote sur la causalité efficiente de l’intellect divin’, in J. Pépin
& H.D. Saffrey (eds), Proclus. Lecteur et interprète des Anciens, Paris 1987.
Westerink, L.G., ‘Proclus, Procopius, Psellus’, Mnemosyne 3, 1941/42, 275–80.
Westerink, L.G. & Saffrey, H.D., Proclus, Théologie platonicienne, Budé edition, vol. 1–4,
1968–87; introductions.
Whittaker, J., ‘The historical background of Proclus’ doctrine of the “authupostata” ’, in
H. Dörrie (ed.), De Jamblique à Proclus, Entretiens sur l’antiquité classique 21,
Vandoeuvres-Genève 1975, 193–230.
Whittaker, J., ‘Varia Procliana’, Greek, Roman and Byzantine Studies 14, 1973, 425–32.

Simplicius
Dodds, E.R., ‘Simplicius’, Oxford Classical Dictionary, Oxford 1970.
Hadot, I., ‘The life and work of Simplicius in Greek and Arabic sources’, Chapter 12
above, translated from following entry.
544 Select Bibliography of Secondary Literature

Hadot, I. (ed.), Simplicius, sa vie, son oeuvre, sa survie, Peripatoi 15, Berlin 1987.
Hadot, I., Le problème du néoplatonisme alexandrin: Hiéroclès et Simplicius, Paris 1978.
Lloyd, A.C., ‘Simplicius’, in P. Edwards (ed.), Encyclopaedia of Philosophy, vol. 7, 1967,
448–9.
Praechter, K., ‘Simplikios’, RE 3 A, 1 (Zweite Reihe), 1927, cols 204–13.
Verbeke, G., ‘Simplicius’, Dictionary of Scientific Biography 12, 1975, 440–3.

*
Becker, O., ‘Formallogisches und mathematisches in griechischen philosophischen
Texten’, Philologus 100, 1956, 108–12.
Becker, O., ‘Zum Text eines mathematischen Beweises im eudemischen Bericht über die
Quadraturen der Möndchen durch Hippokrates von Chios bei Simplikios (Arist.
Phys. S. 66 Diels)’, Philologus 99, 1955, 312–16.
Blumenthal, H.J., ‘The psychology of (?) Simplicius’ commentary on the de Anima’, in
H.J. Blumenthal & A.C. Lloyd (eds), Soul and the Structure of Being in Late
Neoplatonism, Liverpool 1982, 73–93.
Blumenthal, H.J., ‘Neoplatonic elements in the de Anima commentaries’, Chapter 13
above, repr. with revisions from Phronesis 21, 1976, 64–87.
Bossier, F. and Steel, C., ‘Priscianus Lydus en de In de Anima van Pseudo (?) Simplicius’,
Tijdschrift voor Filosofie 34, 1972, 761–822.
Cameron, A., ‘The last days of the Academy at Athens’, Proceedings of the Cambridge
Philological Society 195, 1969, 7–29; French abridgment, ‘Le fin de l’académie’, in
Le néoplatonisme, Actes du colloque de Royaumont 9–13 Juin 1969, Paris, 1971,
281–90.
Clagett, M., ‘A thirteenth-century fragment of Simplicius’ commentary on the Physics of
Aristotle, Quadratura per lunulas’, in J.H. Mundy, R.W. Emery & B.N. Nelson (eds),
Essays in Medieval Life and Thought presented in honour of A.P. Evans, New York
1955, 99–108.
Gätje, H., ‘Simplikios in der arabischen Überlieferung’, Der Islam 59, 1982, 6–31.
Hadot, I., Simplicius, commentaire sur les Catégories, traduction commentée sous la
direction d’Ilsetraut Hadot, 1 (Introduction, first part), 3 (commentary on the first
chapter of the Categories), Leiden 1989.
Hadot, I., ‘La vie et l’oeuvre de Simplicius d’après des sources grecques et arabes’, in id.
(ed.), Simplicius, sa vie, son oeuvre, sa survie, Berlin 1987, 3–29.
Hadot, I., ‘La doctrine de Simplicius sur l’âme raisonnable humaine dans le
commentaire sur le Manuel d’Epictète’, in H.J. Blumenthal & A.C. Lloyd (eds), Soul
and the Structure of Being in Late Neoplatonism, Liverpool 1982, 46–70.
Hoffmann, P., ‘Catégories et langage selon Simplicius – la question du “skopos” du traité
aristotélicien des Catégories’, in I. Hadot (ed.), Simplicius, sa vie, son oeuvre, sa survie,
Berlin-New York 1987, 61–90.
Hoffmann, P., ‘Simplicius’ polemics. Some aspects of Simplicius’ polemical writings
against John Philoponus, from invective to a reaffirmation of the transcendency of
 Select Bibliography of Secondary Literature 545

the heavens’, in R. Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science,
London-Ithaca N.Y. 1987, 57–83.
Hoffmann, P., ‘Sens et dénomination. Homonymie, analogie, métaphore selon le
commentaire de Simplicius sur les Catégories d’Aristote’, École pratique des hautes
études annuaire 93, 1984–85, 343–56.
Hoffmann, P., ‘Les catégories “où” et “quand” chez Aristote et Simplicius’, in P. Aubenque
(ed.), Concepts et catégories dans la pensée antique, Paris 1980, 217–45.
Hoffmann, P., ‘Simplicius: Corollarium de Loco’, in L’astronomie dans l’antiquité classique,
Actes du colloque tenu à l’Université de Toulouse-Le Mirail, 21–33 Oct. 1977, Paris
1979, 143–63.
Meyer, H., Das Corollarium de Tempore des Simplikios und die Aporien des Aristoteles zur
Zeit, Monographien zur Naturphilosophie 8, Meisenheim am Glan 1969.
Montoya Saenz, J., ‘Interpretacion por Simplicio de la teoria aristotelica del “Nous” ’,
Annales del Seminario del Metafisico, Madrid 1968, 75–95.
Sabra, A.I., ‘Simplicius’ proof of Euclid’s parallels postulate’, Journal of the Warburg and
Courtauld Institutes 32, 1969, 1–24.
Sambursky, S. & Pinès, S., The Concept of Time in Late Neoplatonism. Texts with
Translation, Introduction and Notes, Jerusalem 1981; repr. Leiden 1987.
Sambursky, S., The Concept of Place in Late Neoplatonism. Texts with Translation,
Introduction and Notes, Jerusalem 1982.
Sonderegger, E., Simplikios. Über die Zeit: Ein Kommentar zum Corollarium de Tempore,
Hypomnemata 70, Göttingen 1982.
Sorabji, R., Matter, Space and Motion, London-Ithaca N.Y. 1988, chs 1 and 12.
Sorabji, R., Time, Creation and the Continuum, London-Ithaca N.Y. 1983, ch. 5.
Steel, C., The Changing Self. A Study on the Soul in Later Neoplatonism, Iamblichus,
Damascius and Priscianus, Brussels 1978.
Tardieu, M., Coutumes nautiques mésopotamiennes chez Simplicius, forthcoming.
Tardieu, M., ‘Les calendriers en usage à H.arrân d’après les sources arabes et le
commentaire de Simplicius à la Physique d’Aristote’, in I. Hadot (ed.), Simplicius, sa
vie, son oeuvre, sa survie, Berlin-New York 1987, 40–57.
Tardieu, M., ‘S.âbiens coraniques et “S.âbiens” de H.arrân’, Journal asiatique 274, 1986, 1–44.
Tsouyopoulos N., ‘Die Entstehung physikalischer Terminologie aus der neoplatonischen
Metaphysik’, Archiv für Begriffsgeschichte 13, 1969, 7–33.
Vamvoukakis N., ‘Les catégories aristotéliciennes vues par Simplicius’, in P. Aubenque
(ed.), Concepts et catégories dans la pensée antique, Paris 1980, 253–69.
Verbeke, G., ‘Ort und Raum nach Aristoteles und Simplicius. Eine philosophische
topologie’, in J. von Irmscher & I. Mueller (eds), Aristoteles als
Wissenschaftstheoretiker, Berlin 1983, 113–22.
Verbeke, G., ‘Le commentaire de Simplicius sur les Catégories’, Revue philosophique de
Louvain 70, 1972, 279–82.
Wieland, W., ‘Die Ewigkeit der Welt (der Streit zwischen Joannes Philoponus und
Simplicius)’, in D. Heinrich, W. Schulz, K.H. Volkmann-Schluck (eds), Die Gegenwart
546 Select Bibliography of Secondary Literature

der Griechen im neueren Denken, Festschrift für H.-G. Gadamer zum 60. Geburtstag,
Tübingen 1960, 291–316.
Zahlfleisch, J., ‘Einige Bemerkung zur Corollarien des Simplicius in seinem Kommentar
zu Aristoteles Physik’, Archiv für Geschichte der Philosophie 15, n.s. 8, 1902, 186–213.
Zahlfleisch, J., ‘Die polemik des Simplicius gegen Alexander und andere in dem
Kommentar des ersteren zu der aristotelischen Schrift de Caelo dargestelt’, Archiv für
Geschichte der Philosophie 10, n.s. 3, 1897, 191–227.
Zahlfleisch, J, ‘Die Polemik des Simplicius gegen Aristoteles Physik 4, 1–15 über den
Raum dargestelt’, Archiv für Geschichte der Philosophie 10, 1897, 85–109.

Stephanus
Lumpe, A., ‘Stephanos von Alexandrie und Kaizer Herakleios’, Classical and Medieval
Dissertationes 9, 1973, 150–9.
Usener, H., ‘De Stephano Alexandrino’, Bonner Lektionskatal, Sommer 1879, Bonn, 1880;
repr. in his Kleine Schriften 3, Leipzig-Berlin 1914, 247–322.
Blumenthal, H.J., ‘John Philoponus and Stephanus of Alexandria, two Neoplatonist
Christian commentators on Aristotle?’, in D.J. O’Meara (ed.), Neoplatonism and
Christian Thought, Norfolk 1982, 54–66; 244–6.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981 (in 7.1, pp. 133–70,
vocabulary); pp. 133–70 repr. as ‘Porphyry’s legacy to logic: a reconstruction’,
Chapter 7 above.
Noica, C., Ammonius, Stephanus. Comentarii la Tratatul Despre Interpretare al lui Aristotel,
însotite de Textul Comentat, Traducere, Curînt înainte, note si Comentariu, Bucarest 1971.
Roueché, Mossman, ‘The definitions of philosophy and a new fragment of Stephanus of
Alexandria’, in preparation.
Roueché, Mossman, ‘Byzantine philosophical texts of the seventh century’, Jahrbuch der
Oesterreichischen Byzantinistik 23, 1974, 61–76.
Vancourt, R., Les derniers commentateurs alexandrins d’Aristote: l’école d’Olympiodore,
Étienne d’Alexandrie, Lille 1941.
Westerink, L.G., Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,
Amsterdam 1967.

Syrianus
Cardullo, L., ‘Siriano nella storiografia filosofica moderna e contemporanea’, Siculorum
Gymnasium n.s. 40, 1987, 71–182.
Cardullo, L., ‘Syrianus’ lost commentaries on Aristotle’, Bulletin of the Institute of
Classical Studies 33, 1986, 112–24.
 Select Bibliography of Secondary Literature 547

Kremer, K., Der Metaphysikbegriff in den Aristoteles-Kommentaren der Ammonius-

Schule, Beiträge zur Geschichte der Philosophie und Theologie des Mittelalters 39, 1,
Münster 1961.
Madigan, A., ‘Syrianus and Asclepius on forms and intermediates in Plato and Aristotle’,
Journal of the History of Philosophy 24, 1986, 49–71.
O’Meara, D.J., ‘Le problème de la métaphysique dans l’antiquité tardive’, Freiburger
Zeitschrift für Theologie und Philosophie 23, 1986, 3–22.
Praechter, K., ‘Syrianos’ (1), in RE 4 A, 2, 1932, cols 1728–75.
Praechter, K., Review of Kroll, Syrianus in Metaphysica, CAG 6, Göttingische gelehrte
Anzeigen 1903; repr. in his Kleine Schriften, H. Dörrie (ed.), Hildesheim 1973, 246–63.
Saffrey, H.D., ‘How did Syrianus regard Aristotle?’, Chapter 8 above, trans. from
J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2,
Berlin-New York 1987, 205–14.
Sheppard, Anne, ‘Monad and dyad as cosmic principles in Syrianus’, in H.J. Blumenthal
& A.C. Lloyd (eds), Soul and the Structure of Being in Late Neoplatonism, Liverpool
1982, 1–14.
Sorabji, R., Matter, Space and Motion, London-Ithaca N.Y. 1988.

Themistius
Blumenthal, H.J., ‘Themistius: the last Peripatetic commentator on Aristotle?’, Chapter 5
above; repr. from G. Bowersock et al. (eds), Arktouros, Hellenic Studies presented to
B. Knox, Berlin 1979, 391–400.
Stegemann, W., ‘Themistius’, RE 5, 1934, cols 1642–80.
Verbeke, G., ‘Themistius’, Dictionary of Scientific Biography 13, 1976, 307–9.

Badawi, A., La transmission de la philosophie grecque au monde arabe, Paris 1968, 100–4,
166–80 (French translation of Themistius’ ‘Reply to Maximus on the reduction of the
second and third syllogistic figures to the first’).
Blumenthal, H.J., ‘Photius on Themistius (Cod.74): did Themistius write commentaries
on Aristotle?’, Hermes 107, 1979, 178–82.
Brown, H.V.B, ‘Avicenna and the Christian philosophers in Baghdad’, in S.M. Stern et al.
(eds), Islamic Philosophy and the Classical Tradition, Essays presented to Richard
Walzer, Oriental Studies 5, Oxford 1972, 35–48.
Browne, G. M., ‘Ad Themistium Arabum’, Illinois Classical Studies 11, 1986, 223–45.
Frank, R.M., ‘Some textual notes of the oriental versions of Themistius’ paraphrase of
Book 1 of the Metaphysics’, Cahiers de Byrsá 8, 1958–9, 215–30.
Lyons, M.C. (ed.), Themistius Paraphrasis in libros Aristotelis de Anima, London 1973.
Lyons, M.C., ‘An Arabic translation of the commentary of Themistius’, Bulletin of the
School of Oriental and African Studies 17, 1955, 426–35.
548 Select Bibliography of Secondary Literature

Mahoney, E.P., ‘Neoplatonism, the Greek commentators and Renaissance Aristotelianism’,

in D.J. O’ Meara (ed.), Neoplatonism and Christian Thought, Norfolk 1982, 169–84,
264–82.
Mahoney, E.P., ‘Themistius and the Agent Intellect in James of Viterbo and other
thirteenth-century philosophers’, Augustiniana 23, 1973, 422–67.
Mattock, J.N., ‘The supposed epitome by Themistius of Aristotle’s zoological works’,
Akten des VII Kongress für Arabistik und Islamwissenschaften, 260–7.
Minio-Paluello, L. (ed.), Aristoteles Latinus 1.1–5, Paris 1961; includes Pseudo-Augustini
Paraphrasis Themistiana.
Moraux, P., ‘Le “de Anima” dans la tradition grecque. Quelques aspects de
l’interprétation du traité, de Théophraste à Thémistius’, in G.E.R. Lloyd & G.E.L
Owen (eds), Aristotle on Mind and the Senses, Cambridge 1978, 281–324.
O’Donnell, J.R., ‘Themistius’ paraphrasis of the Posterior Analytics in Gerard of
Cremona’s translation’, Mediaeval Studies 20, 1958, 239–315.
Pinès, S., ‘Some distinctive metaphysical conceptions in Themistius’ commentary on Book
Lambda and their place in the history of philosophy’, in J. Wiesner (ed.), Aristoteles:
Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987, 177–206.
Rose, V., ‘Über eine angebliche Paraphrase des Themistius’, Hermes 2, 1867, 212.
Schroeder, F.M., ‘Themistius’ paraphrase of Aristotle de Anima III , 4–8’, in R.B. Todd &
F.M. Schroeder (eds), Two Greek Commentators on the Active Intellect, Pontifical
Institute, Toronto, forthcoming.
Steel, C., ‘Des commentaires d’Aristote par Thémistius?’, Revue philosophique de Louvain
71, 1973, 669–80.
Todd, R.B. & Shroeder, F.M., Two Greek Aristotelian Commentators on the Active
Intellect, Pontifical Institute, Toronto, forthcoming.
Todd, R.B., ‘Themistius and the traditional intepretation of Aristotle’s theory of
phantasia’, Acta Classica 24, 1981, 49–59.
Verbeke, G., Thémistius. Commentaire sur le traité de l’âme d’Aristote, Paris 1957 (with an
introduction), Corpus Latinum Commentariorum in Aristotelem Graecorum 1.
Zimmermann, F.W. & Brown, H.V.B., Review of A. Badawi, Commentaires sur Aristote
perdus en grec et autres épitres, Beirut 1971, ‘Neue arabischen Übersetzungen aus
dem Bereich der spätantiken griechischen Philosophie’, Der Islam 50, 1973, 313–24.

Philosophical subjects

Logic
General
Benakis, L., ‘Commentaries and commentators on the logical works of Aristotle in
Byzantium’, in R. Claussen & R. Daube-Schackat (eds), Gedankenzeichen. Festschrift
für Klaus Oehler, Tübingen 1988, 3–12.
 Select Bibliography of Secondary Literature 549

Brandis, C.A., ‘Über die Reihenfolge der Bücher des aristotelischen Organons und ihre
griechischen Ausleger’, Abh. Akad. Berlin, 1833, 249–99.
Dürr, K., The Propositional Logic of Boethius, Studies in Logic and the Foundations of
Mathematics, Amsterdam 1951.
Ebbesen, S., ‘Ancient scholastic logic as the source of medieval scholastic logic’, in
N. Kretzmann, A. Kenny & J. Pinborg (eds), The Cambridge History of Later Medieval
Philosophy, Cambridge 1982, 101–27.
Ebbesen, S., ‘Philoponus, “Alexander” and the origins of medieval logic’, Chapter 19 above.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Ebbesen, S., ‘The contribution of the Greek Commentators on the Organon to the
formation of Western scholasticism’, in Proceedings of the 1978 World Congress on
Aristotle, vol. 1, Athens 1981, 183–6.
Ebbesen, S., ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and Boethius;
ancient scholasticism and twelfth-century Western Europe. With an edition of the
Tractatus de Paralogismis’, Cahiers de l’Institut du moyen-âge grec et latin 16,
Copenhagen 1976, 1–128.
Kneale, W. & M., The Development of Logic, Oxford 1962.
Lloyd, A.C., ‘Neoplatonic logic and Aristotelian logic’, Phronesis 1, 1955–56, 58–72,
146–60.
Mueller, Ian, ‘Stoic and Peripatetic logic’, Archiv für Geschichte der Philosophie 51, 1969,
173–87.
Rijk, L.M. de, ‘Boèce logicien et philosophe, ses positions et sa métaphysique de l’être’, in
L. Obertello (ed.), Congresso Internazionale di Studi Boeziani – Atti, Roma 1981,
141–56.
Rijk, L.M. de, Logica Modernorum, Assen, vol. 1, 1962; vol. 2, 1967.
Solmsen, F., ‘Boethius and the history of the Organon’, American Journal of Philology 31,
1944, 69–74.

Isagoge and Categories

Anton, J.P., ‘Ancient interpretations of Aristotle’s doctrine of homonyma’, Journal of the
History of Philosophy 7, 1969, 1–18.
Aubenque, P., ‘Plotin et Dexippe, exégètes des Categories d’Aristote’, in C. Rutten & A.
Motte (eds), Aristotelica. Mélanges offerts à Marcel de Corte, Brussels-Liège 1985,
7–40.
Blumenthal, H.J., ‘Pseudo-Elias and the Isagoge commentaries again’, Rheinisches
Museum 124, 1981, 188–92.
Evangeliou, C., Aristotle’s Categories and Porphyry, Philosophia Antiqua 48, Leiden 1988.
Evangeliou, C., ‘The Plotinian reduction of Aristotle’s Categories’, Ancient Philosophy 7,
1987, 147–62.
550 Select Bibliography of Secondary Literature

Evangeliou, C., ‘Aristotle’s doctrine of predicables and Porphyry’s Isagoge’, Journal of the
History of Philosophy 23, 1985, 15–34.
Gottschalk, H.B., ‘The earliest Aristotelian commentators’, Chapter 3 above, repr. with
revisions from W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2,
Berlin 1987, 1079–174.
Hadot, P., ‘The harmony of Plotinus and Aristotle according to Porphyry’, Chapter 6
above, translated from Atti del Convegno Internazionale sul tema Plotino e il
Neoplatonismo in Oriente e Occidente (Roma 5–9 ottobre 1970), Accademia
Nazionale dei Lincei, Rome 1974, 31–47.
Hadot, P., ‘Un fragment du commentaire perdu de Boèce sur les Catégories d’Aristote
dans le codex Bernensis 363’, Archives d’histoire doctrinale et littéraire du moyen-âge
26, 1959, 11–27.
Henry, P., ‘Trois apories orales de Plotin sur les Catégories’, Zetesis. Mélanges de Strycker,
Anvers 1973, 234–65.
Hoffmann, P., ‘Catégories et langage selon Simplicius – la question du “skopos” du traité
aristotélicien des Catégories’, in I. Hadot (ed.), Simplicius, sa vie, son oeuvre, sa survie,
Berlin-New York 1987, 61–90.
Hoffmann, P., ‘Sens et dénomination. Homonymie, analogie, métaphore selon le
commentaire de Simplicius sur les Catégories d’Aristote’, École pratique des hautes
études annuaire 93, 1984–85, 343–56.
Hoffmann, P., ‘Les catégories “où” et “quand” chez Aristote et Simplicius’, in P. Aubenque
(ed.), Concepts et catégories dans la pensée antique, Paris 1980, 217–45.
Luna, C., ‘La relation chez Simplicius’, in I. Hadot (ed.), Simplicius, sa vie, son oeuvre,
sa survie, Berlin-New York 1987, 113–47.
Marcovich, M., ‘Pseudo-Elias on Heraclitus’, American Journal of Philology 96, 1975,
31–4.
Mignucci, M., ‘Aristotle’s definitions of relatives in Categories 7’, Phronesis 31, 1986,
101–27.
Noica, C., Porfir, Dexip, Ammonius. Comentarii la Categoriile lui Aristotel, însotite de
Textul Comentat Traducere, Curînt înainte si Note, Bucarest 1968.
Plezia, M., De Commentariis Isagogicis, Cracow 1947.
Praechter, K., review of Prolegomena and in Isagogem (CAG ), Göttingische gelehrte
Anzeigen 170, Göttingen 1908.
Strange, S.K., ‘Plotinus, Porphyry and the Neoplatonic interpretation of the Categories’,
in W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2, Berlin 1987,
955–74.
Usener, H., Review of L. Spengel (CAG edition of Dexippus), Litterar. Centralblatt, 1860,
124–5.
Vamvoukakis N., ‘Les catégories aristotéliciennes vues par Simplicius’, in P. Aubenque
(ed.), Concepts et catégories dans la pensée antique, Paris 1980, 253–69.
Verbeke, G., ‘Le commentaire de Simplicius sur les Catégories’, Revue philosophique de
Louvain 70, 1972, 279–82.
 Select Bibliography of Secondary Literature 551

Westerink, L.G., Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,

Amsterdam 1967.
Wurm, K., Substanz und Qualität: ein Beitrag zur Interpretation der plötinischen Traktat
VI 1, 2 und 3, Berlin 1973.

de Interpretatione
Arens, H., Aristotle’s Theory of Language and its Tradition, Amsterdam Studies in the
Theory and History of Linguistic Science 29, Amsterdam 1984.
Busse, A., Die neuplatonischen Ausleger der Isagoge des Porphyrius (repeated with
addition in CAG 18.2), Berlin 1892.
Hoffmann, J.G.E., de Hermeneuticis apud Syros Aristotelis, Leipzig 1869.
Isaak, I., Le Peri hermeneias en Occident, de Boèce à Saint Thomas. Histoire littéraire d’un
traité d’Aristote, Bibliothèque thomiste 29, Paris 1953; includes the first 10 chapters
of William of Moerbeke’s translation, pp. 160–9.
Kustas, G.L., Studies in Byzantine Rhetoric [pp. 101–26, The Commentators on Aristotle’s
Categories and on Porphyry’s Isagoge], Analecta Vlatadon 17, Thessaloniki 1973.
Noica, C., Ammonius, Stephanus. Comentarii la Tratatul Despre Interpretare al lui
Aristotel, însotite de Textul Comentat, Traducere, Curînt înainte, note si Comentariu,
Bucarest 1971.
Nuchelmans, G., Theories of the Proposition. Ancient and Medieval Conceptions of the
Bearers of Truth and Falsity, Amsterdam 1973.
Prior, A.N., ‘The logic of negative terms in Boethius’, Franciscan Studies 13, 1953, 1–16.
Sheppard, Anne, ‘Proclus’ philosophical method of exegesis: the use of Aristotle and the
Stoics in the Commentary on the Cratylus’, in J. Pépin & P. Saffrey (eds), Proclus.
Lecteur et interprète des Anciens, Paris 1987, 137–51.
Sullivan, M.W., Apuleian Logic. The Nature, Sources and Influences of Apuleius’ Peri
Hermeneias, Amsterdam 1967.
Tarán, L., Anonymous’ Commentary on Aristotle’s de Interpretatione (Codex Parisinus
Graecus 2064), Beiträge zür Klassische Philologie 95, Meisenheim am Glan 1978.
Zimmermann, F., Al-Farabi’s Commentary and Short Treatise on Aristotle’s de
Interpretatione, translated with Introduction and Notes, Oxford 1981.

Analytics
Ebbesen, S., ‘Analyzing syllogisms, or Anonymous Aurelianensis III – the (presumably)
earliest extant Latin commentary on the Prior Analytics, and its Greek model’,
Cahiers de l’Institut du moyen-âge grec et latin 37, Copenhagen 1981, 1–20.
Lee, T.-S., Die griechische Tradition der aristotelischen Syllogistik in der Spätantike, eine
Untersuchung über die Kommentare zu den Analytica Priora von Alexander
Aphrodisiensis, Ammonius und Philoponus, Hypomnemata 79, Göttingen 1984.
Minio-Paluello, L., ‘A Latin commentary (?translated by Boethius) on the Prior
Analytics and its Greek sources’, Journal of Hellenic Studies 77, 1957, part 1, 93–102.
552 Select Bibliography of Secondary Literature

Moraux, P., Le commentaire d’Alexandre d’Aphrodise aux Secondes Analytiques d’Aristote,

Peripatoi 13, Berlin 1979. Review by E. Berti, Archiv für Geschichte der Philosophie 63,
1981, 84–6.
O’Donnell, J.R., ‘Themistius’ paraphrasis of the Posterior Analytics in Gerard of
Cremona’s translation’, Mediaeval Studies 20, 1958, 239–315.
Rossi, P., ‘Tracce della versione latina di un commento greco ai Secondi Analitici nel
Commentarius in Posteriorium Analyticorum di Grossatesta’, Rivista di Filosofia
Neoscolastica 1978, 433–9.
Shiel, J., ‘A recent discovery, Boethius’ notes on the Prior Analytics’, Vivarium 20, 1982,
128–41.
Westerink, L.G. (ed.), ‘Elias on the Prior Analytics’, Mnemosyne 14, 1961, 126–39; repr. in his
Texts and Studies in Neoplatonism and Byzantine Literature, Amsterdam 1980, 59–72.

Topics and Sophistici Elenchi

Ebbesen, S., ‘The theory of loci in Antiquity and the Middle Ages’, in K. Jacobi (ed.), Acts
of the Eighth European Symposium on Medieval Logic and Semantics, Munich,
forthcoming.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Edlow, R.B., Galen on language and ambiguity; an English translation of Galen’s De
captionibus, with introduction, text and commentary, Philosophia Antiqua 31, Leiden
1977.
Green-Pedersen, N.J., The Tradition of the Topics in the Middle Ages: the commentaries
on Aristotle’s and Boethius’ Topics, Analytica, Munich 1984.
Stump, E., ‘Boethius works on the Topics’, Vivarium 12, 1974, 77–93.
Wallies, M., ‘Die griechischen Ausleger der aristotelischen Topik’, Wissenschaftliche
Beilage zum Programm des Sophien–Gymnasiums zu Berlin 65, 1891, 1–27.

Metaphysics
Benakis, L., ‘The problem of general concepts in Neoplatonism and Byzantine thought’,
in D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Norfolk 1982, 75–86,
248–9.
Frank, R.M., ‘Some textual notes of the oriental versions of Themistius’ paraphrase of
Book 1 of the Metaphysics’, Cahiers de Byrsá 8, 1958–9, 215–30.
Freudenthal, J., ‘Die durch Averroes erhaltenen Fragmente Alexanders zur Metaphysik
des Aristoteles untersucht und übersetzt’, Abhandlung der (k.) preussischen Akademie
der Wissenschaften, Berlin, 1884.
Giocarinis, K., ‘Eustratius of Nicaea’s defence of the doctrine of Ideas’, Franciscan Studies
24, 1964, 159–204.
 Select Bibliography of Secondary Literature 553

Ioannou, P., ‘Die Definition des Seins bei Eustratius von Nicaea. Die Universalienlehre
in der Byzantinischen Theologie im XI Jahrhundert’, Byzantinische Zeitschrift 47,
1954, 258–368.
Lloyd, A.C., ‘The significance of Alexander’s theory of universals’, in Proceedings of the
World Congress on Aristotle, vol. 1, Athens 1981, 155–9.
Madigan, A., ‘Syrianus and Asclepius on forms and intermediates in Plato and Aristotle’,
Journal of the History of Philosophy 24, 1986, 49–71.
Mignucci, M., ‘Puzzles about identity, Aristotle and his Greek commentators’, in
J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 1,
Berlin-New York 1985, 57–97.
Mueller, I., ‘Aristotle’s doctrine of abstraction in the commentators’, Chapter 20 above.
O’Meara, D.J., ‘Le problème de la métaphysique dans l’antiquité tardive’, Freiburger
Zeitschrift für Theologie und Philosophie 23, 1986, 3–22.
Pinès, S., ‘Some distinctive metaphysical conceptions in Themistius’ commentary on
Book Lambda and their place in the history of philosophy’, in J. Wiesner (ed.),
Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987,
177–206.
Pinès, S., ‘Les limites de la métaphysique selon Al-Farabi, Ibn-Bassa et Maimonides:
sources et antithèses de ces doctrines chez Alexandre d’Aphrodise et chez
Thémistius’, Sprache und Erkenntnis im Mittelalter 1, Miscellanea Medievalia 13.1,
Berlin-New York 1981.
Ruland, H.-J., ‘Zwei arabische Fassungen der Abhandlung des Alexander von
Aphrodisias über die universalia (Quaestio I 11a)’, Nachrichten der Akademie der
Wissenschaften in Göttingen, 1979, 243–74.
Saffrey, H.D., ‘How did Syrianus regard Aristotle?’, Chapter 8 above, transl. from
J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2,
Berlin-New York 1987, 205–14.
Tarán, L., ‘Syrianus and Pseudo-Alexander’s commentary on Metaphysics E-N’, in
J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2,
Berlin-New York 1987, 215–32.
Tweedale, M.M., ‘Alexander of Aphrodisias’ views on universals’, Phronesis 29, 1984,
279–303.
Verbeke, G., ‘Aristotle’s Metaphysics as viewed by the ancient Greek commentators’, in
D.J. O’ Meara (ed.), Studies in Aristotle, Washington, D.C. 1981, 107–27.

Physics and Mathematics

Becker, O., ‘Formallogisches und Mathematisches in griechischen philosophischen
Texten’, Philologus 100, 1956, 108–12.
Becker, O., ‘Zum Text eines mathematischen Beweises im eudemischen Bericht über die
Quadraturen der Möndchen durch Hippokrates von Chios bei Simplikios (Arist.
Phys. S. 66 Diels)’, Philologus 99, 1955, 313–16.
554 Select Bibliography of Secondary Literature

Christensen de Groot, J., ‘Philoponus on de Anima 2.5, Physics 3.3 and the propagation
of light’, Phronesis 28, 1983, 177–96.
Clagett, M., ‘A thirteenth-century fragment of Simplicius’ commentary on the Physics of
Aristotle, Quadratura per lunulas’, in J.H. Mundy, R.W. Emery & B.N. Nelson (eds),
Essays in Medieval Life and Thought presented in honour of A.P. Evans, New York
1955, 99–108.
Coutant, V.C., Alexander of Aphrodisias, Commentary on Book IV of Aristotle’s
Meteorologica, diss. Columbia, 1936; reviewed by B. Einarson in Classical Weekly 31,
1937, 33.
Davidson, H.A., ‘The principle that a finite body can contain only a finite power’, S. Stein
& R. Loewe (eds), Studies in Jewish Religious and Intellectual History Presented to
Alexander Altmann, Alabama 1979, 75–92.
Davidson, H.A., ‘John Philoponus as a source of Medieval Islamic and Jewish proofs of
creation’, Journal of the American Oriental Society 89, 1969, 357–91.
Duhem, P.L., Le système du monde I and II. Histoire des doctrines cosmologiques de
Platon à Copernique, Paris 1913–17; repr. 1 Paris 1954; 2 Paris 1974.
Évrard, É, ‘Les convictions religieuses de Jean Philopon et la date de son commentaire
aux Météorologiques’, Bulletin de l’academie royale de Belgique, Classe des lettres et
sciences politiques et morales 5, ser. 39, 1953, 299–357.
Fazzo, S., Alessandro d’Afrodisia e Tolomeo: Aristotelismo e astrologia fra il II e il III
Secolo D.C.’, Rivista di storia della filosofia 4, 1988, 627–49.
Genequand, C., ‘Quelques aspects de l’idée de nature d’Aristote à Al-Ghazali’, Revue de
théologie et de philosophie 116, 1984, 105–29.
Grant, E., ‘Aristotle, Philoponus, Avempace and Galileo’s Pisan dynamics’, Centaurus 11,
1965, 79–95.
Hoffmann, P., ‘Paratasis. De la description aspectuelle des verbes grecs à la définition du
temps dans le néoplatonisme tardif ’, Revue des études grecques 96, 1983, 1–26.
Hoffmann, P., ‘Jamblique exégète du pythagoricien Archytas: trois originalités d’une
doctrine du temps’, Les études philosophiques, 1980, 307–23.
Hoffmann, P., ‘Simplicius, Corollarium de Loco’, L’astronomie dans l’antiquité classique,
Actes du Colloque tenu à l’Université de Toulouse-Le Mirail, 21–33 Oct. 1977, Paris
1979, 143–63.
Judson, L., ‘God or Nature? Philoponus on generability and perishability’, in R. Sorabji
(ed.), Philoponus and the Rejection of Aristotelian Science, London-Ithaca N.Y. 1987,
179–96.
Kraemer, J.L., ‘A lost passage from Philoponus’ Contra Aristotelem in arabic translation
(Simpl. de Caelo, I, 3, 270b5–11)’, Journal of the American Oriental Society 85, 1965,
318–27.
Lucchetta, G.A., Una Fisica Senza Matematica, Democrito, Aristotele, Filopono, Trento
1978.
Mahdi, M., ‘Alfarabi against Philoponus’, Journal of Near Eastern Studies 26, 1967,
233–60.
 Select Bibliography of Secondary Literature 555

McGuire, J.E., ‘Philoponus on Physics ii 1, phusis, hormê emphutos and the motion of
simple bodies’, Ancient Philosophy 5, 1985, 241–67.
Meyer, H., Das Corollarium de Tempore des Simplikios und die Aporien des Aristoteles zur
Zeit, Monographien zur Naturphilosophie 8, Meisenheim am Glan 1969.
Moody, E.A., ‘Galileo and Avempace, the dynamics of the leaning tower experiment’,
Journal of the History of Ideas 13, 1951, 163–93; 375–422.
Moraux, P., ‘Porphyre, commentateur de la Physique d’Aristote’, in C. Rutten & A.
Motte (eds), Aristotelica. Mélanges offerts à Marcel de Corte, Brussels-Liège 1985,
227–39.
Moraux, P., ‘Notes sur la tradition indirecte du de Caelo d’Aristote’, Hermes 82, 1954,
145–82.
Mueller, I., ‘Aristotle’s doctrine of abstraction in the commentators’, Chapter 20 above.
Pinès, S., ‘An Arabic summary of a lost work of John Philoponus’, Israel Oriental Studies
2, 1972, 320–52; repr. in S. Pinès, Studies in Arabic Versions of Greek Texts and in
Medieval Science, vol. 2, Leiden-Jerusalem 1986, 294–326.
Pinès, S., ‘La dynamique d’Ibn Bajja’, in Mélanges Alexandre Koyré I, Paris, 1964, 442–68;
repr. in S. Pinès, Studies in Arabic Versions of Greek Texts and in Medieval Science, vol.
2, Leiden-Jerusalem 1986, 440–66.
Pinès, S., ‘Les précurseurs musulmans de la théorie de l’impetus’, Archeion 21, 1938,
298–306; repr. in S. Pinès, Studies in Arabic Versions of Greek Texts and in Medieval
Science, vol. 2, Leiden-Jerusalem 1986, 409–17.
Pinès, S., ‘Omne quod movetur necesse est ab aliquo moveri, a refutation of Galen by
Alexander of Aphrodisias and the theory of motion’, Isis 52, 1961, 21–54; repr. in
S. Pinès, Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2,
Leiden-Jerusalem 1986, 218–51.
Pinès, S., ‘Un précurseur bagdadien de la théorie de l’impetus’, Isis 44, 1953, 246–51; repr.
in S. Pinès, Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2,
Leiden-Jerusalem 1986, 418–22.
Pinès, S., ‘A tenth-century philosophical correspondence’, Proceedings of the American
Academy for Jewish Research 24, 1955, 103–36; repr. in A. Hyman (ed.), Essays in
Medieval Jewish and Islamic Philosophy, New York 1977, 357–90.
Pinès, S., ‘Études sur Awhad al-Zamân Abu’ 1-Barakât al-Baghdâdî’, Revue des études
juives n.s. 3, 1938, 3–64 and n.s. 4, 1938, 1–33.
Rescher, N. & Marmura, M., Alexander of Aphrodisias, the Refutation by Alexander of
Aphrodisias of Galen’s Treatise of the Theory of Motion, Islamabad (Islamic Research
Institute) 1965.
Romano, F., Porfirio e la Fisica Aristotelica, Catania 1985.
Ruland, H.-J., ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias
über das Wachstum (Quaestio I 5)’, Nachrichten der Akademie der Wissenschaften in
Göttingen, 1981, 51–74.
Sabra, A.I., ‘Simplicius’ proof of Euclid’s parallels postulate’, Journal of the Warburg and
Courtauld Institutes 32, 1969, 1–24.
556 Select Bibliography of Secondary Literature

Sambursky, S. & Pinès, S., The Concept of Time in Late Neoplatonism. Texts with
Translation, Introduction and Notes, Jerusalem 1981; repr. Leiden 1987.
Sambursky, S., The Concept of Place in Late Neoplatonism. Texts with Translation,
Introduction and Notes, Jerusalem 1982.
Sambursky, S., The Physical World of Late Antiquity, London 1962.
Schmitt, C.B., Gianfrancesco Pico della Mirandola (1469–1533) and his Critique of
Aristotle, The Hague 1967.
Sedley, D., ‘Philoponus’ conception of space’, in R. Sorabji (ed.), Philoponus and the
Rejection of Aristotelian Science, London-Ithaca N.Y. 1987, 140–53.
Sharples, R.W., ‘The Unmoved Mover and the motion of the heavens in Alexander of
Aphrodisias’, Apeiron 17, 1983, 62–6.
Sharples, R.W., Alexander of Aphrodisias on Fate, London 1983.
Sonderegger, E., Simplikios. Über die Zeit: Ein Kommentar zum Corollarium de Tempore,
Hypomnemata 70, Göttingen 1982.
Sorabji, R., Matter, Space and Motion, London-Ithaca N.Y. 1988.
Sorabji. R., ‘Infinite power impressed, the transformation of Aristotle’s physics and
theology’, Chapter 9 above (fuller version in his Matter, Space and Motion, London-
Ithaca N.Y. 1988, ch. 15).
Sorabji, R. (ed.), Philoponus and the Rejection of Aristotelian Science, London-Ithaca N.Y.
1987.
Sorabji, R., Time, Creation and the Continuum, London-Ithaca N.Y. 1983.
Steel, C., ‘Proclus et Aristote sur la causalité efficiente de l’intellect divin’, in J. Pépin &
H.D. Saffrey (eds), Proclus. Lecteur et interprète des Anciens, Paris 1987.
Todd, R.B., ‘Some concepts in physical theory in John Philoponus’ Aristotelian
commentaries’, Archiv für Begriffsgeschichte 24, 1980, 151–70.
Todd, R.B., Alexander of Aphrodisias on Stoic Physics, Leiden 1976.
Verbeke, G., ‘Ort und Raum nach Aristoteles und Simplicius. Eine philosophische
topologie’, in J. von Irmscher & I. Mueller (eds), Aristoteles als
Wissenschaftstheoretiker, Berlin 1983, 113–22.
Verbeke, G., ‘Some later neoplatonic views on divine creation and the eternity of the
world’, in D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Norfolk 1982,
45–53 and 241–4.
Verbeke, G., ‘La physique d’Aristote et les anciens commentateurs grecs’, in Proceedings
of the 1978 World Congress on Aristotle, vol. 1, Athens 1981, 187–92.
Wieland, W., ‘Zur Raumtheorie des Johannes Philoponus’, in E. Fries (ed.), Festschrift für
J. Klein, Göttingen 1967, 114–35.
Wieland, W., ‘Die Ewigkeit der Welt (der Streit zwischen Joannes Philoponus und
Simplicius)’, in D. Heinrich, W. Schulz, K.H. Volkmann-Schluck (eds), Die Gegenwart
der Griechen im neueren Denken, Festschrift für H.-G. Gadamer zum 60. Geburtstag,
Tübingen 1960, 291–316.
Wildberg, C., John Philoponus’ Criticism of Aristotle’s Theory of Aether, Peripatoi 18,
Berlin 1988.
 Select Bibliography of Secondary Literature 557

Wildberg, C., Philoponus, Against Aristotle on the Eternity of the World, London-Ithaca
N.Y. 1987.
Wohlwill, E., ‘Ein Vorgänger Galileis im 6. Jahrhundert’, Verhandlungen der Gesellschaft
deutscher Naturforscher und Ärzte 77, part 2, 1905, 80–2; repr. in Physicalische
Zeitschrift 7, 1906, 23–32.
Wolff, M., ‘Philoponus and the rise of preclassical dynamics’, in R. Sorabji, (ed.),
Philoponus and the Rejection of Aristotelian Science, London-Ithaca N.Y. 1987,
84–120.
Wolff, M., Geschichte der Impetustheorie. Untersuchungen zum Ursprung der klassischen
Mechanik, Frankfurt 1978.
Wolff, M., Fallgesetz und Massebegriff. Zwei wissenschaftshistorische Untersuchungen zur
Kosmologie des Johannes Philoponus, Quellen und Studien zur Geschichte der
Philosophie 21, Berlin 1971.
Wolfson, H.A., The Philosophy of the Kalam, Cambridge Mass. 1976.
Wolfson, H.A., ‘The problem of the souls of the spheres from the Byzantine
commentaries on Aristotle through the Arabs and St Thomas to Kepler’, Dumbarton
Oaks Papers 16, 1965, 67–93.
Wolfson, H.A., ‘The Kalam arguments for creation in Saadia, Averroes, Maimonides and
St Thomas’, Saadia Anniversary Volume, American Academy of Jewish Research, Texts
and Studies 2, New York 1943, 197–245.
Wolfson, H.A., Crescas’ Critique of Aristotle. Problems in Aristotle’s Physics in Jewish and
Arabic Philosophy, Cambridge Mass. 1929.
Zahlfleisch, J., ‘Einige Bemerkungen zur Corollarien des Simplicius in seinem
Kommentar zu Aristoteles Physik’, Archiv für Geschichte der Philosophie 15, n.s. 8,
1902, 186–213.
Zahlfleisch, J., ‘Die polemik des Simplicius gegen Alexander und andere in dem
Kommentar des ersteren zu der aristotelischen Schrift de Caelo dargestelt’, Archiv für
Geschichte der Philosophie 10, n.s. 3, 1897, 191–227.
Zahlfleisch, J., ‘Die Polemik des Simplicius gegen Aristoteles Physik 4, 1–15 über den
Raum dargestelt’, Archiv für Geschichte der Philosophie 10, 1897, 85–109.
Zimmermann, F., ‘Philoponus’ impetus theory in the Arabic tradition’, in R. Sorabji (ed.),
Philoponus and the Rejection of Aristotelian Science, London-Ithaca N.Y. 1987, 121–9.

Psychology
General
Blumenthal, H.J., ‘Alexander of Aphrodisias in the later Greek commentaries on
Aristotle’s de Anima’, in J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux
gewidmet, vol. 2, Berlin-New York 1987, 90–106.
Blumenthal, H.J., ‘The psychology of (?) Simplicius’ commentary on the de Anima’, in
H.J. Blumenthal & A.C. Lloyd (eds), Soul and the Structure of Being in Late
Neoplatonism, Liverpool 1982, 73–93.
558 Select Bibliography of Secondary Literature

Blumenthal, H.J., ‘Neoplatonic interpretations of Aristotelian phantasia’, Review of

Metaphysics 31, 1977, 242–57.
Blumenthal, H.J., ‘Neoplatonic elements in the de Anima commentaries’, Chapter 13
above, repr. with revisions from Phronesis 21, 1976, 64–87.
Blumenthal, H.J., ‘Plutarch’s exposition of the de Anima and the psychology of Proclus’,
in H. Dörrie (ed.), De Jamblique à Proclus, Entretiens sur l’antiquité classique 21,
Vandoeuvres-Genève 1975, 123–47.
Blumenthal, H.J., ‘Did Iamblichus write a commentary on the de Anima?’, Hermes 102,
1974, 540–56.
Donini, P.L., ‘L’anima e gli elementi nel “de Anima” di Alessandro di Afrodisia’, Atti della
Accademia delle Scienze di Torino, Classe di Scienze Morali, Storiche e Filologiche
105, 1970–71, 61–107.
Donini, P.L., ‘Il de Anima di Alessandro di Afrodisia e Michele Efesio’, Rivista di Filologia
e di Instruzione Classica 96, 1968, 316–23.
Gätje, H., ‘Die Shrift des Alexander von Aphrodisias über das Sahen’ (with G. Losseren),
Studien zur Überlieferung der aristotelischen Psychologie im Islam, Annales
Universitatis Saraviensis 11, Heidelberg 1971, 140–72.
Gätje, H., ‘Die arabische Übersetzung der Schrift des Alexander von Afrodisias über die
Farbe’, Nachrichten der Academie der Wissenschaften in Göttingen, 1967, 341–82.
Gauthier, R., Introduction to Thomas Aquinas Opera Omnia, Leonine Edition 45.1,
1984.
Hager, F.P., ‘Die Aristotelesinterpretation des Alexander von Aphrodisias und die
Aristoteleskritik Plotins bezüglich der Lehre vom Geist’, Archiv für Geschichte der
Philosophie 46, 1964, 174–87.
Lyons, M.C., ‘An Arabic translation of the commentary of Themistius’, Bulletin of the
School of Oriental and African Studies 17, 1955, 426–35.
Moraux, P., ‘Le “de Anima” dans la tradition grecque. Quelques aspects de
l’interprétation du traité, de Théophraste à Thémistius’, in G.E.R. Lloyd & G.E.L.
Owen (eds), Aristotle on Mind and the Senses, Cambridge 1978, 281–324.
Ruland, H.-J., ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias
über die Sinneswahrnehmung’, Nachrichten der Akademie der Wissenschaften in
Göttingen, 1978, 159–225.
Sorabji, R., ‘From physiology to intentionality’, in M. Nussbaum & A. Rorty (eds),
Aristotle’s de Anima, Oxford 1989.
Steel, C., The Changing Self. A Study on the Soul in Later Neoplatonism, Iamblichus,
Damascius and Priscianus, Brussels 1978.
Todd, R.B., ‘Themistius and the traditional interpretation of Aristotle’s theory of
phantasia’, Acta Classica 24, 1981, 49–59.
Verbeke, G., ‘Levels of human thinking in Philoponus’, in C. Laga, J.A. Munitz, L. van
Rompay (eds), After Chalcedon, Studies in Theology and Church History
offered to Professor Albert van Roey for his seventieth birthday, Louvain 1986,
451–70.
 Select Bibliography of Secondary Literature 559

Intellect
Balleriaux, O., Thémistius. Son interpretation de la noétique aristotélicienne, diss., Liège
1941.
Davidson, H.A., ‘Al-Farabi and Avicenna on the active intellect’, Viator 3, 1972, 109–78.
Finnegan, J., ‘Al-Farabi et le “Peri Nou” d’Alexandre d’Aphrodise’, Mélanges L. Massignon,
1957, 133–52.
Grünaz, A., Die Abhandlung Alexanders von Aphrodisias über den Intellect. Aus
handschriftlichen Quellen zum ersten Male herausgegeben und durch die Abhandlung,
Die Nuslehre Alexanders von Aphrodisias und ihr Einfluss auf die arabisch-jüdische
Philosophie des Mittelalters eingeleitet, Berlin 1886.
Hadot, I., ‘La doctrine de Simplicius sur l’âme raisonnable humaine dans le
commentaire sur le Manuel d’Epictète’, in H.J. Blumenthal & A.C. Lloyd (eds), Soul
and the Structure of Being in Late Neoplatonism, Liverpool 1982, 46–70.
Hamelin, O., La théorie de l’intellect d’après Aristote et ses commentateurs, posthumous
ed. E. Barbotin, Paris 1953.
Hicks, R.D., Introduction to Aristotle’s De Anima, Cambridge 1907.
Jolivet, J., L’intellect selon Al-Kindi, Leiden 1971.
Hyman, A., ‘Aristotle’s theory of the intellect and its interpretation by Averroes’, in
D. J. O’Meara (ed.), Studies in Aristotle, Washington D.C. 1981, 161–90.
Kurfess, H., Zur Geschichte der Erklärung der aristotelischen Lehre vom sogenannten
‘Nous Poietikos’ und ‘Pathetikos’, diss., Tübingen 1911.
Mahoney, E.P., ‘Themistius and the Agent Intellect in James of Viterbo and other
thirteenth-century philosophers’, Augustiniana 23, 1973, 422–67.
Montoya Saenz, J., ‘Interpretacion por Simplicio de la teoria aristotelica del “Nous” ’,
Annales del Seminario del Metafisico, Madrid 1968, 75–95.
Moraux, P., Alexandre d’Aphrodise: exégète de la noétique d’Aristote, Liège-Paris
1942.
Nardi, B., ‘Il commento di Simplicio al de Anima nelle controversie della Fine del secolo
XVe al secolo XVI ’, in his Saggi dell’ Aristotelismo Padovano dal Secolo XIV al XVI ,
Florence 1958, 365–42.
Riet, A. van, ‘Fragments de l’originel grec du de Intellectu de Philopon dans une
compilation de Sophonias’, Revue philosophique de Louvain 63, 1965, 5–40.
Schroeder, F.M., ‘Light and the active intellect in Alexander and Plotinus’, Hermes 112,
1984, 239–48.
Schroeder, F.M., ‘The potential or material intellect and the authorship of the de
Intellectu, a reply to B.C. Bazán’, Symbolae Osloenses 57, 1982, 115–25.
Schroeder, F.M., ‘The analogy of the active intellect to light in the de Anima of
Alexander of Aphrodisias’, Hermes 109, 1981, 215–25.
Théry, G., Autour du décret de 1210: I. David de Dinant. Étude sur son panthéisme
matérialiste, Bibliothèque thomiste 7, Kain 1926.
Thillet, P., ‘Matérialisme et théorie de l’âme et de l’intellect chez Alexandre d’Aphrodise’,
Revue philosophique de la France et de l’étranger 171, 1981, 5–24.
560 Select Bibliography of Secondary Literature

Todd, R.B. & Shroeder, F.M., Two Greek Aristotelian Commentators on the Active
Intellect, Pontifical Institute, Toronto, forthcoming.
Verbeke, G., ‘Levels of human thinking in Philoponus’, in C. Laga, J.A. Munitz, L. van
Rompay (eds), After Chalcedon. Studies in Theology and Church History offered to
Professor Albert van Roey for his seventieth birthday, Louvain 1986, 451–70.
Wilpert, P., ‘Die Ausgestaltung der aristotelischen Lehre vom Intellectus agens bei den
griechischen Kommentatoren und in der Scholastik des 13. Jahrhunderts’, Aus der
Geisteswelt des Mittelalters. Festschrift für M. Grabmann, Beiträge zur Geschichte der
Philosophie und Theologie des Mittelalters, supp. vol. 3 i, 1935, 447–62.

History of the commentaries

(Subjects in this section appear in roughly chronological order)

The early commentators

Accatino, P., ‘Alessandro di Afrodisia e Aristotele di Mitilene’, Elenchos 6, 1985, 67–74.
Arnim, H. von, ‘Herminos’ (2), RE 8, 1, 1913, col 835.
Arnim, H. von, ‘Boethos’ (4), RE 3, 1, 1899, cols 601–3.
Becchi, F., ‘Aspasio e i peripatetici posteriori, la formula definitiva della passione’,
Prometheus 9, 1983, 83–104.
Brink, K.O., ‘Peripatos’, RE Suppl. 7, 1940, cols 899–949.
Busse, A., ‘Peripatos und Peripatetiker’, Hermes 61, 1926, 335–42.
Capelle, W., ‘Lukios’ (1), RE 13, 2, 1927, cols 1791–7.
Dalsgaard-Larsen, B., Jamblique de Chalcis: exégète et philosophe, 2 vols, Århus 1972.
Donini, P.L., Le scuole, l’anima, l’impero: la filosofia antica da Antioco a Plotino, Turin 1982.
Donini, P.L., Tre studi sull’Aristotelismo nel II secolo D.C., Historica, Politica,
Philosophica 7, Turin 1974.
Drossaart Lulofs, H.J., Nicolaus Damascenus on the Philosophy of Aristotle, Philosophia
Antiqua 13, Leiden 1965; 2nd ed. 1969.
Drossaart Lulofs, H.J., ‘Aristotle, Bar Hebraeus and Nicolaos Damascenus on animals’, in
A. Gotthelf (ed.), Aristotle on Nature and Living Things, Pittsburg-Bristol 1986,
345–57.
Düring, I., Aristotle in the Ancient Biographical Tradition, Göteborg 1957.
Fritz, K. von, ‘Nikostratos’ (26), RE 13, 1, 1936, 547–51.
Gercke, A., ‘Alexandros (von Aigai)’ (2), RE 1, 2, col. 1452.
Gercke, A., ‘Boethos’ (9), RE 3, 1, 1899, cols 603–4.
Glucker, J., Antiochus and the Late Academy, Hypomnemata 56, Göttingen 1978.
Gottschalk, H.B., ‘The earliest Aristotelian commentators’, Chapter 3 above, repr. with
revisions from W. Haase (ed.), Aufstieg und Niedergang der römischen Welt, II , 36.2,
Berlin 1987, 1079–174.
 Select Bibliography of Secondary Literature 561

Gottschalk, H.B., ‘Boethus’ psychology and the Neoplatonists’, Phronesis 31, 1986,
243–57.
Gottschalk, H.B., review of P. Moraux, Der Aristotelismus, vol. 2, in Liverpool Classical
Monthly 10.8, 1985, 122–8.
Huby, P.M., ‘An excerpt from Boethus of Sidon’s commentary on the Categories?’,
Classical Quarterly 31, 1981, 398–40.
Littig, Fr., Andronikos von Rhodos (3 vols), Erlangen 1, 1890; 2, 1894; 3, 1985.
Moraux, P., ‘Diogène Laërce et le Peripatos’, Elenchos 7, 1986, 247–94.
Moraux, P., Der Aristotelismus bei den Griechen: von Andronikos bis Alexander von
Aphrodisias, vol. 1, Die Renaissance des Aristotelismus im I. Jh. v. Chr., Peripatoi 5,
Berlin-New York 1973; vol. 2, Der Aristotelismus im I. und II. Jh. n. Chr., Peripatoi 6,
Berlin-New York 1984.
Moraux, P., ‘Die Anfänge der griechischen Aristoteles-Interpretation’, Proceedings of the
1978 World Congress on Aristotle, vol. 1, Athens 1981, 95–100.
Moraux, P., ‘Einige Aspekte des Aristotelismus von Andronikos bis Alexander von
Aphrodisias’, in J. Burian & L. Vidman (eds), Antiquitas Graeco-Romana et Tempora
Nostra, Prague 1968, 203–8.
Moraux, P., ‘L’expose de la philosophie d’Aristote chez Diogene Laërce’, Revue
philosophique de Louvain 47, 1949, 5–43.
Plezia, M., De Andronici Rhodii Studiis Aristotelicis, Polska Academia Unmiectnosci,
Archiwum Filologiczne 20, Cracow 1946.
Praechter, K., ‘Nikostratos der Platoniker’, Hermes 57, 1922, 481–517; repr. in his Kleine
Schriften, H. Dörrie (ed.), Collectanea 7, Hildesheim-New York 1973, 101–37.
Sullivan, M.W., Apuleian Logic. The Nature, Sources and Influences of Apuleius’ Peri
Hermeneias, Amsterdam 1967.
Tarán, L., review of Paul Moraux, Der Aristotelismus, vol. 1, Gnomon 46, 1981, 721–50.
Wacholder, B.Z., Nikolaos of Damascus, Berkeley 1962.

School of Athens
Beutler, R., ‘Plutarchos von Athen’, RE 21, 1, 1951, cols 962–75.
Cameron, A., ‘The last days of the Academy at Athens’, Proceedings of the Cambridge
Philological Society 195, 1969, 7–29; French abridgment, ‘Le fin de l’académie’ in
Le néoplatonisme, Actes du colloque de Royaumont 9–13 Juin, 1969, Paris, 1971,
281–90.
Cameron, A., ‘Iamblichus at Athens’, Athenaeum 45, 1967, 143–53.
Évrard, É., ‘Le maitre de Plutarque d’Athènes et les origines du néoplatonisme
athénien’, L’antiquité classique 29, 1960, 108–33; ‘Plutarque, Jamblique, Porphyre’, ibid.,
392–406.
Frantz, Alison, ‘Pagan philosophers in Christian Athens’, Proceedings of the American
Philosophical Society 119, 1975, 29–38.
Immisch, O., ‘Attikoi exegetai’, Philologus 63, 1904, 31–40.
562 Select Bibliography of Secondary Literature

Tardieu, M., ‘Les calendriers en usage à H.arrân d’après les sources arabes et le
commentaire de Simplicius à la Physique d’Aristote’, in I. Hadot (ed.), Simplicius, sa
vie, son oeuvre, sa survie, Berlin-New York 1987, 40–57.
Tardieu, M., ‘S.âbiens coraniques et “S.âbiens” de H.arrân’, Journal asiatique 274, 1986,
1–44.
Westerink, L.G. & Saffrey, H.D., Proclus. Théologie platonicienne, Budé edition, vol. 1,
Paris 1968; introduction.

School of Alexandria
Blumenthal, H.J., ‘Pseudo-Elias and the Isagoge commentaries again’, Rheinisches
Museum 124, 1981, 188–92.
Évrard, É., L’école d’Olympiodore et la composition du commentaire à la Physique de Jean
Philopon, diss., Liège 1957.
Kremer, K., ‘Die Anschauung der Ammonius (Hermeiou)-Schule über den
Wirklichkeitscharacter des Intelligiblen. Über einen Beitrag der Spätantike zur
platonisch-aristotelischen Metaphysik’, Philosophisches Jahrbuch 69, 1961, 46–63.
Kremer, K., Der Metaphysikbegriff in den Aristoteles-Kommentaren der Ammonius-
Schule, Beiträge zur Geschichte der Philosophie und Theologie des Mittelalters 39, 1,
Münister 1961.
Marcovich, M., ‘Pseudo-Elias on Heraclitus’, American Journal of Philology 96, 1975, 31–4.
Meyerhof, M., ‘La fin de l’école d’Alexandrie d’après quelques auteurs arabes’, Archeion
15, 1933, 1–15.
Meyerhof, M., ‘Von Alexandrien nach Bagdad’, Sitzungsberichte der preussischen
Akademie der Wissenschaften, 1930, 389–429.
Praechter, K., ‘Hermeias’, RE 8, 1, 1913, cols 516–38.
Roueché, Mossman, ‘The definitions of philosophy and a new fragment of Stephanus of
Alexandria’, in preparation.
Saffrey, H.D., ‘Le chrétien Jean Philopon et la survivance de l’école d’Alexandrie au VIe
siècle’, Revue des études grecques 67, 1954, 396–410.
Simon, Jules, Histoire de l’école d’alexandrie, 2 vols, Paris 1843–45.
Tarán, L., Anonymous Commentary on Aristotle’s de Interpretatione (Codex Parisinus
Graecus 2064), Beiträge zur Klassische Philologie 95, Meisenheim am Glan 1978.
Vacherot, E., Histoire critique de l’école d’Alexandrie, 2 vols, Paris 1846–51.
Vancourt, R., Les derniers commentateurs alexandrins d’Aristote: l’école d’Olympiodore,
Étienne d’Alexandrie, Lille 1941.
Westerink, L.G., ‘Academic practice about 500 A.D. in Alexandria’, unpublished paper
delivered to Dumbarton Oaks conference.
Westerink, L.G., The Greek Commentaries on Plato’s Phaedo, vol. 1, Olympiodorus,
Amsterdam 1973.
Westerink, L.G., ‘Ein astrologisches Kolleg aus dem Jahre 564’, Byzantinische Zeitschrift
64, 1971, 6–21.
 Select Bibliography of Secondary Literature 563

Westerink, L.G., Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,

Amsterdam 1967.
Westerink, L.G., Anonymous Prolegomena to Platonic Philosophy, Amsterdam 1962; new
edition in preparation; pp. ix-xxxii repr. with the revisions prepared for the new
Budé edition in Chapter 14 above.

The Arab tradition

Badawi, A., Commentaires sur Aristote perdus en grec et autres épitres, Beirut 1971.
Badawi, A., La transmission de la philosophie grecque au monde arabe, Études de
philosophie médiévale 56, Paris 1968.
Baumstark, A., Aristoteles bei den Syrern vom V bis VIII Jahrhundert, Leipzig 1900.
Booth, E.G.T., Aristotelian Aporetic Ontology in Islamic and Christian Thinkers,
Cambridge 1983.
Brown, H.V.B, ‘Avicenna and the Christian philosophers in Baghdad’, in S.M. Stern et al.
(eds), Islamic Philosophy and the Classical Tradition, Essays presented to Richard
Walzer, Oriental Studies 5, Oxford 1972, 35–48.
Browne, G.M., ‘Ad Themistium Arabum’, Illinois Classical Studies 11, 1986, 223–45.
Craig, W.L., The Kalam Cosmological Argument, London 1979.
Daiber, H., Ein Kompendium der Aristotelischen Meteorologie in der Fassung des Hunain
ibn Is hâq, Amsterdam 1975.
˙
Davidson, H.A., Proofs for Eternity, Creation and the Existence of God in Medieval
Islamic and Jewish Philosophy, New York-Oxford 1987.
Davidson, H.A., ‘The principle that a finite body can contain only a finite power’, in S.
Stein & R. Loewe (eds), Studies in Jewish Religious and Intellectual History Presented
to Alexander Altmann, Alabama 1979, 75–92.
Davidson, H.A., ‘Al-Farabi and Avicenna on the active intellect’, Mediaeval and
Renaissance Studies 3, 1972, 109–78.
Davidson, H.A., Averroes’ Middle Commentary on Porphyry’s Isagoge and on Aristotle’s
Categoriae, Cambridge Mass. 1969.
Davidson, H.A., ‘Philoponus as a source of Medieval Islamic and Jewish proofs of
creation’, Journal of the American Oriental Society 89, 1969, 357–91.
Endress, G., The Works of Yahyâ ibn ‘Adi, Wiesbaden 1977.
Fakhry, M., ‘The “antinomy” of the eternity of the world in Averroes, Maimonides and
Aquinas’, Le Muséon 66, 1953, 139–55.
Feldman, S., ‘The theory of eternal creation in Hasdai Crescas and some of his
predecessors’, Medieval and Renaissance Studies 11, 1980, 289–320.
Finnegan, J., ‘Al-Farabi et le “Peri Nou” d’Alexandre d’Aphrodise’, Mélanges L. Massignon,
1957, 133–52.
Finnegan, J., ‘Avicenna’s refutation of Porphyrius’, in Avicenna Commemoration Volume,
Calcutta 1956, 187–203.
564 Select Bibliography of Secondary Literature

Frank, R.M., ‘Some textual notes of the oriental versions of Themistius’ paraphrase of
Book 1 of the Metaphysics’, Cahiers de Byrsá 8, 1958–9, 215–30.
Gabrieli, F., ‘Plotino e Porfirio in un eresiografo musulmano’, La Parola del Passato 1,
1946, 338–46.
Gätje, H., ‘Simplikios in der arabischen Überlieferung’, Der Islam 59, 1982, 6–31.
Gätje, H., Studien zur Überlieferung des Aristotelischen Psychologie im Islam, Annales
Universitatis Saraviensis 11, Heidelberg 1971.
Gätje, H., ‘Zur arabischen Überlieferung des Alexander von Aphrodisias’, Zeitschrift der
deutschen morgenländischen Gesellschaft 116, 1966, 255–78.
Gauthier, R., Introduction to Thomas Aquinas Opera Omnia, Leonine Edition 45.1, 1984.
Genequand, C., ‘Quelques aspects de l’idée de la nature d’Aristote à Al-Ghazali’, Revue
de théologie et de philosophie 116, 1984, 105–9.
Ghorab, A.A., ‘The Greek commentators on Aristotle quoted in Al-Almiri’s ‘As-sa ‘âda
Wa’l-is’âd’, in S.M. Stern et al. (eds), Islamic Philosophy and the Classical Tradition.
Essays presented to R. Walzer on his 70th birthday, Oxford 1972, 77–88.
Grant, E., ‘Aristotle, Philoponus, Avempace and Galileo’s Pisan dynamics’, Centaurus 11,
1965, 79–95.
Lyons, M.C., ‘An Arabic translation of the commentary of Themistius’, Bulletin of the
School of Oriental and African Studies 17, 1955, 426–35.
Madkour, I., L’Organon d’Aristote dans le monde arabe, Paris 1969.
Mahdi, M., ‘Alfarabi against Philoponus’, Journal of Near Eastern Studies 26, 1967,
233–60.
Moody, E.A., ‘Galileo and Avempace, the dynamics of the leaning tower experiment’,
Journal of the History of Ideas 13, 1951, 163–93; 375–422.
Peters, F.E., Aristoteles Arabus. The Oriental Translations and Commentaries on the
Aristotelian Corpus, Monographs on Mediterranean Antiquity 2, Leiden 1968.
Pinès, S., ‘Les précurseurs musulmans de la théorie de l’impetus’, Archeion 21, 1938,
298–306; repr. in S. Pinès (ed.), Studies in Arabic Versions of Greek Texts and in
Medieval Science, vol. 2, Leiden-Jerusalem 1986, 409–17.
Pinès, S., ‘Les textes arabes dits plotiniens et le “courant porphyrien” dans le
néoplatonisme grec’, in Le néoplatonisme, Actes du Colloque de Royaumont 9–13
Juin 1969, Paris 1971, 303–17; repr. in S. Pinès, Studies in Arabic Versions of Greek
Texts and in Medieval Science, vol. 2, Leiden-Jerusalem 1986, 264–77.
Pinès, S., ‘The spiritual force permeating the Cosmos according to a passage in the
Treatise on the Principles of the All ascribed to Alexander of Aphrodisias’, in S. Pinès,
Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2, Leiden-
Jerusalem 1986, 252–5.
Pinès, S., ‘Un précurseur bagdadien de la théorie de l’impetus’, Isis 44, 1953, 246–51; repr.
in S. Pinès, Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2,
Leiden-Jerusalem 1986, 418–22.
Pinès, S., Les limites de la Métaphysique selon Al-Farabi, Ibn-Bassa et Maimonides;
sources et antithèses de ces doctrines chez Alexandre d’Aphrodise et chez
 Select Bibliography of Secondary Literature 565

Themistius, Sprache und Erkenntnis im Mittelalter 1, Miscellanea Medievalia 13.1,

Berlin-New York 1981.
Pinès, S., ‘Omne quod movetur necesse est ab aliquo moveri, a refutation of Galen by
Alexander of Aphrodisias and the theory of motion’, Isis 52, 1961, 21–54.
Rescher, N. & Marmura, M., Alexander of Aphrodisias. The Refutation by Alexander of
Aphrodisias of Galen’s Treatise of the Theory of Motion, Islamabad (Islamic Research
Institute), 1965.
Rescher, N., Al-Farabi’s Short Commentary on Aristotle’s Prior Analytics, Pittsburgh 1963.
Sorabji, R., ‘Infinite power impressed, the transformation of Aristotle’s physics and
theology’, Chapter 9 above (fuller version in his Matter, Space and Motion, London-
Ithaca N.Y. 1988, ch. 15).
Steinschneider, M., ‘Johannes Philoponus bei den Arabern’, in ‘Al-Farabi-Alpharabius,
des arabischen Philosophen Leben und Schriften’, Mémoires de l’academie impériale
des sciences de St. Petersbourg, series 7, 13, 4, 1869, 152–76; 220–4; 250–2.
Steinschneider, M., Die arabischen Übersetzungen aus dem Griechischen, Leipzig 1893;
repr. Graz 1960.
Stern, S.M., ‘ “The first in thought is the last in action”, the history of a saying attributed
to Aristotle’, Journal of Semitic Studies 7, 1962, 234–52.
Tardieu, M., ‘Sâbiens coraniques et “Sâbiens” de Harrân’, Journal asiatique 274, 1986,
˙ ˙ ˙
1–44.
Thillet, P., ‘Indices porphyriens dans la théologie d’Aristote’, Le néoplatonisme, Actes du
Colloque de Royaumont 9–13 Juin, 1969, Paris 1971, 293–302.
Ullmann, M., Die Natur- und Geheimwissenschaften in Islam, Leiden 1972.
Walzer, R., ‘Porphyry and the Arabic tradition’, in Porphyry the Philosopher, Entretiens
sur l’antiquité classique 12, Vandoeuvres-Genève 1966, 273–97.
Walzer, R., Greek into Arabic, Oxford 1962.
Wenrich, J.G., De auctorum graecorum versionibus et commentariis syriacis, arabicis,
armenicis, persisque, Leipzig 1842.
Wolfson, H.A., The Philosophy of the Kalam, Cambridge Mass. 1976.
Wolfson, H.A., ‘The problem of the souls of the spheres from the Byzantine
commentaries on Aristotle through the Arabs and St Thomas to Kepler’, Dumbarton
Oaks Papers 16, 1965, 67–93.
Wolfson, H.A., ‘The Kalam arguments for creation in Saadia, Averroes, Maimonides and
St Thomas’, in Saadia Anniversary Volume, American Academy of Jewish Research,
Texts and Studies 2, New York 1943, 197–245.
Wolfson, H.A., Crescas’ Critique of Aristotle. Problems in Aristotle’s Physics in Jewish and
Arabic Philosophy, Cambridge Mass. 1929.
Zimmermann, F., ‘Philoponus’ impetus theory in the Arabic tradition’, in R. Sorabji
(ed.), Philoponus and the Rejection of Aristotelian Science, London-Ithaca N.Y. 1987,
121–9.
Zimmermann, F., Al-Farabi’s Commentary and Short Treatise on Aristotle’s de
Interpretatione, translated with Introduction and Notes, Oxford 1981.
566 Select Bibliography of Secondary Literature

Zimmermann, F. & Brown, H.V.B., Review of A. Badawi, Commentaires sur Aristote

perdus en grec et autres épitres, Beirut 1971, ‘Neue arabischen Übersetzungen
aus dem Bereich der spätantiken griechischen Philosophie’, Der Islam 50, 1973,
313–24.

Middle and Late Byzantium

Benakis, L., ‘Commentaries and commentators on the logical works of Aristotle in
Byzantium’, in R. Claussen & R. Daube-Schackat (eds), Gedankenzeichen. Festschrift
für Klaus Oehler, Tübingen 1988, 3–12.
Benakis, L., ‘Grundbibliographie zum Aristoteles-Studium in Byzanz’, in J. Wiesner (ed.),
Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2, Berlin-New York 1987,
352–79.
Benakis, L., ‘The problem of general concepts in Neoplatonism and Byzantine thought’,
in D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Norfolk 1982, 75–86,
248–9.
Benakis, L., ‘Studien zu den Aristoteles Kommentaren des Michael Psellos 1. Ein
unedierte Kommentar zur Physik des Aristoteles von Michael Psellos 2. Die
aristotelischen Begriffe “Physis”, “Materie”, “Form” nach Michael Psellos’, Archiv für
Geschichte der Philosophie 43, 1961, 215–38; 44, 1962, 33–61.
Browning, R., ‘An unpublished funeral oration on Anna Comnena’, Chapter 17 above,
repr. with revisions from Proceedings of the Cambridge Philological Society 188, 1962,
1–12.
Dräseke, J., ‘Zu Michael Psellos’, Zeitschrift für wissenschaftliche Theologie 32, 1899,
303–30.
Ebbesen, S., ‘Philoponus, “Alexander” and the origins of medieval logic’, Chapter 19
above.
Ebbesen, S., ‘Boethius, Jacobus Veneticus, Michael Ephesius and “Alexander” ’, Cahiers de
l’Institut du moyen-âge grec et latin 34, Copenhagen 1979, v–xiii.
Ebbesen, S., ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and Boethius:
ancient scholasticism and twelfth-century Western Europe. With an edition of the
Tractatus de Paralogismis’, Cahiers de l’Institut du moyen-âge grec et latin 16,
Copenhagen 1976, 1–128.
Ebbesen, S., ‘Ho Psellos kai hoi Sophistikoi Elenchoi’, Byzantina 5, 1973, 429–44.
Hunger, H., Die hochsprachliche profane Literatur der Byzantiner, Munich 1978; cf. esp.
vol. 1, 11–41, Platonismus und Aristotelismus in Byzanz, 54–62; Bibliography,
Handbuch der Altertumswissenschaften 12, Byzantinisches Handbuch 5.
Krumbacher, K., Geshichte der byzantinischen Litteratur von Justinian bis zum Ende des
Oströmischen Reiches (527–1453), Handbuch der klassischen Altertumswissenschaft
9, 1, Fortleben des Aristoteles, 2nd ed. Munich 1897, 430ff.
Lackner, W., ‘Zum Lehrbuch der Physik des Nikephoros Blemmydes’, Byzantinische
Forschungen 4, 1972, 157–69.
 Select Bibliography of Secondary Literature 567

Moraux, P., D’Aristote à Bessarion: trois études sur l’histoire et la transmission de

l’aristotélisme grec, Laval, Quebec 1970.
Oehler, K., ‘Aristotle in Byzantium’, Greek, Roman and Byzantine Studies 5, 1964, 133–46;
German transl. by P. Moraux, Aristoteles in der neuren Forschung, Darmstadt 1968,
381–99; reprinted in K. Oehler, Antike Philosophie und byzantinisches Mittelalter,
Munich 1969, 272–86.
Roueché, Mossman, ‘A middle Byzantine handbook of logical terminology’, Jahrbuch der
Oesterreichischen Byzantinistik 29, 1980, 71–98.
Roueché, Mossman, ‘Byzantine philosophical texts of the seventh century’, Jahrbuch der
Oesterreichischen Byzantinistik 23, 1974, 61–76.
Tarán, L., Review of K. Oehler, Antike Philosophie und byzantinisches Mittelalter,
Gnomon 46, 1974, 534–47.
Tatakis, B.N., ‘La philosophie byzantine’, in E. Brehier, Histoire de la philosophie, suppl. 2,
Paris 1949, 161–210; repr. 1959.
Wolfson, H.A., ‘The problem of the souls of the spheres from the Byzantine
commentaries on Aristotle through the Arabs and St Thomas to Kepler’, Dumbarton
Oaks Papers 16, 1965, 67–93.
Zervos, C., Un philosophe néoplatonicien du XIe siècle, Michael Psellos, sa vie, son oeuvre,
ses luttes philosophiques, son influence, diss., University of Paris, Paris 1920; repr. New
York 1973.

Translation into Latin

(see also Boethius)

Allan, D.J., ‘Medieval versions of Aristotle, de Caelo, and of the Commentary of

Simplicius’, Mediaeval and Renaissance Studies 2, 1950, 82–120.
Bossier, F. & Brams, J. (ed.), Méthodes de traduction et problèmes de chronologie. Recueil
d’articles sur Moerbeke, Louvain 1988.
Bossier, F., ‘Traductions latines et influences du commentaire in de Caelo en Occident
(XIIIe-XIVe siècle)’, in I. Hadot (ed.), Simplicius, sa vie, son oeuvre, sa survie,
Berlin-New York 1987, 289–325.
Courcelle, P., Late Latin Writers and their Greek Sources, Cambridge, Mass. 1969; translated
from Les lettres grecques en Occident, de Macrobe à Cassiodore, 2nd ed., Paris 1948.
Cranz, F.E., ‘Alexander Aphrodisiensis’, in P.O. Kristeller & F.E. Cranz (eds), Catalogus
Translationum et Commentariorum, Medieval and Renaissance Latin Translations
and Commentaries, Washington D.C., vol. 1, 1960, 77–135; vol. 2, 1971, 411–22.
Cranz, F.E., ‘The prefaces to the Greek editions and Latin translations of Alexander of
Aphrodisias, 1450–1575’, Proceedings of the American Philosophical Society 102, 1958,
510–46.
Ebbesen, S., ‘Philoponus, “Alexander” and the origins of medieval logic’, Chapter 19
above.
568 Select Bibliography of Secondary Literature

Ebbesen, S., ‘Analyzing syllogisms, or Anonymous Aurelianensis III – the (presumably)

earliest extant Latin commentary on the Prior Analytics, and its Greek model’,
Cahiers de l’Institut du moyen-âge grec et latin 37, Copenhagen 1981, 1–20.
Ebbesen, S., Commentators and Commentaries on Aristotle’s Sophistici Elenchi. A Study of
Post-Aristotelian and Medieval Writings on Fallacies, 3 vols, Corpus Latinum
Commentariorum in Aristotelem Graecorum 7.1–3, Leiden 1981.
Ebbesen, S., ‘Anonymus Aurelianensis II , Aristotle, Alexander, Porphyry and Boethius;
ancient scholasticism and twelfth-century Western Europe. With an edition of the
Tractatus de Paralogismis’, Cahiers de l’Institut du moyen-âge grec et latin 16,
Copenhagen 1976, 1–128.
Gersh, S., Middle Platonism and Neoplatonism. The Latin Tradition, 2 vols, Notre Dame
1986.
Gilson, E., History of Christian Philosophy in the Middle Ages, London 1955.
Grabmann, M., Guglielmo di Moerbeke O.P., il Traduttore delle Opere di Aristotele (I Papi
del Duecento e l’Aristotelismo, II .), Miscellanea Historiae Pontificae XI .20, Rome
1946.
Grabmann, M., Die Aristoteleskommentare des Heinrich von Brüssel und die Einfluss
Alberts des Grossen auf die mittelalterlichen Aristoteleserklärung, Sitzungsberichte der
bayerischen Akademie der Wissenschaften, Munich 1943.
Grabmann, M., Mittelalterliche lateinische Übersetzungen von Schriften der
Aristoteles-Kommentatoren Johannes Philoponos, Alexander von Aphrodisias und
Themistios, Sitzungsberichte der bayerischen Akademie der Wissenschaften,
Munich 1929.
Hadot, P., Marius Victorinus, Paris 1971.
Hadot, P., Porphyre et Victorinus, Paris 1968.
Isaak, I., Le Peri hermeneias en Occident de Boèce à Saint Thomas. Histoire littéraire d’un
traité d’Aristote, Bibliothèque thomiste 29, Paris 1953; includes the first 10 chapters
of William of Moerbeke’s translation, pp. 160–9.
Jolivet, J., ‘The Arab inheritance’, in P. Dronke (ed.), A History of Twelfth-Century
Philosophy, Cambridge 1988.
Jourdain, A., Recherches critiques sur l’âge et l’origine des traductions latines d’Aristote et
sur des commentateurs grecs ou arabes employés par les docteurs scholastiques, new ed.
revised and expanded by C. Jourdain (from original 1819), Paris 1843.
Kristeller, P.O. & Cranz, F.G. (eds), Catalogus Translationum et Commentariorum,
Medieval and Renaissance Latin Translations and Commentaries, Washington D.C.,
vol. 1, 1960; vol. 2, 1971.
Lohr, C.H., ‘Aristotle in the West; some recent works’, Traditio 25, 1969, 417–31.
McEvoy, J., The Philosophy of Robert Grosseteste, Oxford 1982.
Mercken, H.P.F., ‘The Greek commentaries on Aristotle’s Ethics’, Chapter 18 above,
revised from ‘Introduction’, Corpus Latinum Commentariorum in Aristotelem
Graecorum 6.1.
Minio-Paluello, L., Opuscula, the Latin Aristotle, Amsterdam 1972.
 Select Bibliography of Secondary Literature 569

Minio-Paluello, L., ‘Giacomo Veneto e l’Aristotelismo latino’, in Pertuzi (ed.), Venezia e

l’Oriento fra tardo Medioevo e Rinascimento, Florence 1966, 53–74; repr. in his
Opuscula, the Latin Aristotle, Amsterdam 1972.
Minio-Paluello, L., ‘Jacobus Veneticus Graecus, canonist and translator of Aristotle’,
Traditio 8, 1952, 265–304; repr. in his Opuscula, the Latin Aristotle, Amsterdam 1972.
Moraux, P., D’Aristote à Bessarion: trois études sur l’histoire et la transmission de
l’aristotélisme grec, Laval, Quebec 1970.
Oehler, K., Antike Philosophie und byzantisches Mittelalter, Munich 1969, ch. 13,
‘Aristotle in Byzantium’; repr. from Greek, Roman and Byzantine Studies 5, 1964,
133–46.
Rossi, P., ‘Tracce della versione latina di un commento greco ai Secondi Analitici nel
Commentarius in Posteriorium Analyticorum di Grossatesta’, Rivista di Filosofia
Neoscolastica, 1978, 433–9.
Shiel, J., Aristoteles Latinus III: Scholiorum in Analytica priora supplementa, Bulletin de
la Société internationale pour l’étude de la philosophie médiévale 26, 1984, 119–26.
Shiel, J., ‘The Latin Aristotle’, Medium Aevum 42, 1973, 147–52.
Steenberghen, F. van, Aristotle in the West. The Origins of Latin Aristotelianism,
translation of the Louvain 1955 edition, Aristote en Occident. Les origines de
l’Aristotélisme parisien, by L. Johnston, New York 1970.
Steinschneider, M., Die europäischen Übersetzungen aus dem arabischen bis Mitte des 17
Jahrhunderts; repr. Graz 1956.
Théry, G., Autour du décret de 1210: I. David de Dinant. Étude sur son panthéisme
matérialiste, Bibliothèque thomiste 7, Kain 1926.
Thillet, P., Alexandre d’Aphrodise, de Fato ad Imperatores, version de Guillaume de
Moerbeke, Études de philosophie médiévale 51, Paris 1963.
Verbeke, G., Thémistius, Commentaire sur le traité de l’âme d’Aristote, Corpus Latinum
Commentariorum in Aristotelem Graecorum 1, Paris 1957 (with an introduction).
Verbeke, G. (ed.), Corpus Latinum Commentariorum in Aristotelem Graecorum, the
Medieval Latin translations of the commentaries, Paris 1957–.

The Renaissance
Grant, E., ‘Aristotle, Philoponus, Avempace and Galileo’s Pisan dynamics’, Centaurus 11,
1965, 79–95.
Kristeller, P.O. & Cranz, F.G. (eds), Catalogus Translationum et Commentariorum,
Medieval and Renaissance Latin Translations and Commentaries, Washington D.C.,
vol. 1, 1960; vol. 2, 1971.
Kristeller, P.O., ‘Paduan Averroism and Alexandrism in the light of recent studies’,
Aristotelismo padovano e filosofia aristotelica, Atti del XII congresso internazionale
di filosofia IX , Florence 1960, 147–155; repr. in his Renaissance Thought 2, New York
1965, 111–18.
Lohr, C.H., ‘Comment. d’Aristote au moyen-âge’, Vestigia 2, Fribourg 1988–.
570 Select Bibliography of Secondary Literature

Lohr, C.H., ‘Renaissance Latin Aristotle commentaries’, Studies in the Renaissance 21,
1974; Renaissance Quarterly 28, 1975; 29, 1976; 30, 1977, 681–741; 31, 1978, 532–603;
32, 1979, 529–80; 33, 1980, 623–74; 35, 1982, 164–76.
Lohr, C.H. (ed.), Commentaria in Aristotelem Graeca, versiones Latinae, the Renaissance
Latin Translations Reprinted, Frankfurt 1978–.
Mahoney, E.P., ‘Marcilio Ficino’s influence on Nicoleto Vernia, Augustino Nifo and
Marcantonio Zimara’, in Gian Carlo Giamparini (ed.), Marcilio Ficino e il Ritorno di
Platone, Padua 1986, 509–31.
Mahoney, E.P., ‘Philosophy and science in Nicoleto Vernia and Augustino Nifo’, in
Antonino Poppi (ed.), Scienza e Filosofia all’ Universita di Padova nel Quattrocento,
Padua 1985, 135–202.
Mahoney, E.P., ‘Neoplatonism, the Greek commentators and Renaissance
Aristotelianism’, in D.J. O’Meara (ed.), Neoplatonism and Christian Thought, Norfolk
1982, 169–84, 264–82.
Mahoney, E.P., ‘Nicoletto Vernia on the soul and immortality’, in E.P. Mahoney (ed.),
Renaissance Essays in Honour of P.O. Kristeller, Leiden 1976, 144–53.
Mahoney, E.P., ‘Themistius and the agent intellect in James of Viterbo and other
thirteenth-century philosophers’, Augustiniana 23, 1973, 422–67.
Moody, E.A., ‘Galileo and Avempace, the dynamics of the leaning tower experiment’,
Journal of the History of Ideas 13, 1951, 163–93; 375–422.
Nardi, B., ‘Alessandrismo’, Enciclopedia Cattolica, Vatican City 1968, vol. 1, 778f.
Nardi, B., ‘Il commento di Simplicio al de Anima nelle controversie della fine del secolo
XVe al secolo XVI ’, Archivio di Filosofia, 1951, 139–206; repr. in his Saggi
sull’Aristotelismo Padovano dal Secolo XIV al XVI , Florence, 1958, 365–442.
Schmitt, C.B., ‘Philoponus’ commentary on Aristotle’s Physics in the sixteenth century’,
in R. Sorabji (ed.), Philoponus and the Rejection of Aristotelian Science, London-
Ithaca N.Y. 1987, 210–27.
Schmitt, C.B., Aristotle and the Renaissance, Cambridge, Mass.-London 1983.
Schmitt, C.B., Gianfrancesco Pico della Mirandola (1469–1533) and his Critique of
Aristotle, The Hague 1967.

Manuscripts
Argyropoulos, R.D. & Caras, I., Inventaire des manuscrits grecs d’Aristote et de ses
commentateurs. Contribution à l’histoire du texte d’Aristote. Supplement, Paris 1980.
Canart, P., ‘Manuscrits d’Aristote et de ses commentateurs sur papier occidental ancien’,
in J. Wiesner (ed.), Aristoteles: Werk und Wirkung, Paul Moraux gewidmet, vol. 2,
Berlin-New York 1987, 418–33.
Harlfinger, D. & Wiesner, J., ‘Die griechischen Handschriften des Aristoteles und seiner
Kommentatoren’, Scriptorium 18, 1964, 238–57.
Hoffmann, P., Review of R.D. Argyropoulos and I. Caras, Inventaire des manuscrits grecs
d’Aristote, Revue des études anciennes 83, 1981, 352–6.
 Select Bibliography of Secondary Literature 571

Huby, P.M., ‘The transmission of Aristotle’s writings and the places where copies of his
works existed’, Classica et Mediaevalia 30, 1969, 241–57.
Minio-Paluello, L. (ed.), Aristoteles Latinus. Codices. Supplementa, Paris 1961.
Mioni, E., ‘Manoscritti Graeci-Veneti di Aristotele e Commentatori’, in C. Mango,
Catalogo Manoscritti e stampe venete dell’Aristotelismo e Averroismo, Venice 1958.
Moraux, P. et al., Aristoteles Graecus I-, Berlin 1976–.
Thillet, P., ‘Les MSS grecs d’Aristote et de ses commentateurs’, Bulletin de l’association
Guillaume Budé, 1963, 351–5.
Wartelle, A., Inventaire des manuscrits grecs d’Aristote et de ses commentateurs, Paris
1963. Supplement by R.D. Argyropoulos & I. Caras, Paris 1980. Corrections by
D. Harlfinger & J. Wiesner, ‘Die griechischen Handschriften des Aristoteles und
seiner Kommentatoren’, Scriptorium 18, 1963, 238–57.
572
 Index Locorum

This index is to the commentaries on Aristotle included in CAG plus the works by Boethius
and ps.-Elias listed below after the abbreviations.
The following abbreviations are used in this index:

in An. Post. in Analytica Posteriora

in An. Pr. in Analytica Priora
in Cael. in de Caelo
in Cat. in Categoriae
in DA in de Anima
in EN in Ethica Nicomachea
in GA in de Generatione Animalium
in GC in de Generatione et Corruptione
in HA in Historia Animalium
in IA in de Incessu Animalium
in Int. in de Interpretatione
in Isag. in Porphyrii Isagogen
in Metaph. in Metaphysica
in Meteor. in Meteorologica
in MA in de Motu Animalium
in PA in de Partibus Animalium
in Phys. in Physica
in SE in de Sophisticis Elenchis
in Sens. in de Sensu et Sensibilia
in Top. in Topica
Proleg. Prolegomena

Full references for these works, for individual authors, are given on pp. 30–2 of this volume,
with the exception of the following:

Boethius, in Cat.: Patrologia Latina (Migne) 64, 159–294.

Boethius, in Int.1 (=ed. prima, in vol. 1) and in Int.2 (=ed. secunda, in vol. 2):
Boethii commentarii in librum Aristotelis peri Hermeneias, Meiser (ed.), Teubner
1877–80.
Boethius, in Isag.1 (=ed. prima) and in Isag.2 (=ed. secunda): Anicii Manlii Severini Boethii
In Isagogen Porphyrii commenta, S. Brandt (ed.), Vienna and Leipzig 1906 (Corpus
Scriptorum Ecclesiasticorum Latinorum 48).
ps.-Elias, in Isag.: Pseudo-Elias (Pseudo-David), Lectures on Porphyry’s Isagoge,
L.G. Westerink (ed.), Amsterdam 1967.

References to the pages of this volume appear in bold type.

573
574 Index Locorum

Alexander in Cat., 56, 77, 232, 256–7, 297, 313,

in An. Pr., 93–5, 102; 1,1–6,12, 102; 350, 367, 371, 387; 4,28ff., 73;
4,3–14, 509; 8,2ff., 52; 9,5, 52; 42,23, 6,2ff., 46; 6,9–16, 6; 6,10–12, 232;
52; 72,27, 93; 89,34, 93; 91,2, 93; 6,10–16, 247; 6,17–20, 246; 7,7–14,
125,3–6, 93; 125,16, 93; 125,30ff., 6; 8,6, 384; 11, 159; 25,12, 384;
92; 156,11–157,10, 103; 160,32, 26,2–3, 233; 29,9–10, 265;
102; 161,1, 95; 161,3–26, 103; 35,18–36,4, 249, 255; 35,20–1, 233;
177,19–182,8, 99, 102; 183, 36,25–6, 232; 37,1–10, 233;
34–184,18, 99, 102; 191,19ff., 110; 37,17–18, 244; 49,6–11, 167;
191,24, 110; 192,4, 110; 193,13, 58,16–17, 265; 66,6, 51; 80,15, 51;
110; 193,20, 111; 218,7ff., 111; 88,20–2, 409; 106, 384
218,13, 111; 218,20, 111; 220,3ff., in Int., 229–32, 297, 350, 385; 1,3–11,
111; 250,2, 91 298; 1,6, 384; 1,6–11, 351; 1,11,
in DA 16,18–17,8, 107; 24,21–3, 109; 229; 5, 160; 5,15, 160; 5,24, 66;
40,4–10, 107; 90,6–7, 109 5,28ff., 62, 66, 73; 7,15, 386; 8,19,
in Metaph., 102; 52,13–16, 506; 405; 12–13, 169; 14,11, 405;
52,15–19, 506; 58,31–59,8, 96; 18,2–12, 182; 18,31, 160; 21,4ff.,
59,28ff., 98; 79,3–98,24, 96; 182; 21,21ff., 170; 24,24–9, 230;
136,13–138,23, 102; 138,6ff., 24,29–30, 175; 32,4, 388; 34ff., 420;
102; 141,21ff., 103; 152,6ff., 107; 39,13–17, 104; 39,17–32, 104; 44,11,
159,9, 103; 162,6–10, 103; 164,24, 388; 47,26, 388; 50,10, 388; 51,20–2,
103; 165,4, 103; 165,21, 103; 170; 52, 172; 55–7, 169; 56,16, 388;
166,19–167,3, 95; 167,5ff., 103; 62, 184; 70,4, 388; 72–5, 177, 182;
168,7ff, 103; 169,11, 103; 170, 73,8–11, 178; 73,11–14, 183; 84–6,
5–171,22, 102; 172,21–2, 102; 182; 84,30–85,3, 183; 94,29, 388;
196,19–26, 102; 199,19–22, 507; 126,11–14, 178; 128,3–14, 183;
199,35–9, 508; 201,4–11, 506; 128,4–14, 184; 130,27–138,10, 230;
208,31ff., 107; 210,6–8, 107; 132,14–16, 230–231; 132,15, 231;
220,24ff, 103; 289,29–33, 102; 133,19–23, 230; 135,28–32,
337,29, 103; 524,9–12, 165 230–231; 136,15–17, 230; 136,20–4,
in Meteor., 102; 1,1–4,11, 102; 14,25ff., 230; 181,30–1, 190; 181,30–2, 351;
105; 66–89, 106; 143,13, 90, 92 213,1–5, 230; 213,8–10, 225, 230,
in Sens., 23–4, 104; 2,1, 52; 10,2, 94; 237; 213,33f., 405; 217,28–9, 229;
31,29, 91; 82,16–17, 94; 111,17–19, 240,10–11, 229; 243,2–4, 230, 231;
506; 164,13ff., 104 248,7–249,26, 231; 248,17–18, 231;
in Top. 1,3, 52; 26,24, 52; 74,26ff., 73; 248,17–31, 255, 286; 248,18, 231;
93,20ff., 72; 94,8ff., 73; 99,2–20, 248,28, 231; 248,31, 231; 249,17,
94; 100,31–101,14, 94; 125,4–8, 230; 250,12–19, 229; 250,34–251,1,
52; 133,24, 111; 133,26, 111; 255, 286; 253,12–17, 191; 259,9,
181,21–187,9, 94; 181,21ff., 101; 388
187,9–10, 92; 242,4, 112; 250,26, in Isag., 47, 59, 232, 350, 370; 1,2, 47;
111; 434,2, 95; 507,10ff., 112; 569,3, 1,4, 48; 1,4ff., 50; 1,13ff., 48, 50;
93; 574,26, 93 2,28, 47; 3,9–15, 232, 362, 375;
Ammonius 3,9ff., 218; 3,21ff., 50; 3,22–3, 375;
in An. Pr., 99, 232, 256–7, 298, 350, 4,7–8, 232; 4,22–5, 375; 5,31ff., 50;
371; 1,8–9, 160, 162; 5,19–22, 233; 9,7–23, 420; 9,8–15, 375; 9,16–17,
25,11–20, 233; 25,20, 232; 31,11, 66; 232; 9,16f., 218; 9,21, 218; 11,11,
31,15–22, 131; 38,40–39,2, 330; 218; 11,11–13, 232; 11,22–13,7,
39,24–6, 92; 39,31, 93; 50,13–21, 99; 504; 11,31–12,6, 504; 12,19, 47;
67,32–69,28, 350; 68,4–6, 159 12,25–7, 375, 509; 13,10, 47;
 Index Locorum 575

13,24–7, 362, 375; 15,17–16,3, 375; 92,29–30, 221; 98,5–15, 244;

16,17–19, 375; 21, 403; 21,6ff., 50; 98,7–8, 509; 99,5, 221; 99,5–6, 223;
21,8ff., 51; 22,23, 47; 23,12, 392; 100,3, 223; 102,17, 221; 103,3–4,
23,18, 51; 24,2–5, 232; 26,5, 51; 236; 103,9–12, 247; 104,13–15, 509;
39,14ff., 505; 41–2, 165; 41,1–2, 105,30–2, 236; 105,31–2, 221;
265; 41,13–20, 505; 41,20–42,26, 105,32, 222; 106,22, 223; 106,23,
232; 42,6–16, 505; 44–5, 157; 222; 106,33–4, 221; 106,33–107,4,
63,14–16, 232; 101–5, 157; 221; 107,12, 222, 223; 108,23–5,
110,22, 265 222; 108,24, 226; 108,25, 221;
Anon. 113,1–2, 221; 115,37, 221, 222;
in EN 7, 23, 440, 473; 133,35–134,4, 119,33–4, 226; 120,10–11, 221;
94; 134,9–21, 94; 141,10–20, 121,7–8, 221; 121,8–9, 223;
94 122,25–6, 226; 123,15, 222; 123,16,
in Int., 20 222; 123,16–17, 221; 128,21–2, 221;
Asclepius 137,2–3, 221; 140,15–16, 221;
in Metaph., 43, 218, 221–7, 229, 231, 141,22, 221; 142,36–7, 351;
244–5, 257, 296, 350, 352; 1,2–3, 142,38–143,1, 225; 144,28, 221;
221; 11,35–6, 226; 15,8, 221; 17,7, 144,28–30, 222; 144,30–1, 226;
221; 17,29–18,1, 221; 19,34–5, 144,30–2, 222; 145,15, 221;
221; 20,25–6, 226; 20,25–8, 227; 145,17–18, 225; 146,5, 221; 146,6,
20,27–8, 222; 21,18–21, 247; 21,19, 221; 146,7–8, 222; 147 passim, 222;
222; 23,3, 221; 26,6, 222; 28,6, 222, 147,4, 221; 147,4–6, 223; 147,6–11,
226; 28,20–2, 236; 28,26–7, 236; 226; 147,7, 221; 147,8–9, 222;
28,31–2, 236; 33,34–5, 222–223; 147,28, 221; 147,29–30, 221;
34,28–30, 223; 35 passim, 223; 147,31–2, 244; 147,32, 226;
35,2–16, 223; 35,10–11, 221; 35,15, 148,10–11, 221, 236; 148,13–5, 222;
221; 35,15–16, 222; 36,22, 223; 148,24, 221; 148,24–7, 222; 148,27,
38,11–13, 223; 38,13–14, 223; 222; 148,27–8, 221; 148,28–32, 222;
38,13–16, 223; 38,18, 223; 38,18–19, 148,29–31, 238; 148,30–4, 226;
221; 38,21–4, 223; 38,21–9, 223; 148,33, 221; 148,33–4, 224; 151,1,
38,22–4, 223; 39,7–9, 223; 39,8–9, 221; 151,4–6, 509; 151,8–12, 226;
223; 39,17, 221; 41,31–3, 226; 151,9–11, 227; 151,15–32, 236;
42,14, 221; 42,14–15, 224, 247; 151,18–19, 222; 151,19, 222;
44,32–5, 247; 44,32–7, 237; 151,22–32, 221; 151,24–7, 222;
44,35–7, 237; 47,9, 221; 48,5, 223; 151,25, 222; 151,26–7, 222; 151,29,
48,5–6, 221, 225; 48,5–7, 225; 221; 158,18, 221; 158,18–23, 224;
52,1–2, 221; 52,6, 223; 52,6–7, 225; 158,27–8, 226; 159,16, 221; 163,11,
52,7, 221, 222; 52,21–4, 247; 221; 165,35–166,1, 225; 166,29–31,
54,25–6, 222, 225, 247; 55,25–6, 225; 166,31–4, 247; 167,14–34, 237;
247; 56,20, 221; 56,20–1, 222; 167,24–6, 247; 167,29–31, 237;
56,24, 221; 57,8, 226; 57,9, 226; 167,31, 238; 167,32, 225; 168,31,
60,23–4, 221; 69,17–21, 4; 69,17–27, 221; 168,34–5, 225; 171,9–11, 352;
237; 69,19–20, 237; 69,20, 238; 173,22, 225; 173,22–3, 225;
69,20–1, 238; 70,2–4, 225; 70,26–9, 173,31–3, 225; 175,16, 225;
227; 71,28, 4; 73,21–3, 225; 75,27–8, 175,16–18, 225; 176,12, 222; 178,1,
237; 75,32–3, 225; 76 passim, 225; 223; 181,32, 221; 183,14–16, 237;
77,8–17, 226; 77,16–17, 224; 183,21–3, 225; 183,22–3, 221;
80,30–1, 225; 81,2–4, 225–226, 238; 186,1–2, 352; 186,2, 221; 189,12,
88,4–7, 225, 227; 89,4–5, 352; 221; 189,27, 165; 190,5, 221; 191,7,
89,21–2, 225; 90,27–8, 352; 221; 191,7–11, 222; 191,8, 221;
576 Index Locorum

191,11–13, 222; 191,12–13, 222; Aspasius

194,13, 221; 194,13–17, 222; 194,14, in EN, 18, 23, 75, 91, 431, 439–42,
221; 194,14–15, 223; 194,23–6, 474; 2,16–3,2, 476; 5,23–30, 476;
352; 195,26, 221; 195,26–7, 226; 10,29–32, 476; 32,10, 112; 44,20–4,
195,37–196,10, 227; 196,6–7, 222; 476; 59,2–11, 94; 156,14–16, 476;
198,26–7, 223; 198,30, 221; 201,4, 180,8, 477
224; 201,4–12, 224; 201,9, 223;
201,9–10, 222; 201,11–12, 224; Boethius
201,23, 223; 201,25–6, 223; 201,26, in Cat., 297, 378–85, 398, 407, 494;
222; 201,40–202,9, 223; 202 passim, 159A, 407–8, 412; 159A–163C,
223; 202,8–9, 223; 202,11, 222; 382; 159B–C, 414; 160A, 407;
202,11–12, 223; 202,12, 222; 202,19, 160A–B, 22, 408, 418; 160B, 382,
222; 202,19–22, 223; 202,22, 222; 399; 161A–162A, 21; 161B–D,
202,22–3, 223; 202,23–5, 223; 408; 162A, 5, 382, 417–18; 162D,
202,27–32, 225; 202,31–2, 224; 410; 163B–C, 409; 168C, 382;
206,18–19, 222; 206,18–20, 223; 175B, 384; 176D–177A, 413;
206,19, 222; 207,18–19, 223; 180C, 381, 407, 418; 181D, 381;
207,19–20, 223; 207,23–7, 223; 183–4, 410; 183D, 412; 184A, 379;
208,18–22, 223; 208,21, 223; 184A–B, 381; 185D, 410; 190A,
208,21–2, 222; 209,35–7, 225; 379; 192C, 379; 201B, 401;
216,23, 222, 223; 216,24–6, 225; 201D–202B, 407; 202B–C, 416;
217,2–4, 222; 217,3–4, 222; 202C, 409; 206D, 381; 208B, 381;
217,3–5, 222; 217,4–5, 222; 216D, 379, 416; 217A, 409; 224C,
218,3–5, 225; 222,2–3, 221; 380–1; 224C–225B, 382; 224D, 417;
223,33–6, 224; 223,35, 222; 225B, 417; 228B, 379; 230C, 381;
223,35–6, 224; 223,36, 223; 233B–D, 417; 239A, 409, 416;
225,15–17, 222, 224, 236; 225,16, 240C–D, 407; 243C, 410; 250C,
221–2; 225,22, 224; 225,34–226,3, 408; 252B, 409; 252B–C, 22; 252C,
224; 226 passim, 222; 226,2–3, 226; 409; 263B, 391; 264B, 381; 264C,
226,2–4, 224; 226,3, 224, 226; 379; 264D, 379; 266A–C, 381;
226,6, 224; 226,6–8, 224; 226,7, 270A–B, 381; 277D–278D, 381;
222; 226,12–5, 352; 226,13, 222; 280B–D, 381; 283C, 379; 288D,
226,16–17, 224; 226,17–18, 224; 381; 289C, 408, 409; 291B–C, 384;
227,2, 224; 227,2–4, 224; 227,3–4, 294A–B, 409
224, 226; 227,12, 224; 227,37, 222; in Int.1, 385–91; 32,3, 390; 32,7, 389;
230,4, 222, 224; 232,9–10, 224; 46, 412
232,18–25, 224; 233,25–31, 236; in Int.2, 385–91, 399; 1–13, 21; 3,7,
233,29–31, 225; 233,38–40, 237; 397; 4,4, 397; 6, 169; 7, 408; 7–13,
238,5, 224; 249,3, 222; 249,3–4, 418; 7,4, 387; 7,5, 391; 10–11, 160;
222; 301,15, 221; 309,10–15, 227; 10,14, 103; 11,13–15, 102; 12, 408;
309,12–13, 226; 309,14–15, 224; 13, 405; 13,9–11, 418; 18,26, 388,
309,15–18, 224, 227; 309,24–6, 227; 391; 19,20, 391; 21–3, 420; 22,6–13,
363,1–5, 237; 368,13–15, 227; 420; 23, 405; 24, 415; 26–36, 160;
393,34–394,2, 237; 432,29–436,8, 26,17, 184; 27,14, 399; 28, 175, 184;
296; 433,9–436,6, 4; 433,16, 4; 28,18ff., 162; 30, 162; 39–40, 175;
439,26, 224; 439,26–7, 221, 223; 41, 172; 49, 182; 54–6, 405; 55,1–7,
441,27–31, 237; 442,1–2, 237; 412; 64, 412; 65,20, 388; 66,18, 388;
447,25, 165; 450,22, 4; 450,25–7, 66,31, 388; 68,18, 388; 71, 415;
221 75,10, 388; 77, 169; 79–80, 404;
 Index Locorum 577

79,9–80,1, 21; 79,16, 398, 400; 142,22ff., 457; 143,1–3, 461;

79,16ff., 57, 400; 80,1–6, 22; 81–2, 143,26ff., 458; 150,2–4, 459;
183; 87, 382; 87,15–16, 387; 92–4, 153,19–154,8, 458; 154,15–16, 459;
405; 105,15, 388; 106ff., 177; 155,1–17, 459; 155,4–8, 457;
134,20ff., 184; 143–4, 387; 143,1, 168,5ff., 460; 171,23–8, 458; 176–7,
388; 160,5, 388; 182,17–20, 182; 460; 178,30, 460; 180,12–15, 461;
186,2, 390; 190,12ff., 396; 201, 415; 180,16–17, 460; 180,24–5, 460;
201,2, 389; 207,21, 400; 234,10–13, 182,6–13, 462; 182,25–9, 462;
400; 235,6–8, 400; 240,10, 400; 186,10–20, 462; 189,13, 462;
250,20, 390; 250,20–3, 406; 251, 200,8–18, 462; 202,18, 463; 203,8,
404; 251,1, 389–90; 251,8, 394; 463; 203,14, 463; 211,6, 463; 221,24,
272,14ff., 93; 274,12, 93; 293,27, 463; 229,22, 453; 234,7–18, 463;
387; 321,20, 388; 321,20ff., 388; 234,27–34, 463; 240,29–30, 463;
343,7, 398; 361,9, 386; 421,3, 390; 244,36, 463; 246, 463; 246,11, 463;
428,12, 388; 430,24, 388; 438,10, 246,13, 463; 248,24–6, 454; 254,12,
389; 446,11, 389; 449,14, 389; 463; 254,28–255,1, 463
479,6, 389 Dexippus
in Isag.1, 55, 57, 391–3, 398, 404, 407; in Cat., 19, 55, 77, 133, 168, 297, 337,
3,1–16,10, 392; 4,17–5,10, 21, 371; 382; 5,26–30, 52; 6,14ff., 160;
7, 420; 7,11–12,9, 21; 7,11–16, 21; 6,18–20, 160; 7–8, 172; 7–8,10, 172;
8,1, 392; 8,6–19, 22; 8,11, 391; 7–10, 160, 182; 7–16, 159; 7,1–2,
9,21, 392; 10, 404; 10,19–22, 371; 175; 7,30–8,1, 182; 8,28–9,1, 160;
12–15, 408; 12,20, 393; 14,17, 51; 9,14, 160; 9,14–15, 177; 9,16–20,
23,17, 391; 24ff., 411; 34,12, 391; 160; 10,3–6, 181; 10,3–10, 175;
42,14, 393 10,20, 162; 11–16, 159; 11,14–15,
in Isag.2, 391–3, 398, 411; 25, 404; 159; 11,28, 159; 13–16, 174; 15,27,
93, 404; 135, 405, 406; 135, 163; 16,33–17,3, 420; 17,1, 2;
5–143,7, 21; 135,7, 393, 398; 17,1ff., 494; 19, 176; 19,20–2, 159;
140,13ff., 392; 159ff., 411; 20,3–4, 160; 21,18ff., 66; 26, 174,
161, 408; 164, 163; 166, 163; 413, 494; 26,9, 165; 32–3, 169;
167, 408; 168, 157; 168,11, 393; 40,13–42,3, 133; 40,14–18, 134;
319,15–325,7, 393 40,19–25, 2, 134; 40,25–41,3, 135;
40,28–41,3, 196; 41,4, 140; 41,4–6,
David 140; 41,7–18, 141, 196; 41,9–12,
in Isag., 59, 364; 78,28–79,1, 375; 142; 41,18–42,3, 2, 144–145;
80,14, 51; 93,25, 51; 95,10, 51; 41,18ff., 174; 42,2, 139; 44, 2; 44,11,
102,4ff., 76; 115,12–13, 15; 139; 44,15, 139; 44,25, 139;
129,8–10, 15, 365 45,12–31, 108; 45,22, 165; 45,27ff.,
Proleg., 370; 9,2–5, 362; 15,21–26, 365; 83; 51,13, 165
17,2–9, 362, 375; 24,11–12, 365;
24,14–15, 365; 31,34, 44; 32,30–3, Elias
375; 34,21–2, 375; 45,7–9, 375; in Cat., 56, 297, 361, 367; 107–134,
59,17–19, 375, 509 363; 107,8, 48; 107,11ff., 64, 67;
Derivative of Adrastus 107,24, 383; 107,24–6, 6, 48, 351;
in EN 2–5, 18, 23, 440, 453, 454; 122,9, 109,10–14, 363; 110,18–20, 363;
456; 125,16, 456; 126,21, 456; 127,5, 113,17ff., 64; 115,14ff., 73; 117,17ff.,
457; 127,7, 457; 128,32, 456; 46; 117,22ff., 73; 117,24, 65;
129,1–3, 456; 130,25, 459; 136,28, 120,16–17, 364; 120,23–30, 240;
457; 137,8, 457; 138,24, 456; 120,30–121,3, 231; 121,5–19, 246;
578 Index Locorum

123,1–3, 3–4, 329; 123,1–7, 6; 470; 616,35, 470; 619,15, 470;

123,2–3, 242; 123,7–9, 334; 123,9, 619,19, 470; 620,4, 470
46; 123,10ff., 329; 124,21–3, 231; in IA, 23, 58, 74–5, 468
129,1, 51; 130,11–13, 160; 130,28, in MA, 23, 58, 74–5, 468
363; 154,17–19, 165; 187,6–7, 364; in PA, 23, 58, 74–5, 468
236,6, 363 in PN, 23, 57, 74–5, 468; 142,5,
in Isag., 47, 59, 361, 370; 35,7, 51; 467; 149, 74–5; 149,8–16, 468;
39,12–14, 362; 40,8, 51; 69,22, 364 149,16, 58
Proleg. 1–39, 363; 1,19–2,2, 375; in SE, 22–3, 57
3,17–20, 362; 6,7–9, 361; 6,32ff.,
59; 7,3, 364; 12,1, 364; 13,26, 365; Olympiodorus
14,4–7, 375; 14,8, 44; 16,20–5, in Cat., 40, 297, 313, 353, 367, 384;
362, 375; 17,2, 232; 17,7–8, 375; 35,4, 382; 48,14, 382
20,10–11, 352; 20,19–20, 191; 21,4, in Meteor., 350, 353, 358; 2,11–12, 363;
365; 22,19–23, 362; 22,25–6, 362; 52,30–53,2, 353; 55,9, 363;
23,34–24,3, 375; 24,21ff., 59; 27,16, 115,11–13, 358; 118,5–119,8, 358;
362; 31,11–13, 362, 375; 34,10–21, 120,13–14, 358; 120,20–1, 358;
375 141–50, 358; 147,24–148,6, 358;
Eustratius 153,7, 352; 153,19–21, 358;
in An. Post. 2, 446–7; 195,5ff., 451; 301,18–19, 365
196,14–16, 451 Proleg. 1,7, 47; 1,15, 50; 1,16, 48; 1,23,
in EN 1, 6, 23, 25–6, 430–1, 440, 443, 47; 3,8ff., 48, 50; 3ff., 47; 7,24ff., 73;
446, 452, 477; 1,13–23, 448; 4,9, 8,29ff., 46, 50; 9,27–10,2, 509;
448; 4,25–8, 450; 4,25–38, 450; 9,31–10,2, 246; 12,18, 47; 13,24ff.,
6,5–8, 450; 7,9–12, 449; 34,19–31, 76; 17,20, 45; 19,20–4, 160
450; 34,23–8, 449; 35,33–7, 449;
39,24ff., 450; 40,22ff., 451; Philoponus (?)
40,37–41,2, 451; 46,29–31, 451; in An. Post. 2, 421,3–4, 159
63,29–33, 450; 106,19–32, 450; in DA 3, 53, 130, 197, 334, 340, 346,
111,21ff., 447; 272,3–277,17, 449; 365; 448,6–7, 365; 450,8ff., 130;
298,30–31, 447; 400,22–401,20, 452 450,27–33, 130; 457,24–5, 366;
in EN, 6 470,18, 102; 490,18ff., 341;
508,19–25, 124; 508,19ff., 130;
Heliodorus (?), in EN, 22, 26, 42 527,29–32, 365; 535,2ff., 126;
535,5f., 342; 535,8–10, 335;
Michael of Ephesus 535,12f., 335; 535,13–15, 338;
in EN, 23, 25–6, 37, 74–5, 430–1 535,15f., 338; 536,2–5, 125, 338,
in EN 5, 440, 464, 470; 25,28–9, 470; 341; 539,1ff., 338; 540,27, 365;
47,17–18, 470; 66,5–9, 470; 72,5–6, 541,20–542,5, 365; 547,11–14,
470 365; 595,33–598,7, 365
in EN 9–10, 58, 440, 442, 464, 470; Philoponus
483,3,, 470; 515,30, 470; 524,17, in An. Post., 8, 14, 24, 130, 243, 260–2,
470; 529,21, 470; 533,22, 470; 268, 276, 280, 284–5, 350, 481,
533,35, 470; 537,13, 470–1; 542,27, 485, 496; 1–2, 408; 1,3, 262; 3,2–4,
471; 542,27ff., 471; 570,21, 464; 268; 47,24–48,11, 257; 67,17–18,
578,16, 470; 578,19, 58; 579,5, 470; 276; 109–10, 486; 110,14–15, 276;
579,23, 470; 579,30, 470; 585,10, 111,31–112,36, 351; 133,24, 165;
470; 585,13, 470; 587,20, 58; 133,30–134,2, 276; 135,11–15, 276;
598,19ff., 45; 598,23, 470; 601,4, 158,7–11, 257; 171,13ff., 485;
 Index Locorum 579

174,28–9, 485; 177,19–29, 257; 116,1–21, 294; 119,22–4, 254;

179,10–11, 272; 181,11ff., 506; 126,26–7, 286; 132,30–1, 286;
215,4–5, 245; 242,14–243,25, 4, 132,32–133,3, 274; 133,1–2, 274;
243, 255, 276; 242,26–243,13, 243; 138,21–3, 287; 141,2–4, 274;
242,26–243,25, 199; 243,7, 243; 164,11–13, 333; 165,4ff., 347;
243,17–24, 276; 243,20–1, 243; 165,5–9, 342; 165,8–13, 341;
244,18–25, 276; 258,26–30, 257; 195,2ff., 331; 198,22–9, 345;
260,29–261,8, 257; 265,5–6, 272 198,22ff., 123; 202,22ff., 331;
in An. Pr., 8, 257, 262, 268, 284–5, 215,4–216,26, 346; 216,33–5, 286;
350, 481; 29,2, 351; 31,24, 351; 217,9–15, 346; 224,10–225,31, 347;
40,30–1, 351; 126,20–3, 92; 224,28ff., 347; 225,20–31, 347;
165,27–167,30, 99; 169,17–23, 99; 228,16–17, 274; 259,9–10, 274;
243,1–4, 159; 273,4, 160 260,18–25, 255; 265,30–4, 274;
in Cat., 40, 257, 262, 268, 276, 284–5, 268,6–9, 274; 268,37–269,1, 274;
297, 350, 367, 384, 387, 481; 1,7, 269,26–7, 274; 270,2–4, 274;
48; 4,23ff., 73; 5,15ff., 46; 5,16f., 270,29–30, 274; 272,31–2, 274;
73; 5,18ff., 65, 73; 6,3–16, 246; 279,11–12, 274; 286,19–21, 274;
6,15–16, 509; 7,20ff., 76; 8–9, 160; 286,32–4, 274; 297,1–10, 294;
8,7–8, 7; 9,14–15, 162; 9,29–31, 297,10–11, 286; 324,15–16, 274;
160; 12,4–9, 162; 17,2–10, 224; 348,10–19, 92; 359,10, 275;
31,7ff., 329; 32,20–2, 276; 33,22–6, 407,18–422,10, 9; 418,24f., 130;
276; 49,23–51,21, 255; 50,23–31, 422,11–436,33, 9; 433,15ff., 130
276; 67,7–10, 255; 72, 167; in GC, 8, 260–2, 268, 273–6, 278–9,
85,8–86,1, 513; 86,23–7, 277; 284–5, 350; 1,3, 262; 1,9–12, 274;
87,7–10, 276; 99,29–30, 276; 3,1–4, 274; 4,17, 275; 12,6, 101;
102,16, 51; 126,29–31, 352; 37,19–31, 242–243; 37,20–6, 243;
145,8–11, 254; 145,10–146,2, 37,26–8, 243; 37,28–31, 243;
276; 156,8–11, 409; 165,1–2, 45,5–6, 274; 49,12–13, 274; 49,23,
276; 169,19, 352; 205,26–8, 409 274; 50,1–5, 234, 242; 73,8, 275;
in DA, 8–9, 14, 130, 259, 262, 268, 73,28–30, 275; 77,8–15, 506;
273–6, 279, 284–5, 288–9, 328, 116,1–121,10, 249; 129,6–14, 274;
350; 1–3, 332; 1,1ff., 329, 340; 132,12–13, 275; 132,19, 275;
1,3, 262; 3,8–9, 514; 3,8–13, 509; 134,1–4, 275; 134,4–5, 274;
3,12–13, 509; 5,38–6,1, 513; 135,1–2, 274; 135,23–4, 274;
7,12–19, 274; 9,6–7, 274; 9,14–15, 136,33–137,3, 199, 242, 255;
274; 12,14f., 332; 12,15–22, 332; 139,20–1, 274; 144,21–2, 274;
15,65–9, 95; 16,10–26, 253; 147,19–25, 274; 152,23–153,2, 242,
17,19–23, 333; 18,22–33, 333; 18,27, 255; 156,20–5, 275; 157,5–6, 275;
274; 21,1–2, 274; 21,20–3, 6, 339; 178,21–2, 275; 199,4–6, 274;
24,27–9, 274; 27,21f., 62; 27,21ff., 206,2–4, 274; 280,21–4, 275;
66; 35,1–2, 255, 286; 37,18–31, 4, 282,23–7, 274; 283,1–10, 274;
199; 37,19–31, 255; 37,28f., 341; 286,7, 199; 288,19–289,22, 274;
45,53–9, 338; 55,4–7, 338; 55,23ff., 290,12–13, 274; 290,24–5, 274;
503; 56,2–4, 274; 57,28–58,6, 504; 291,21–3, 274; 296,14–17, 274;
58,6–13, 512; 63,12–14, 286; 296,14–298,8, 274; 296,19–21, 254;
78,24–6, 513; 94,18, 286; 100,6–16, 297,15–24, 242; 299,18–19, 274;
275; 101,34–102,31, 287; 102,1, 299,22, 274; 299,25, 274; 300,1–3,
287–8; 102,2, 288; 102,17, 287; 274; 300,7–8, 274; 301,28–9, 275;
102,18, 288; 115,22–121,10, 285; 312,17–18, 274
580 Index Locorum

in Meteor., 14, 249, 252, 260–2, 277, 264; 467,1–468,7, 268; 484,15–19,
284–5, 350, 352; 8,13, 260; 9,8, 260; 264; 490,14, 254; 497,8–9, 264;
11,24–37, 277; 12,24–32, 277; 499,26, 265; 499,26–7, 265;
12,30–1, 277; 15,23, 260; 22,8, 260; 500,28–9, 266; 526,20–3, 265; 536,6,
23,22, 260; 43,7, 260; 43,37, 260; 265; 536,6–7, 265; 539,5–6, 265;
47,30, 260; 48,14, 260; 53,26–7, 542,17–18, 265; 542,18, 265; 542,36,
260; 61,16, 260; 63,8–9, 260; 265; 543,1–4, 265; 546,16–17, 265;
71,3–4, 260; 75,30, 260; 77,23, 260; 555,25–7, 265; 557,8–585,4, 265,
78,31, 260; 79,31, 260; 82,17, 260; 268; 563,20–5, 265; 563,22–5, 265;
86,18–19, 260; 91,3, 280; 96,22, 567,29–33, 265; 583,13–585,4, 257;
260; 106,9, 280; 110,13, 260; 584,14, 266; 592,16–32, 265, 268;
118,25–6, 260 601,12–13, 264; 606,25–675,11,
in Phys., 9, 14, 19, 54, 130, 249, 252–3, 265; 611,16, 266; 619,1–4, 266;
257–77, 280, 284–5, 294, 350; 1,17, 619,10–13, 268; 632,4–634,2, 265,
264; 1,23–4, 264; 2,21–2, 266; 268; 639,3–643,8, 265, 268;
2,25–7, 266; 5,21–5, 231; 9,23–10,2, 639,7–9, 263, 268, 272; 639,7–10,
253; 15,29–30, 253; 15,30, 264; 267; 640,23–5, 268; 641,13–642,20,
16,2, 253; 16,8, 253; 22,13–15, 254; 211; 644,16–22, 213; 661,9–12,
54,8–55,26, 264, 268; 55,20–2, 273; 266; 675,12–695,8, 265, 268;
55,24–6, 267, 271–273; 55,25, 267; 675,21–9, 266; 677,9–686,29, 266;
55,26, 268, 272; 81,25, 98; 111,19ff., 677,9–689,25, 265; 685,23–686,29,
113; 128,22–31, 253; 152,5–7, 264; 267; 686,12–17, 267; 686,16–17,
156,10–12, 253; 156,16–17, 272; 267; 689,4–16, 266; 689,26–694,27,
162,12–20, 237; 163,2–12, 254; 265–266; 690,34–691,5, 212;
173,21–6, 104; 187,6–9, 254; 693,30, 266; 695,9–702,9, 266;
189,10–26, 199; 189,13–17, 242; 697,26, 266; 703,16–17, 258;
191,1–2, 268; 191,9–192,2, 264, 705,21–2, 268; 747,1–3, 264;
268; 191,34–192,1, 267; 191, 761,34–762,9, 268; 762,2–9,
34–192,2, 267; 219,19–22, 253; 264; 762,7–9, 263, 269; 762,8–9,
220,20–5, 253; 225,4–226,11, 4; 266–267; 777,11–12, 264; 797,23–6,
236,29–237,4, 264; 240,18–19, 199; 264; 812,22–3, 264; 812,23, 264;
250,28, 268; 298,6–10, 242; 820,30–2, 264; 820,30–821,4, 264;
298,6–12, 264; 303,1–5, 264; 823,16–20, 264; 824,22–5, 264;
303,18–25, 264; 303,24–5, 264; 832,17–18, 264; 838,14–15, 264;
304,5–10, 242; 340,7–9, 266; 838,20–1, 264; 838,28–33, 264;
340,12–13, 266; 340,31, 253; 859,30, 264; 870,2–9, 264; 870,3–8,
346,8–9, 266; 347,9–11, 266; 264; 873,1–2, 264; 889,17–23,
347,10–11, 266; 362,21–5, 264; 264; 890,2–3, 264; 890,7, 264;
384,29–385,11, 211; 405,3–7, 264; 891,33–892,24, 264; 891,35, 264;
405,28–9, 266; 408,14–15, 266; 892,7, 264; 892,18, 264; 893,6–9,
410,21–4, 264; 414,21–2, 268; 264; 893,6–28, 264; 894,8–10, 264;
428,23–430,10, 264, 268, 273; 894,26, 264; 897,15–17, 264;
430,9–10, 267–268, 272; 438,5–6, 897,23–7, 264; 898,2–4, 264;
264; 438,9–10, 253; 440,16–17, 266; 898,15–16, 264; 905,19–21, 264;
444,5–6, 265; 447,18–20, 265; 906,38–9, 264; 906,38–40, 264
448,20–1, 265; 454,23–4, 265; Porphyry
456,17–458,31, 264; 456,17–459,1, in Cat., 19, 39, 55, 77, 156, 158, 297,
268; 458,15–6, 267; 458,30–1, 263, 322, 379, 410, 417; 55,3–56,13,
267; 463,3–4, 265; 467,1–468,4, 407; 55,8–12, 160; 56–8, 409; 56,6,
 Index Locorum 581

160; 56,6–13, 410; 56,8–13, 169, 159; 23,13, 159; 28,27–8, 159;
409; 56,14, 379; 56,23–p.57, 156; 36,15, 159; 128,14ff., 184; 129,8–9,
56,31, 135; 57, 168; 57–8, 158, 159, 184; 141,6–7, 184; 179,6ff., 184
412; 57,7–8, 2, 159; 57,20–2, 159; ps.-Elias
57,21, 160, 412; 57,23, 160; 57,27, in Isag. (Westerink), 364, 370; 12,5,
160; 57,28, 160; 57,28–31, 160; 375; 12,18, 362; 12,19, 362; 14,27,
57,29–58,3, 414; 57,32, 169; 375; 15,4, 365; 17,5, 375; 18,14,
57,32–3, 174; 58,3, 160; 58,5–7, 2; 362; 18,29, 375; 22,20–1, 375;
58,5ff., 77; 58,7–8, 166; 58,12–14, 34,27, 365; 41,29–31, 365; 42,20,
159; 58,16–18, 169, 410; 58,23, 160; 365
59,5ff., 77; 59,10, 379; 59,10ff., 77; ps.-Philoponus (Michael of Ephesus), in
59,12, 160; 59,34–60,10, 409; GA, 23, 58, 74–5, 468
60,15ff., 160; 60,16–19, 159; 60,17f.,
322; 61,10–12, 156; 62,1, 386; Simplicius
63,16, 160; 64,9–17, 176; 64,30, in Cael., 25, 99, 240, 273, 282–3, 295,
379; 65,33, 160; 66,34, 146; 66,34ff., 310–11, 322–3, 326; 12,28, 17;
86; 67,4, 146; 71,5, 160; 71,29–30, 16,20–1, 17; 21,33ff., 95; 22,18,
160; 75, 159; 77,27–30, 167; 78,35, 95; 23,22ff., 95; 26,17–24, 310;
379; 81, 159, 173; 81,7, 379; 81,16, 26,18–19, 282; 26,19, 283; 26,21–3,
379; 82,23, 379; 86,20–32, 407; 268; 29,7–8, 256; 34,5–11, 289;
86,20ff., 77; 86,23, 85; 88,14, 379; 42,17, 283; 49,24–5, 283; 59,13–15,
90–1, 159, 163, 409–410; 90,4–9, 279; 69,9f., 344; 69,9ff., 129;
167; 90,32–4, 166; 90,33–91,1, 410; 78,17–21, 289; 78,21–79,8, 289;
91,6, 160; 91,8, 147; 91,14ff., 156; 79,8–13, 289; 79,13–14, 289;
91,19, 2; 91,19–27, 2; 91,20, 147; 84,11–13, 248, 283; 84,11–14, 4;
91,22, 160; 93,31, 379; 94,20, 379; 87,3–11, 235; 90,12, 283; 91,17–19,
96,19–20, 167; 98,34, 379; 100, 416; 289; 119,7, 283; 130,31–131,1,
100,12–16, 407; 100,21–3, 409; 17; 140,5, 256; 142,22–5, 199;
100,23–7, 407; 101,24ff., 181; 143,17–29, 200; 153,16ff., 95;
101,33, 379; 104,27, 160; 106,11ff., 159,2–9, 4; 159,3, 256; 159,7, 256;
181; 107,35, 159; 111, 416; 201,3–4, 269; 265,29, 101; 266,29,
111,9–10, 416; 111,12–15, 409; 101; 271,13–14, 234; 271,13–21,
112,17–20, 159; 112,18, 160; 115,25, 233, 350; 271,18–21, 197; 271,19,
159; 115,31, 159; 118,8, 379; 340; 271,21–7, 235; 276,14, 98;
120,33–121,3, 417; 125,18, 160; 279,5ff., 104; 284,25–285,5, 104;
127, 416; 127,3–7, 409; 128,13–15, 285,21–5, 104; 286,25–7, 104;
407; 132,4, 379; 134,25–9, 409; 297,9ff., 97; 301,3, 206; 301,4–7,
153,30–154,2, 322 198; 358,27–360,3, 200, 206;
Isagoge, 2–3, 7, 21–2, 44, 47, 157, 322, 358,27ff., 97; 359,14, 206; 360,23–4,
398, 404, 406, 408, 411, 492; 1, 200; 361,10–15, 200; 378,20f.,
411; 1,3, 47; 1,6, 405; 1,9–14, 505; 329; 380,5, 93; 387,12–19, 287;
1,15, 157; 1,18, 136; 1,23ff., 137; 6, 391,12, 101; 430,29, 93; 430,32,
166; 6–7, 159; 7, 162; 10,22–11,6, 96; 430,32–431,11, 93; 430,32ff.,
157 90, 93; 439,14, 105; 439,27–32,
ps.-Alexander 105; 454,23ff., 329; 462,20, 299,
in Metaph., 23, 41, 57, 468; 415,29–31, 340; 485,19–22, 241; 503,34,
89 142; 503,35, 92; 506,13, 142;
in SE, 482, 487; 11,23–8, 184; 11,25–7, 525,13, 311; 640,27–32, 328;
162; 11,26, 160; 11,27, 160; 23,8, 652,9ff., 96
582 Index Locorum

in Cat., 1, 25, 40, 77, 295, 297, 313, 160,1, 159; 163,6, 81; 164,4, 382;
318–19, 322–4, 326, 328, 330, 337, 165,32ff., 85; 166,8ff., 79; 167,22,
367, 384; 1, 382; 1,8ff., 55; 1,13ff., 81; 174,14ff., 78; 187,24ff., 81;
76; 1,17, 81; 1,18ff., 87; 2,3, 76, 77; 187,28ff., 85; 188,31ff., 84; 191,9ff.,
2,6, 383; 2,8, 379; 2,9, 133; 2,9ff., 340; 195,31ff., 99; 201,34ff., 79;
159; 2,10, 385; 2,13, 2, 296; 3,18, 206,10ff., 79, 84; 206,13ff., 80;
102; 3,19, 48; 3,20ff., 48; 4,13, 102; 230,32–231,1, 410; 231,20, 88;
4,22ff., 73; 6,2f., 46; 6,15–18, 246; 256,16ff., 78; 264,2–4, 409; 302,5ff.,
7,23–32, 4; 7,29–32, 328; 8,9ff., 48; 81; 302,12ff., 83; 302,15, 81;
9–13, 77, 157, 160; 9,4ff., 48; 302,15ff., 81; 306,13ff., 83; 322,15,
10,8–20, 102; 10,19–20, 96; 10,20, 73; 332,14f., 79; 338,21ff., 81;
135; 10,20–13,18, 159; 10,20ff., 77; 342–64, 80; 342,24, 79; 346,14, 82;
10,22–3, 2; 11,2–3, 409; 11,23ff., 347,6, 80; 347,22–36, 82; 348,2, 82;
77–8; 12,12–13,11, 420; 13,11–18, 348,2ff., 82; 348,22ff., 81; 351,21ff.,
96; 13,13ff., 77; 13,21–3, 419; 85; 357,28, 80; 359,1ff., 85;
13,21–6, 420; 14–15, 408; 15, 159; 359,15ff., 85; 363,29–364,6, 2;
15,7–8, 168; 16,17ff., 77; 17,3–7, 368,12ff., 87; 373,7, 81; 379,8, 381;
169; 17,5–7, 409; 18,14–16, 408; 379,8ff., 63; 381,2, 379; 381,20, 379;
18,22ff., 78; 18,26ff., 77; 18,28ff., 78; 385,3f., 73; 406,6ff., 87; 414,34, 379;
21,4, 87; 22,15ff., 159; 23,13, 382; 427–8, 409; 428,3ff., 88; 432,24ff.,
25,10ff., 81; 26–8, 176; 27, 160; 82; 433,20ff., 81–2; 433,28ff., 73,
29–30, 176; 29,29, 87; 29,29ff., 66, 81–2; 434,18f., 82
81; 30,13–15, 139; 30,16ff., 86; in DA, 17, 25, 129–30, 295, 312–15,
32,19ff., 86; 33,11, 146; 36,27, 81; 324, 326, 328, 345, 348, 513;
36,28, 81; 38,19, 81; 38,26, 386; 1,3–21, 329; 4,14ff., 123, 332;
40,5–13, 420; 41,28ff., 77; 42,1, 81; 4,18–20, 347; 6,12–16, 339; 17,2–5,
42,5–6, 162; 42,23–4, 169; 50,2ff., 341; 21,35ff., 339; 73,25ff., 333;
81; 50,28–51,12, 200; 51,3–4, 419; 85,18, 331; 89,33–5, 330; 90,29–
53, 159; 53,4ff., 170; 54,8ff., 80; 91,4, 345; 90,29ff., 123; 96,3–15,
55,24–9, 322; 58,15ff., 81; 58,27ff., 346; 96,8ff., 124; 102,12ff., 341;
81; 62,24ff., 78; 62,25, 85; 62,28, 85; 133,32, 17; 134,6, 17; 134,13–20, 17;
62,28ff., 87; 63,22, 78; 64–5, 169; 141,15–38, 17; 186,26ff., 339;
64,13ff., 78, 87; 64,26, 159; 65,21, 187,36f., 330; 233,7–17, 514;
81; 68,22–8, 419–420; 69,15–16, 240,1ff., 340; 240,2f., 341; 240,2ff.,
162; 73–4, 174; 73,28ff., 88; 76,1ff., 125; 240,4f., 341; 240,8, 341;
79; 76,3, 140; 76,13ff., 82; 76,14ff., 244,37–9, 341; 244,39ff., 341–2;
88; 76,17–22, 142; 76,19ff., 82; 244,40f., 341; 245,12, 329; 245,13–5,
76,20, 142; 76,25, 135; 77,4ff., 141; 342; 245,21f., 342; 245,37, 342;
77,7, 142; 77,7–10, 142; 77,8ff., 82; 247,41f., 341; 248,2ff., 341;
78,4ff., 82; 78,5ff., 83; 78,6, 81; 79, 276,28–277,1, 515; 276,28ff., 514;
159; 82,22–8, 108; 85,5–9, 108; 277,1–6, 515; 277,30–278,6, 516;
91,14ff., 420; 97,28ff., 83; 103,7, 313,6f., 330
163; 104,21–2, 165; 104,26f., 83; in Phys., 17, 25, 36, 38, 94–5, 129, 240,
105,3–4, 420; 116,25, 2; 127,30, 88; 273, 299, 311, 313, 326; 7,17–19,
128,5ff., 87; 128,7, 85; 129,1ff., 85; 235; 10,13, 95; 11,16, 95; 14,4, 509;
132,9–12, 159; 135,10–26, 421; 23,14–16, 96; 26,5ff., 96; 26,13, 96;
154,3ff., 79; 155,6, 159; 156,16ff., 26,15–18, 235; 28,32–29,5, 5; 43,6,
88; 157,18ff., 79; 158,27, 135; 235; 99,29, 95; 131,14, 94; 131,15ff.,
159,14, 81; 159,16, 81; 159,32, 75; 93; 133,21ff., 95; 133,24–5, 95;
 Index Locorum 583

135,16, 98; 136,10, 98; 148,22–4, 1362,11–1363,8, 234; 1362,16–17,

240; 182,19–21, 113; 182,21–6, 113; 235; 1362,17–19, 235; 1362,17–20,
205,8–15, 237; 211,13ff., 95; 234; 1362,20–1363,2, 235; 1362,32,
211,20ff., 104; 238,8, 104; 240,19, 235; 1363,2–4, 234; 1363,2–8, 235;
104; 248,2ff., 78; 249,37, 101; 1363,4–8, 198, 235; 1363,4–12, 197,
257,13–16, 235; 258,13–25, 234; 199; 1363,8–12, 197, 234, 350;
291,21, 101; 310,25–311,37, 96; 1363,9–10, 234; 1363,10–12, 241;
355,13, 98; 372,11, 101; 420,13, 98; 1363,12–24, 235
420,18, 98; 422,25, 93; 440,14, 66; Sophonias
454,19, 98; 467,1, 101; 489,20ff., in DA 1,4ff., 55; 1,11ff., 55; 126,17–19,
101; 511,30, 101; 526,16–18, 507; 255
526,20, 507; 526,21, 507; 547,11, Stephanus
93–4; 558,34, 93–4; 623,11–18, 17; in Int., 385; 5,20–36, 160; 6,17–21, 185;
640,12–18, 5; 671,4, 96; 679,12, 101; 34,1ff., 54; 46,25–47,12, 351; 63,9,
679,32, 101; 700,19, 98; 705,5, 98; 388
727,35, 93; 728,5, 93; 752,15, 93; Syrianus
759,18ff., 82; 759,20, 95; 759,20ff., in Metaph., 19, 35, 187, 194, 296; 7,12,
82; 765,31ff., 129; 766,18, 82; 165; 12,28ff., 509; 12,29–13,3, 510;
766,18ff., 82; 894,12, 98; 923,7ff., 20,9, 165; 28,14, 165; 30,28, 165;
64, 66; 1036,14–17, 129; 1055,15, 32,9–10, 165; 36,28, 165; 43,6–23,
106; 1055,24–1056,6, 106; 223; 80,4–7, 3, 296; 80,4–81,14,
1130,7–1131,7, 271; 1133,10, 256; 188–189; 80,16, 188; 80,28, 189;
1133,17–1135,15, 271; 1141,19, 91,20, 510; 91,29–34, 511; 95,11,
286; 1147,10–1149,4, 271; 165; 95,29ff., 511; 98,34, 165;
1157,6–1159,4, 271; 1159,18–19, 101,22–103,12, 517; 105,21ff., 77;
287; 1164,11–30, 271; 1165,34, 112,14–17, 222; 112,35–6, 222;
329; 1166,37–1167,16, 271; 113,15–16, 222; 117,25–118,11, 4,
1171,30–1175,26, 271; 1174,22–5, 198; 142,19, 341; 155,29, 165
293; 1177,22–6, 293; 1177,38–
1178,5, 293; 1178,7–1179,26, 270; Themistius
1219,3–5, 287; 1327,14–16, 200; in An. Post., 21, 24; 1–2, 407; 1,2–12,
1327,15–16, 292; 1329,17–19, 200; 18; 18,26, 397
1331,30–3, 200; 1333,23–7, 292; in An. Pr., 21, 99; 26,31–29,28, 99
1333,27–30, 293; 1333,28–30, 208; in DA, 314, 347; 1,2–5, 344; 3,4–8, 3;
1334,37–9, 208, 293; 1336,35ff., 24,22ff., 123; 41,11ff., 123; 42,15ff.,
330; 1351,27, 98; 1354,9ff., 95; 346; 42,35–7, 346; 43,25–34, 124,
1355,32, 95; 1358,29–30, 200; 348; 44,21ff., 123; 61,22ff., 92;
1360,24, 234; 1360,24–1363,24, 98,12–109,3, 125, 344; 100,7, 128;
233; 1360,25, 235; 1360,28–31, 234; 100,11, 128; 100,32–5, 127;
1360,28–34, 235; 1360,30–1, 234; 102,30ff., 125; 103,34–5, 128;
1361,11–12, 236; 1361,11–1362,12, 104,15f., 125
234; 1361,11–1363,12, 4, 197; in Metaph. 9,3–11, 122; 32,18ff., 126;
1361,12–1362,10, 234; 1361,31–3, 115,5, 122
206, 234; 1361,32–3, 234; 1362,7–8, in Phys. 4,24, 382; 29,20–3, 506;
235; 1362,11–15, 206, 234; 160,26f., 82; 163,6, 82; 163,6ff., 82
584
 General Index

Abel, A., 252 Ammonius, 3–9, 13–14, 16, 19–20, 25, 41,
Abélard, 490, 495–6 46–8, 50–2, 56–7, 66–7, 90, 92, 99, 104,
Abravanel, 198, 202, 208, 210 130–1, 155, 159–60, 165–6, 169–70,
Abu Bishr Matta, 203 177–8, 182–5, 190, 191, 196–9, 202,
Abû Qurra, Theodore, Bishop of H.arrân, 203, 214; his metaphysics, 215–50;
308 on mathematics, 503–5, 509, 513–14,
Accattino, P., 95 516, 519; school of, 40, 220
Achaicus, 79 Ammonius Saccas, 3, 196, 219
Ackrill, J.L., 71, 127 Anâbû [Anebo], 303
Adrastus, 17, 18, 23, 86, 91, 94, 454–7, 458, Anastos, M.V., 252, 259
459, 460, 461, 462, 463, 464, 474 Anatolius, 354
Aedesia, wife of Hermeias, 349–50 Andronicus of Rhodes, 1, 51–2, 61–80,
Aedesius, 11 81, 82, 84, 86, 95, 102–3, 380–1,
Aeneas of Gaza, 12, 228, 357–8 476, 477
Agapius, a pupil of Proclus, 325, 361 animal sacrifices, 12
Agathias, 299–301, 307, 312, 355 Anscombe, G.E.M., 117
al-Kindî, 307 Anselm, 490
al-Mascûdî, 302–5, 306, 307 Antiochus of Ascalon, 7, 67, 69, 194;
al-Nadîm, 307, 315, 323–4, 325 school of, 68
al-Nayrîzî, 324–5 Apellicon, 61, 67, 68–9
al-Qiftî, 309, 314, 315, 324 Apollonius Dyscolus, 169, 407
Albertus Magnus, 477–80 Apuleius, 21, 404
Albinus, 40, 52, 147, 397 Aquinas, Thomas, 4, 29, 51, 195, 206,
Alcinous, 194 209–10, 212, 214, 478–9, 483
Alexander of Aegae, 95–6 Arabic sources: for Simplicius, 308–10,
Alexander of Aphrodisias, 1, 3, 6, 16, 311–12; of the catalogue of Aristotle’s
17–18, 23–4, 25, 26, 29, 39, 41, 52, 57, writings, 62–3
71, 76, 77, 82, 83, 86, 119–20, 122, 125, Aratus, 101
127, 129, 143, 187, 198, 200, 202–3, Archytas, 75, 80, 133, 142, 380, 417
206–10, 234, 287, 309, 327, 334–7, Arefshatian, S.S., 364
339, 347, 350, 439; his commentaries, Ariston, 67–9, 70, 75, 84
349–75; on mathematics, 506–8, Aristoteles of Mytilene, 90, 94–5
509, 511, 513–14, 519; school of?, Aristotle: obscurity of style, 6; Organon,
89–118 151, 153, 155–7, 217, 220, 227–33,
Alexander the Great, 64, 66 276–7, 321, 352, 361, 364, 368, 377, 391,
Alexander III , Pope, 424, 426 394, 398–401; pedagogy in the
Alexander of Lycopolis, 360 Organon, 408–9; pragmateiai (school-
Alexandrian school of Neoplatonism, treatises), 61, 65; theology and physics
89–118, 327, 334–6, 339, 347, 350 195–214; view of mathematical objects,
ambiguity, 154, 174–8 502–3
Ambrose, 196 Arius Didymus, 75, 196, 112, 196, 439
Amelius, 44, 135 Arnim, H. von, 43, 100

585
586 General Index

Asclepius, 4, 40, 41 43–4, 56, 197, 218, 219, Blumenthal, H.J., 14, 39, 56, 62, 71, 76, 90,
221–7, 229, 231, 233, 236–50, 252, 256, 130, 222, 241, 242, 256, 283, 295, 312,
257, 309, 350–1, 352 313, 348, 505, 513
Asmus, 350, 351 Bochenski, I.M., 76, 252
Aspasius, 17–18, 23, 41, 75, 86, 93–4, 96, Bodeus, R., 75
386, 431, 441–3, 454, 455, 474–7 body, 117; finite, 198–9, 201, 207; its
astrology, 366, 429 relation to soul, 109–10, 122–4; ousia
Athanasius II , Patriarch, 351 148–9
Athenian school, 19, 349–51, 353–4, 382, Boethius, 7, 15–16, 21–2, 25–6, 51, 55, 57,
400; of Neoplatonism, 216–17, 248, 84, 93, 102, 155, 157, 159, 160, 162, 163,
296, 341 169, 172, 175–7, 182–4, 219, 241, 248,
Athenodorus, 68, 69, 75, 77, 78; his 297, 371, 377–401, 403–21, 486,
criticism of Aristotle’s Categories, 488–90, 492–5
85, 86 Boethus of Sidon, 81–2
Athens: libraries, 61–2; sack by the Boethus (Peripatetic), 476
Heruls, 61–2 Böhm, W., 253, 256, 262
Atticus, 67, 75, 86, 97, 98, 99, 146, 194 Boissonade, J.F., 192
Augustine, 7, 12, 139 Bonaventure, 26, 27, 195, 209, 477–80
Aujoulat, N., 219 Booth, E.G.T., 252
authenticity, 368–9, 371, 374, 383, 389–90; Bossier, F., 17, 130, 312–13, 317–19, 348,
of Alexander’s works, 102–3; of 513
Aristotle’s works, 7, 65; of Brams, J., 395
ps.-Pythagorean works, 4–5; of Brandis, C.A., 23, 35, 52, 64, 66, 74, 75, 76,
Simplicius’ works, 311–16 85, 86
Averroes, 24, 26, 202, 203, 205–10, 212, Brandt, S., 51, 57, 157, 163, 371, 391–3, 394,
478, 479, 488, 523 400, 403
Avicenna, 24, 29, 197, 202–5, 206, 206, 211, Brehier, E., 187
213, 278, 478, 488 Bréhier, L., 252
Axelos, K., 252 Brentano, F., 29
Brink, K.O., 68, 327
Bacon, R., 195, 209 Brockhoff, 424
Badawi, A., 23 Brown, H.V.B., 23
Baghdad, 20, 305 Browning, R., 58, 75, 423–38, 447, 465,
Baltes, M., 86, 97 472, 487
Balty, J. & J., 9–10 Bruns, I., 46, 91, 92, 100, 103, 104, 108,
Bardy, G., 227, 228 112, 113, 116, 441
Barker, E., 468, 473 Brunschwig, J., 72
Barnes, J., 127, 385, 390 Buridan, J., 155, 177, 213, 490
Baumstark, A., 50 Burkert, W., 4
Becchi, F., 459–60, 476 Busse, A., 44, 73, 136, 190, 191, 232,
Behler, E., 233 297, 352, 364, 371, 378, 379, 384,
Bernardinello, S., 74 392, 393, 482
Bessarion, 318, 321 Byrne, P., 209
Beutler, R., 225, 323 Byzantium (see also Constantinople), 18,
Bidez, J., 155, 344, 379, 380, 381, 384, 385, 22, 24, 399, 435, 487, 492
387, 394, 400
Bielawski, J., 62 Calcidius, 52, 216
Bielmeier, A., 349 Cameron, A., 13, 20, 248, 257, 259, 281,
Blackwell, R.J., 209, 212 308, 353
 General Index 587

Cantarelli, L., 354 Cosmas Indicopleustes, 248, 252, 293–4,

Capelle, W., 87 354
Cappuyns, M., 377, 394 Cougny, E., 322–3
Cardullo, L., 19 Courcelle, P., 12, 228, 366, 384–7,
Cassiodorus, 391, 400–1, 415 397, 400
Caston, V., 8, 35, 95, 133, 187, 295 Coutant, V.C., 106
Categories: 5–8; ta pro tôn topôn, 76, 103 Cranz, E., 522
Cave, G., 252 Cratippus, 66–71
Censorinus the Academic, 100–1 Creation, 3, 15–16, 19–20, 27, 97, 196, 199
Chadwick, H., 390, 399 Crescas, 202, 206, 210
Chaldean Oracles, 360 Creuzer, 353
Chennos, P., 63 Ctesiphon, 15, 20
Chortasmenos, John, 320–2
Chosroes, 60, 299, 300, 306–7, 311–12, Dagron, G., 119
315, 355 Dalsgaard-Larsen, B., 19, 152
Christ, two natures of, 443–6 Damascius, 8, 13, 15, 16, 19, 20, 165, 189,
Christianity, 5, 11–16, 18, 23, 27–8, 40, 216, 218, 220, 247–8, 281, 298–301, 307,
56–7, 199–200, 214, 216–19, 233, 311–12, 316, 323, 336, 340–1, 349–51,
238–9, 241–4 353–4, 355, 358–9
Chrysaorius, 156 Darrouzès, J., 427, 435, 465
Chrysippus, 99, 102 David, 8–9, 15, 20–1, 22, 40, 41, 44, 47, 51,
Chrysolanus, Peter, 430 53, 56–7, 59, 76, 216, 250, 252, 259, 297,
Chwolsohn, D., 302, 305, 307 322, 362, 364–5, 370
Cicero, 1, 55, 61–2, 67–71, 72, 190, 194, Davidson, H.A., 199, 206, 208, 210
328, 396–7, 398, 400, 401 Deichgräber, K., 309
Claggett, M., 213 Deitz, L., 87
common sense, 104, 113–14, 448 Demosthenes, 49
Comnena, Anna, 423–38, 447, 453, 465, des Places, E., 75, 86–7, 97, 99
467, 472–3 determinism, and free will, 116
Comnenus, Alexius, Emperor, 424–5, 426, Deuse, W., 12
435, 444–5 Dexippus, 2, 9, 19, 25, 39–40, 41, 52, 66, 83,
Comnenus, John II , Emperor, 428 108, 133–49, 159–175, 177, 181–2, 196,
Comnenus, Manuel I, Emperor, 424, 426, 297, 337, 382, 420, 494–5
429, 423, 434 diagram, 384, 388, 393, 394, 397
conceptualism, 26, 451, 495 dianoia, 160, 165, 377, 469, 504–5,
Constantine the Great, Emperor, 11, 12, 39 511–13
Constantine IX Monomachus, Emperor, didascalica or kephalaia (headings), 382–4,
437 386, 392
Constantine VII Porphyrogenitus, Diehl, E., 38
Emperor, 60 Diels, H., 35–6, 38–9, 42–3, 48, 94, 271,
Constantinople (see also Byzantium), 13, 293, 299
18, 20–1, 22, 41, 308, 334, 354, 361, 366, Dihle, A., 63
426–7, 428; Patriarchal school in, 424, Dillon, J., 9, 11, 19, 78, 84, 86, 87, 96, 97,
432, 434 165, 330–2
Cornford, F.M., 97 Dio Chrysostom, 43
Cornutus, 77, 78, 85, 87, 379 Diocletian, Emperor, 354
Corrigan, K., 76 Diodorus Cronus, 99, 354, 400
Corte, M. de, 334 Diogenes, 299, 300
Cosmas II Atticus, Patriarch, 425, 432 Diogenes Laertius 50, 52, 63–4, 73
588 General Index

Dionysius of Halicarnassus, 69, 159 Eutocius, 257, 261, 352

Dod, B.G., 24, 486–7 Évrard, É., 8–9, 54, 119, 241–2, 248, 252–4,
Dodds, E.R., 255, 230–1, 327, 333 257, 271–2, 277
Donati, S., 295 exêgêsis tês lexeôs, 9
Donini, P.L., 97, 108, 119, 476
Dooley, W.E., 104 Fabricius, J.A., 192, 319
Dörrie, H., 71, 75, 77, 78, 84, 96, 252, Farabi, 4, 197, 233
316–17, 473 Favorinus, 101
Doxopatres, John, 320–3, 434–5 Feldman, S., 208
Draeseke, J., 430, 443 Feliciano, G.B., 442–3
Drossaart Lulofs, H.J., 75 Festugière, A.J., 200, 234, 246, 332
Ducci, E., 241 fifth element, 14, 98–9, 213, 255, 264,
Duhem, P., 213 266–7, 274, 277, 287, 289, 352
Dunbabin, J., 478 fire, 105
Düring, I., 61–3, 64, 65–7, 68–9, 71, 328 Follet, S., 99–100
Fraser, P.M., 69, 70, 84
Ebbesen, S., 22–4, 25–6, 57, 151–85, 381, Frede, M., 76, 103
385, 390–1, 393, 395, 403–21, 466–7, Freudenthal, G., 324
481–97 Freudenthal, J., 43
Edel, A., 503 Friedländer, M., 197
Elias, 6, 8–9, 15, 20–1, 22, 40, 41, 44, 46, 47, Frisius, J., 318
48–9, 53, 56–7, 64–7, 197, 216, 239–41, Fritz, K. von, 77, 79, 83, 87
252, 261, 297, 329, 351–2, 361–6, 368,
370, 380, 383, 430 Gaiser, K., 78, 84
Elias, Prefect of Illyricum, 361 Galen, 77, 79, 90, 93, 96, 325
Ellis, J.O., 28, 117 Galileo, 27, 29
entelechy, 109–10, 345, 346–7, 433 Gandillac, M. de, 213
Epictetus, 14, 46 Gätje, H., 309–102, 311, 314–15,
Epicurus, 39, 100, 458, 459 323–4, 325
epidromê, 105 Gauthier, R.A., 440, 443, 452, 475
epigraphê, 47–8, 52–3, 383, 41 Geach, P.T., 117
epikheirêma, 386, 415 Geminus, 101
epinoia, 170, 504, 506, 507, 513–14 Genequand, C., 206–7
epoptia, see mystical union Gerard of Cremona, 24, 324
essential predication, 169–73 Gersonides, 210
eternity: of movement, 263, 264, 267, Gesner, C., 317, 318
269–71; of the world, 227–9, 241–2, Gessius, 366
252–4, 257, 259, 263–4 Ghazali, 24
Euclid, 101 Gilbert the Porretan, 496
Eudemus, 39, 64, 66, 75, 76, 94, 95 Gilson, E., 24
Eudorus, 68, 69, 75, 82, 84–5, 86, 96 Giocarinis, K., 443, 450–1
Eulamius (or Eulalius), 299 Girolamo da Cremona, 522
Eusebius, 86, 454 Girolamo Donati, 522
Eustathius, 322, 323, 369 Glucker, J., 67, 70, 101
Eustratius of Nicaea, 3, 6, 18, 22–3, 26, 59, Gnostics, 305, 309
430, 431–2, 439, 440, 443–53 God, 22, 139, 194; of Aristotle’s as Creator
Euthymius Malakes, 424 or Sustainer, 25, 27, 195–214; the
Euthymius Tornikes, 424 Intellect, 13, 215, 225–7; the One, 13,
Euthymius Zigabenus, 434 215, 218, 222–5
 General Index 589

Gottschalk, H.B., 1–2, 18, 61–88, 91, 475–6 Hierocles of Alexandria, 3, 13–14, 39, 44,
Goulet-Cazé, M.-O., 62 52, 196–7, 216–20, 244
grammar, see philology Hipparchus, 101
Green-Pedersen, N.J., 414, 486, 490 Hippocrates of Chios, 324
Gregory of Nazianzus, 120–1, 356, 425 Hoffmann, J.G.E., 388
Grosseteste, R., 24, 209, 439–41, 442, Hoffmann, P., 15, 38, 40, 56, 282, 295, 310
448–9, 453, 457, 473, 477 Housman, A.E., 67
Gudeman, A., 245, 252–3, 254, 257, 260, Hunger, H., 22, 75, 320–2, 424
263, 271, 283 huparxis, 508, 509
Guérard, C., 5, 309 hupekkauma, 105
Guilland, R., 361 hupokeimenon, 86, 104, 117
Gutas, D., 63 hupomnêma, 49, 92, 298, 367
Guthrie, W.K.C., 79, 105, 327 Hypatia, 216, 301
Guzzo, A., 127 hypostasis, 125, 143, 177, 194, 217

Haase, W., 92, 245, 263, 331, 337, 476 Iamblichus, 3, 5–6, 9–13, 16, 19, 22, 52,
Hadot, I., 7, 14, 17, 20, 130, 219–20, 244, 56, 82, 120, 125, 128, 131, 133, 135–6,
295–326, 348, 353, 355, 381–3 141–2, 159, 165, 216, 242, 296, 297–8,
Hadot, P., 39, 133–49, 148, 170, 196, 220, 316–21, 323, 326, 328–32, 338–40, 344,
295, 297, 381 369, 379, 380–2, 384–5, 417–21, 509,
haireseis (sects), 100 517–20; his school at Apamea, 9–12
Hamelin, O., 122, 328 Ideas, 2, 3–5, 7, 156, 194
Hanquet, R., 474–5 imagination, 130
Harles, G.C., 319–20 Immisch, O., 41, 57, 58, 468
Harlfinger, D., 90 impetus, 27, 29, 211, 213, 264, 268
harmony: of Aristotle with Christianity, India, 29
195–214; of Plato and Aristotle (and infinite: force, 14; past, 29; power, 195–214
others), 3–5, 197–8, 244, 385–6, inherence, 80
399–400; of Plotinus and Aristotle, Intellect, 120, 124–8, 334–44, 347–8; active,
133–50 18–19, 27, 125, 127, 334–6; demonic or
H.arrân (Carrhae), 20, 301–2 angelic, 343; divine, 341–2; from
Hartmann, N., 71 outside (nous thurathen), 94–5;
Hayduck, M., 57, 98, 221, 238, 260, 315, material, 95; see also God
334–5, 430, 439 intellective theory (noera theôria), 19, 22,
Heinze, R., 78 141–3, 145
Heliodorus of Prusa, 22, 26, 42, 349, intention, 29
353, 440 Irene Ducas, Empress, 426
Henry, P., 104, 187, 337 Irene, Sebastocratorissa, 432
Hephaestus, 354–5 Isaac, J., 21
Heraclides of Pontus, 98–9 Isagoge, 7, 21–2, 40, 44, 46–7, 136, 152,
Heraclides, Aurelius Eupyrides, 99–100 156–7, 166–7, 218, 232–3, 257–8, 361,
Heraclius, Emperor, 41, 366 363–4, 367
Hermeias, 8, 44, 216, 303, 340 Isidore, 299, 300
Herminus, 18, 77, 86, 90, 93–4, 111, Islam, 29
175, 369, 378 Iwakuma, Y., 495
Herod the Great, 74
hexis, 458 Jacobi, K., 496
Heylbut, G., 439, 440–1, 455 Jacoby, 62
Hicks, R.D., 328 Jaeger, W., 74, 102
590 General Index

Jahn, A., 356 Lloyd, A.C., 2–3, 26, 77, 108, 130, 152, 189,
James of Venice, 24, 26, 395, 432, 487–8 219, 222, 241, 245, 313
James of Viterbo, 126 Lloyd, G.E.R., 116, 125
Joachim, H.H., 72 logic, 392, 393; Aristotelian, 39–40, 133,
Joannou, P., 253, 256, 443, 451, 452 155–7, 170, 193, 377; as a tool, 45,
John Damascene, 430, 434 47, 73; for beginners, 420; formalist
John Italus, 430, 431–5 view of, 77; from ancient to medieval,
John Pagus, 497 481–2; introduction to (see Isagoge);
Jones, A.H.M., 119 medieval, 481–500; part of philosophy,
Joseph, H.W.B., 79 71, 102, 383; Porphyry’s, 151–86; Stoic,
Judson, L., 199, 208 415–16
Julian, Emperor, 9–10, 11, 130–1, 301–2 logos, 160–1, 177–8, 181–5, 200; basilikos,
Julius Zosimianus, 100 426; epitaphios, 426; exôterikoi logoi,
Justinian, Emperor, 15, 20, 192, 279–80, 447, 367; hupographoi logoi, 393; logoi
299, 354 as Platonic Ideas, 215–16, 242–3,
247, 255
Kahn, C.H., 75 Longinus, 217
Kalbfleisch, C., 85 Lucchetta, G.A., 248, 253, 256, 259, 264
Kappelmacher, A., 400 Lucian, 18, 453–4, 456, 458
Kennedy, G.A., 322–3 Lucius, 78, 86–8
Kenny, A., 9, 18, 24, 63, 75, 90–1, 403, 454, Luna, C., 295
462, 476 Lydus, John, 138, 354
Khatchadourian, H., 15, 40 Lynch, J.P., 18, 66, 67, 70, 89, 90, 99–100
Khostikian, M., 364 Lyons, J., 161
khreia, 418, 461
Kneale, W. and M., 61 McDiarmid, J.B., 96
Know Thyself (gnôthi sauton), 303 MacLeod, C.M.K., 245, 253, 256, 262
kosmos vs. to en kosmôi, 84 Macrobius, 397
Kötting, B., 190 Madigan, A.J., 92, 101
Kraemer, J.L., 253, 259, 267, 277–8 magnet, 105–6
Kraye, J., 7, 63 Mahdi, M., 233, 241, 248, 256, 259, 278–9
Kremer, K., 220, 231, 232, 237 Mahé, J.-P., 297
Kristeller, P.O., 126 Mahoney, E.P., 3, 19, 25, 126–8
Krumbacher, K., 41, 57, 424 Maimonides, 197
Kurfess, A., 127 Mâlik b. cUqbûn, 302, 304, 305
Manandean, Y., 297
Lambeck, P., 320 Manicheanism, 301, 308–9
Landauer, S., 122 manuscripts: Copenhagen, Thott 403, 406,
Latin translations (see also Boethius and 418; Paris gr. 184–5, 322, 432, 441–2,
James of Venice), 27 2162, 401; Vatican, Urb. 441–2; Vienna,
Lemerle, P., 366 phil. graec. 320–1, 423–4
lemma or quotation, 9, 54, 55, 388 Marcellinus, 322
Leo, Fr., 49 Marcus Aurelius, Emperor, 18
Leo Magentinus, 160 Marinus, 192–3, 245, 329
Leon of Chalcedon, 444 Mariotti, I., 84
Leszl, W., 96 Marius Victorinus, 21
lexis (text), 53–4, 298, 469 Marrou, H.I., 281
Libanius, 120–1 Martindale, J.R., 119
Littig, Fr., 50–1, 61, 64–5, 67, 70, 79–80 Maspéro, J., 256
 General Index 591

mathematics 501–20 Neoplatonism: Christian, 449–52;

Mathieu, V., 127 curriculum, 5, 16–17, 89, 102; elements
Maximus of Ephesus, 216 in the de Anima 327–48; pagan, 15,
Maximus Planudes, 321 241–2, 357, 433; school of H.arrân,
medieval commentators, 414 302–5; school practices, 5–11,
Mekler, S., 69 28; theology, 12, 16, 25, 196–8,
Mercken, P., 91 215–50
Merlan, P., 92, 93, 97, 116, 119, 127, 141, Neumann, C., 364
218–19, 229, 248, 337, 344 Nicephorus Bryennius, 426–7
metempsychosis, 433 Nicetas Acominatus, 434–5
Metochites, T., 57 Nicetas Choniates, 428–9
Meyer, W., 60 Nicetas of Serrai, 445
Meyerhof, M., 252, 256, 305 Nicolaus of Damascus, 74, 96, 128
Michael Anchialus, 434, 472 Nicolaus of Methone, 434
Michael Choniates, 423, 425–6 Nicostratus, 78, 82, 86–7
Michael of Ephesus, 3, 18, 22–3, 26, 41, 45, Nisibis, 300, 306
57–8, 59, 74–5, 151, 184, 429, 430–1, Nock, A.D., 190
432, 439, 440–3, 464–72 noêma, 386
Michael Glycas, 434 nominalism, 26, 155, 490, 492–5
Michael Italicus, 425–6, 432 Norvin, K.P.W.J., 356
Michael Psellus, 431, 433, 446 Nuchelmans, G., 161
Michael VII Ducas, Emperor, 466–7 Nutton, V., 96
Mignucci, M., 109
Minio-Paluello, L., 24, 377, 381, 384, O’Brien-Moore, A., 70
395, 397 O’Cleiright, P., 76
Minnis, A.J., 7 O’Meara, D.J., 5, 13
Minniti Colonna, M., 227 Obertello, L., 230, 403, 404, 410, 420
modism, 491 Ockham, 491, 493, 497
Mohammed, 449 Olympiodorus, 8–9, 15, 16, 20–1, 40–1,
Molinier, E., 521 44–8, 52, 53–4, 73, 76, 160, 189, 192–3,
Montanari, E., 98 197, 241, 245–6, 248, 252, 257; school
Moody, E.A., 212 of, 259
Moraux, P., 18, 23, 57, 61–6, 68–70, 73, Omont, H., 317
75–96, 102–4, 108–9, 125, 128, One, the, and the Indefinite Dyad, 78
187, 321 ontology, 139; behind logic, 451–9
Moreau, J., 194 oral teaching (sunanagnôsis), 6, 383, 384
Moreschini, C., 86 orality vs. text, 7, 42–9
Morris, J., 119 Oribasius, 53
movement, 81–3, 85, 88, 129–30, 138, 143, Origen, 7, 52–3
145–6, 26–7, 234–5, 240, 263–4, 266–7, Origen, the pagan, 217, 219
269–71, 274–7, 287–91 Owen, G.E.L., 80, 125
Movia, G., 83
mystical union, 12–13, 450 Pachymeres, G., 57, 58, 59, 366
Packard, D., 89
nature (phusis), 123, 146, 223, 332–3 Parente, I., 78
Neckham, A., 485 Parmenides, 98, 165, 246, 307, 309, 340
negative theology, 148, 223–4, 232–3, 356 Paschalis II , Pope, 444
Neilus Doxopatres, 434–5 Peck, A.L., 334
Nemesius, 478 Pellat, C., 302
592 General Index

Peripatetics, 379 Praechter, K., 7–9, 13–14, 23–4, 61, 77–8,

Perler, O., 252 84–5, 87, 187–8, 216–20, 233, 237, 242,
Pernot, L., 321 245, 283, 296–7, 308, 316–17, 430–1,
Peter III Mongus, Patriarch, 351 464–7, 472–3
Peters, F.E., 314, 324 Praetextatus, 21, 397, 404
Pétridès, S., 256 Prantl, K., 61, 73, 79, 81, 83–5
Petrus Grossolanus, 444, 445–6 praxis (lecture), 9, 53–4
Pfeiffer, R., 62, 65 Preisigke, F., 325
phantasia, 404 Presocratics, 4–5, 16, 27, 38–9, 295, 313
Philology, 10–11, 36, 38, 43 Preus, A., 23, 75, 468
Philoponus, 4, 8–10, 14–16, 19–20, 24, 40, prime matter, 14, 29, 196
41, 46–8, 53, 56–7, 92, 95, 99, 101–2, prime mover, 95, 143, 196
120, 123, 124, 130–1, 159–60, 162, 167, Priscian, 15, 17, 20, 60, 130, 299, 312–13,
197–8, 199–209, 211–14, 297, 310, 311, 315, 339, 348, 513
328–9, 332–4, 338–40, 345–7, 350–3, Probus of Antioch, 387, 388
363–6, 386–7, 407–8, 433, 481–97; Proclus, 3–4, 5–8, 12–15, 16, 19, 25, 41,
chronology of his commentaries, 44–6, 48, 55, 188, 190, 192–4, 197–203,
251–94; development of his thought, 210–11, 216–17, 222–3, 225, 228–32,
251–94; on mathematics, 503–5, 506, 234, 244, 246, 248, 249–50, 270, 280–1,
509, 512–14, 519 297–8, 309, 318–19, 321, 330–4,
Photius, 121, 129, 220, 351, 492 338–43; on mathematics, 509–10,
Piccolomini, F., 312, 348 512–13, 517–18, 520; school of,
Pinborg, J., 403, 488, 491, 493 385, 396, 399, 400
Pines, S., 197, 199, 200, 201–3, 211, 248, Procopius, 306, 354–5, 364
273 pronoia, 399
place, 2, 14, 80, 82, 84 prooemium, 232, 382–4, 392
Plato’s theology, 5–6, 247; Receptacle, 97; propositions, 394, 405, 408, 411, 413–16;
Timaeus, 5–6, 45, 55, 96–8, 137, 189, categoric, 394, 410
190, 196, 200, 218, 246, 290–1, 302, protheôria, discussion of doctrine versus
318–19, 330–1, 342–3, 433, 455 discussion of text, 9
pleasure, 101, 112–13, 115 prothesis, 51–2, 381
Pletho, 156 providence, 95, 97–8, 104, 108–9, 111,
Plezia, M., 61–6, 73, 79 114, 217
Plotinus, 2, 3, 5, 62, 65, 76–7, 81–3, 86–8, ps.-Alexander, 41–2, 151, 159–60, 162, 165,
120, 121, 123, 127, 187, 216, 218–19, 184
298, 309, 335–8, 339–40; objections to ps.-Archytas, 79, 419
Aristotle’s Categories, 133–50; school ps.-Dionysius, 356
of, 90 ps.-Elias, 362, 365, 370
Plutarch of Athens, 6, 8, 15, 19, 44, 46, 193, ps.-Philoponus, 159, 241, 328, 340, 346
196–7, 216, 338–40 ps.-Pythagorus, 4–5
pneuma, 333–4 ps.-Rhabanus, 495
Pomponazzi, P., 522 Ptolemy, the astronomer, 85
Pope-Henessey, J., 521, 524 Ptolemy-al-Garîb, his catalogue of
Porphyry, 1, 2–3, 5, 7, 12, 16, 19, 22, 26, Aristotle’s writings, 62–3
39–40, 41, 47, 55, 62, 66, 76–7, 120, 123, Pythagoras, 188–9, 191–2, 194, 174, 177
128, 133–50, 377, 378–81, 383–5, Pythagorean mathematics, 296, 516–18
493–5; legacy to logic, 151–86, on
mathematics, 504–5, 518–19 quadrivium, 465
Potamo, 96 Quain, E.A., 383
 General Index 593

qualities: distinguished from substances, scholia: ancient, 453–6; Books 2, 3, 4 and 5,

80; particular, 28 456–64, Books 5, 9 and 10 464–72,
quantum physics, 28 Book 7 473
Quine, W.V.O., 501 Schwyzer, H.-R., 332, 337
Scot, M., 206, 209
Rabe, H., 292, 322–3, 329, 366, 416 Scott, W., 307
Rand, E.K., 377, 400 Scotus Eriugena, 492
Ranke, H., 325 Scotus, Duns, 491
Ratramnus, 492 Seeck, O., 119
Ravaisson, F., 16 Segonds, A., 256, 260
Reale, G., 252 sensation, 144–5, 147
realism, 26, 488, 490, 492, 497 senses, five, 28–9
reason, discursive, 128, 337–8, 342 sententia, 138–40, 388–9, 390, 392, 416
Regenbogen, O., 63 Sergius, Patriarch, 278
relatives, 28 Setton, K.M., 190
Rémondon, R., 248 Sharples, R.W., 89–118, 198, 200, 439, 459
Renaissance, 27, 41–2, 57–8, 126, Sheldon-Williams, I.P., 248
194 Sheppard, A., 297
Repici, L., 76 Shiel, J., 377–401, 406–7, 486
rhetoric, 10–11, 23, 58, 63, 69, 74–5, Sicherl, M., 356
189, 320–3, 382, 414 Simler, J., 318
Richard, M., 8, 43, 350 Simplicius, 1, 4–5, 6, 8, 9, 14–15, 16, 17,
Riemannian geometry, 18 20, 28, 29, 36, 38–41, 47–8, 55, 56, 75,
Rijk, L.M. de, 404, 408, 418–20 78–80, 87–8, 92–6, 99, 101–6, 120, 123,
Rist, J.M., 77 124, 129–30, 133, 135–6, 140–2, 146,
Romano, F., 189, 296–7 159–60, 168–70, 192, 197–200, 206,
Roscelin, 490 210, 216, 217–18, 219–20, 233–5,
Rose, V., 442–3, 448 238–43, 245, 248, 251, 256, 279, 327–30,
Ross, W.D., 113, 327 332–3, 339–47, 348, 350, 367, 378–9,
Rossi, P., 486 380–4, 409, 419–20, 432, 439, 455; his
Roueché, M., 10, 20–1, 22, 41, 366 acquaintance with Philoponus, 282–4;
Ruland, H.-J., 98, 108 his life and work, 295–326; on
Russell, D.A., 71 mathematics, 505–7, 513–18, 520
Singer, S., 53
Sâbians of Harrân, 302–5 skopos, 47–8, 50–2, 125, 338, 383, 403
˙Sabra, A.I., ˙310, 324–5 Skowronski, L., 356
Saffrey, H.D., 13, 25, 119, 187–94, 198, smell, sense of, 28, 29
222–3, 230–2, 234, 241, 245, 247–8, Smith, A., 12, 76
253, 256–8, 259, 261 Socrates, 109, 121, 131, 179, 281, 359
Salaville, S., 443 solipsism, 413
Sambursky, S., 85, 101, 253 Solmsen, F., 74, 399
Sandbach, F.H., 69 Sopatros, 323
Sanjian, A.K., 15, 40, 364 Sophonias, 22, 39, 42
Sarton, G., 213 Sorabji, R., 14–15, 16–18, 20, 26–8, 40, 61,
Schleiermacher, F., 443, 447–8, 452, 91, 97, 112, 127, 195–214, 233–4, 251,
464, 473 253, 277, 282, 295, 481
Schmitt, C., 27 Sosicrates, 95
Scholasticism: ancient 403, 408; medieval Sosigenes, Alexander’s teacher, 86, 90, 92,
182, 377; Western 182, 488 94, 111
594 General Index

Sosigenes, the Stoic, 92 syllogism, 92–3, 111, 130–1, 157, 489;

Soterichus Panteugenes, 425, 434 constituent parts of, 378; forms 447,
Sotion, 95 449; treatises on, 394–5
soul, 107–8, 117, 122–8, 130, 137, 141, Symmachus of Phlya, 100
143, 215, 223, 237–8, 244, 253–5, 285, Symmachus, 397
287–92, 329–31; definition of, 109–10, Synesius, 281, 310
123, 331, 345; divisions of, 433; four Syrian school of Neoplatonism, 216
stages of, 450; immortality of, 27, 98; Syrianus, 3–4, 5, 8, 19, 25, 29, 35–6, 41, 44,
logoi as constituents of, 510–12; powers 187–94, 197–9, 202, 203, 208, 349–50,
of, 393 384; on mathematics, 509–13; school
Southern, R., 484–5 of, 382
space, 18, 29, 82 Szlezák, T.A., 5, 79–80, 82–5, 187
Spath, R.J., 209, 212
Spawforth, A.J.S., 100 Tâbit b. Qurra, 20, 305, 307
Speusippus, 4, 81 Tannery, P., 56, 233, 241, 260, 350, 355
Spyridon Lampros, 423–4 Tarán, L., 6, 20, 23, 38, 57, 78–9, 81, 92, 221,
Steel, C., 17, 129–30, 198, 199, 209–10, 229, 248, 256–7, 281, 285, 295, 386
312–13, 348, 513 Tardieu, M., 20, 297, 301–2, 305–9, 311
Steinschneider, M., 24, 314–15 Tatakis, B.N., 252, 430
Stephanus, 15, 18, 20–1, 22, 41, 53–4, 124, Taurus, 97
130, 160, 185, 197, 236, 252, 334–44, taxis tês anagnôseôs, 51
365–6, 385, 430 Teller, P., 28
Stewart, H.F., 377 theia dunamis, 104
Stobaeus, 75, 439 Theiler, W., 84, 121, 139, 219, 327, 519
Stoics, 1–2, 45, 190, 379, 381, 387–8, Themistius, 18–19, 21, 23, 24, 39, 41, 56,
389–90, 408–9, 415–16; doctrine of 82, 92, 99, 119–31, 193, 203, 314, 344,
‘Categories’, 68; earliest commentators, 346, 347–8, 382, 397, 404, 406–7, 414,
68, 71–5, 77–88 ethics, 453–4, 461–2, 486, 506,
475; hostile to Aristotle’s doctrine, 85–6; Theodore of Smyrna, 435
on Aristotle’s God, 196; their criticism Theodosius, Emperor, 12
of Aristotle’s Categories, 77; tripartite Theophrastus, 17, 60, 62, 63, 64, 66, 74, 76,
division of philosophy, 72–3 95, 147, 198, 313–14, 326, 344, 453
Strabo, 67, 68–9, 78 Theophylact, 361, 424
Strange, S.K., 76 Theosebios, 44
Strato, 81 Thesleff, H., 5
Stump, E., 394, 396–7 theurgy, 13
substance, 167–8, 230–1, 274, 286–7; Thillet, P., 23–4, 57, 94, 108
distinguished from the other Thirlkel, W.E., 209
categories, 72–3; hierarchy of Thompson, H.A., 310
substances, 255; Neoplatonist vs. Timothy, of Gaza, 60
Aristotelian, 156, 196–7; primary and Todd, R.B., 3, 16, 71, 74, 86, 89, 98, 101,
secondary, 410 104, 124, 274
suicide, 7, 358, 373 Tornikes, G., 423–35, 465, 467, 472
Sulla, 61, 67, 68, 310–11 Torstrik, A., 35
Sullivan, M.W., 21 Touati, C., 210, 213
Sulowski, 399 Trebonius, C., 70
sunanagnôsis (reading of a text with a Trevor-Roper, H., 401
master), see oral teaching Trifolgi, C., 295
Susemihl, F., 68 tropos tês didaskalias, 53
 General Index 595

Troupeau, G., 200, 202 Wallis, R.T., 125, 189, 330–1

truth, 153, 154–5, 171, 173, 175, 178–85 Waltz, P., 312
Tuilier, A., 259 Walz, C., 320, 322
Tweedale, M.M., 108, 508 Warn, J., 353
Tyrannio of Amisus, 61, 66–9 Wartelle, A., 319
Tzetzes, John, 432 Wegenast, K., 228
Überweg, F., 61, 296 Wehrli, F., 82
weight, 85, 87–8
Ulocrino, 521–4 Weissert, H., 68
universals, 165–7, 170, 171, 173, 380, Wendland, P., 43, 91
409–11, 451; Alexander’s theory of, Westerink, L.G., 5, 20–1, 25, 101–2, 119,
108–9; in mathematics, 501–2, 508 160, 165, 189, 193, 218, 221–3, 229–32,
universe, finite or infinite, 101 234, 236, 241, 245–8, 253, 256–61, 299,
univocal, 380 318–19, 320, 331–2, 349–75
Usener, H., 35–6, 39, 41, 57, 60, 67, 69, Whittaker, J., 309
159, 366 Wieland, W., 253, 253, 283
Uspenskij, F., 435 Wiesner, J., 63, 76, 90, 96, 129, 187, 321
Ustillianus Hesychius, 44 Wildberg, C., 277, 293
utility (see also khreia), 389 Wilpert, P., 328
Wolff, M., 199, 213, 233, 256, 263–4, 271–2
vacuum, 14, 29, 211–12, 213 Wolfson, H.A., 199, 202, 206–8, 210, 287
Vancourt, R., 233–4, 241, 356 Wolska, W., 248, 262, 278, 292
Verbeke, G., 224, 241, 245, 259, 334
Verginius Rufus, 95 Xenarchus, 95
Verrycken, K., 199, 215–50, 251–94 Xenocrates, 4, 72, 78–9
Victorinus, 391, 393, 396, 397 Xenophanes, 96
Vitelli, G. (H.), 19, 98, 263, 264, 268
vocabulary, 127–8, 147, 158–9, 294, 341 Zacharias of Mytilene (Scholasticus),
voice of, from the, 8, 221, 256–7, 351 219–20, 221, 227–32, 237
Vries-van der Velden, E. de, 472 Zeller, E., 56, 61, 66, 69, 73, 83–5, 216, 233,
Vyver, A. van de, 394 353
Zenobius, 101
Wacht, M., 288 Zimmermann, F.W., 23, 27, 202
Wallace, W.A., 27 Zintzen, C., 247, 281, 340
Wallies, M., 251, 262, 395, 397 Zonaras, John, 428
596
597
598
599
600