Plato's Philosophy of Science

This page intentionally left blank

 Plato's
Philosophy of
Science
Andrew Gregory

Bloomsbury Academic
An imprint of Bloomsbury Publishing Pic

B L O O M S B U R Y
L O N D O N • NEW D E L H I • NEW Y O R K • SYDNEY
 Bloomsbury Academic
A n imprint of Bloomsbury Publishing Pic

50 Bedford Square 1385 Broadway

London New York
WC1B 3DP N Y 10018
UK USA

www.bloomsbury.com

BLOOMSBURY and the Diana logo are trademarks

of Bloomsbury Publishing Pic

First published in 2001 by

Gerald Duckworth & Co. Ltd.

© Andrew Gregory, 2000

Andrew Gregory has asserted his right under the Copyright, Designs
and Patents Act, 1988, to be identified as Author of this work.

A l l 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 author.

British L i b r a r y Cataloguing-in-Publication Data

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

ISBN: HB: 978-0-7156-2987-1

ePDF: 978-1-4725-0237-7
ePub: 978-1-4725-0238-4

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress.

Typeset by Ray Davies

 Contents

Acknowledgments ix

Introduction 1
I The development of Plato's views 3
II W h y teleology? 5
III Plato and scientific realism 8
IV Observation and experiment 10
V Geometrical atomism 13

1. Cosmology, C o s m o g o n y a n d Teleology 17
I Plato and modern cosmology 17
II The chaos of the elements 22
III Chaos and cosmogony 24
IV Atomists, physiologoi and cosmoi 26
V Recurring cosmological problems 30
VI Biology and zoogony 34
VII Empedocles 37
VIII The shapes and sizes of atoms 39
IX A teleological cosmos vs. m u l t i p l e entities 42
X Cosmological reasons for teleology 45

2. Astronomy, O b s e r v a t i o n a n d E x p e r i m e n t 48
I Two worlds and investigation 49
II The dynamics of investigation 51
III Vlastos on investigation 53
IV Doing astronomy and teaching astronomy 55
V 'Real'astronomy 58
VI The Republic and other works 61
VII Celestial motion i n the Republic 64
VIII Timaeus 68b-d 67
IX Plato and experiment 70
vi Contents

3. Meno's P a r a d o x a n d U n d e r d e t e r m i n a t i o n 74
I The modern underdetermination problem 75
II The Duhem-Quine thesis 78
III Meno's paradox 80
IV Socrates' autobiography 82
V Teleology and the Phaedo 85
VI H e a l t h y hypotheses 87
VII Teleology and epistemology 90
VIII D u h e m and i n s t r u m e n t a l i s m 93
IX Saving the phenomena 97

4. C e l e s t i a l M o t i o n i n the Timaeus 101

I Disorderly and degenerating cosmologies 101
II The m y t h of the Politicus 103
III Celestial motion i n the Timaeus 105
IV Celestial motion i n later Plato 109
V The Politicus m y t h and the Timaeus 111
VI Reason and necessity 113
VII Implications for causation 115
VIII Political, social and moral decline 118
IX The implications of a stable cosmology 121

5. P l a t o a n d the D e v e l o p m e n t of G r e e k A s t r o n o m y 124
I The Republic and the m y t h of E r 125
II The Timaeus model 128
III Retrogression and deviation i n l a t i t u d e 131
IV Venus, M e r c u r y and the contrary power 136
V Eclipses and occlusions 139
VI Variations i n planetary velocity and distance 145
VII Prototype or finished article? 148
VIII The astronomical and cosmological traditions 153
IX Plato and Eudoxus 155
X Meno's paradox and astronomy 157

6. P l a t o a n d the D e v e l o p m e n t of G r e e k C o s m o l o g y 159
I Plato and the Presocratics 160
II Plato and geocentrism 162
III The Greeks and geocentrism 164
IV The animate heavens 167
V Nomological realism 170
VI C i v i l l a w and physical law 173
 Contents vii
V I I Analogues for physical processes 175
V I I I Plato and subsequent astronomy 178
I X Plato and subsequent cosmology 183

7. G e o m e t r i c a l A t o m i s m - F l u x a n d L a n g u a g e 187
I The receptacle and language 188
II Some paradoxes 191
III Geometrical atomism 194
IV Letters, syllables and triangles 196
V The evidence of Aristotle 200
VI The ageing process 203
VII The gold example 205
VIII Geometrical atomism and Socrates' dream 208
IX Geometrical atomism and language 210

8. G e o m e t r i c a l A t o m i s m - M a t t e r a n d S p a c e 214
I The receptacle and metaphors 215
II The Timaeus and empty space 217
III M a t e r i a l metaphors 218
IV Spatial metaphors 221
V The receptacle and qualities 223
VI Perception i n the Timaeus and Theaetetus 227
VII The evidence of Aristotle 229
VIII Plato, Democritus and Descartes 231
IX Teleology and bonding 235
X The significance of geometrical atomism 238

9. E p i s t e m o l o g y i n the Timaeus a n d Philebus 241

I World soul and h u m a n souls 242
II Souls and common concepts 245
III Souls and epistemology 247
IV Timaeus 27c-29d and two worlds (TW) theory 250
V Two worlds and stability 253
VI W r i t i n g and paidia 255
VII Plato and m y t h 258
VIII I n t e r p r e t i n g the Timaeus 259
IX The Philebus and divine cosmology 262
viii Contents

10. C o n c l u s i o n 265
I The role of empirical w o r k 265
II Plato as a scientific (proto-)realist 266
III Reasons for teleology 267
IV The aims of the investigation of nature 269
V Plato's developing philosophy of science 270
VI Criticisms of the atomists and physiologoi 271
VII Plato relative to the atomists and physiologoi 273
VIII Plato's philosophy of science 274

Notes 275
Bibliography 307
Index Locorum 325
Index of Names 330
General Index 332
 Acknowledgements

A book always owes a great many debts of gratitude, so let me begin w i t h

a blanket t h a n k you to a l l those w i t h w h o m I have discussed Plato and
ancient science over the past few years. Special thanks are due to Bob
Sharpies, Verity Harte, M . M . McCabe and Hasok Chang who have a l l
read, commented on, and discussed parts of this book. I would also like to
t h a n k the ancient philosophy reading group at K C L , and audiences to
w h o m I have given papers, especially at the I n s t i t u t e of Classical Studies
and at U C L .
O n the technical side, Amanda dos Santos has done sterling w o r k i n
preparing many of the diagrams. I would also like to t h a n k Deborah Blake
and Ray Davies at D u c k w o r t h for their patience and efficiency i n prepar­
i n g the book for production.
Last, b u t by no means least, I would like to t h a n k Sheelagh Doherty for
her support and patience d u r i n g the long process of w r i t i n g this book.
W i t h o u t her, and those mentioned above, this would have been a poorer
effort.

Department of Science and Technology Studies A n d r e w Gregory

U n i v e r s i t y College London
December 2000
This page intentionally left blank
For Sheelagh, with love
This page intentionally left blank
 Introduction

Plato's conception of the n a t u r a l world, and how we ought to investigate

and explain i t have provoked some sharply divergent evaluations. He has
received m u c h criticism, the m a i n accusations being t h a t he was antipa­
thetic to the investigation of nature, feeling t h a t i t could produce no
w o r t h w h i l e results, t h a t his views were overly teleological, t h a t his views
were retrogressive relative to the atomists and the other physiologoi, t h a t
his methodology was anti-empirical, t h a t he discouraged careful observa­
t i o n and t h a t he effectively substituted mathematics for physics. Others
have praised h i m , largely for w h a t they see as his role i n the m a t h e m a t i -
sation of nature and of science, while some have sought to strike a balance.
The a i m of this book w i l l be to produce a reappraisal of Plato's philosophy
of science, and to attempt to clarify the basis, nature and justification of
these evaluations. The m a i n arguments of the book, w h i c h I outline i n
more detail below, are as follows:

1. Plato's views on the nature of science, and his views on astronomy

and cosmology i n particular develop, and do so i n interesting ways.
2. Plato can best be seen as someone who is struggling w i t h the
foundations of scientific realism.
3. Plato has interesting epistemological, cosmological and nomological
reasons for teleology, i n addition to any other reasons he may have.
4. The anti-empirical charges against Plato's conception of science are
i l l founded.
5. Plato believes the investigation of nature w i l l produce w o r t h w h i l e
results.
6. Plato has a series of i m p o r t a n t and cogent criticisms of the atomists
and the other physiologoi.
7. Plato's views on science, and his geometrical atomism i n particular,
are undervalued relative to the atomists and physiologoi.

Some comments on m y use of 'science' are perhaps i n order. I n m y view i t

is quite proper to speak of Greek, and indeed Presocratic science. Clearly
there were significant weaknesses, b u t I believe there is enough i n com­
mon between the activities of the Greeks and the moderns to call b o t h
science. I w o u l d argue t h a t the key factor is the Greek conception of the
2 Plato's Philosophy of Science
cosmos and t h e i r insistence on n a t u r a l (as opposed to mythological or
supernatural) explanation. Whether i t is proper to speak of Plato and
science w i l l be a matter for this book. Whether he invokes supernatural
and mythological explanations w i l l be an i m p o r t a n t issue for his cosmo­
logy. W h i l e Plato was clearly not an empirical scientist himself, t h a t is not
a bar to his being a theoretical scientist or philosopher of science. I argue
t h a t he made considerable contributions on both counts. Where Plato fits
into the general development of Greek science and philosophy of science,
especially i n relation to astronomy and cosmology, w i l l be another impor­
t a n t issue for this book. O f course there are weaknesses i n his views too,
w h i c h we w i l l need to consider carefully i n relation to contemporary Greek
science. I shall be interested i n two means of assessing Plato's views on
science:

A. Is t h i s position or strategy relevant to modern debates?

B . Was this view viable and plausible i n the context of fourth-century
Greek science?

C r i t i c a l to A is the attempt to see Plato as the progenitor of scientific

realism. Critical to B is Plato's relation to the atomists and other physiolo-
goi. O n some issues, these strategies may be combined. So Plato's position
may contain a kernel w h i c h is s t i l l significant i n modern debates, b u t
relative to modern views Plato overplays his position (though the overplay­
ing may be defensible contextually). So, for instance, I shall argue t h a t
Plato's strategy for the solution of certain perennial cosmological problems
is v i r t u a l l y identical to some modern strategies, though his actual answers
are distinctly more teleological. T h a t teleology m i g h t be defended as a
historical first instance (no hindsight to judge i t w i t h ) or as being as
plausible i n context as other approaches, or as being constrained by other
areas of Greek thought.
The nature and role of the Timaeus w i l l of course be a major issue. One
sub-theme w i l l be an argument t h a t the Timaeus, i n terms of content, style
and concerns is somewhat better integrated into the mainstream of Plato's
later w o r k t h a n is generally allowed. I n particular I believe there to be
some i m p o r t a n t parallels w i t h the Theaetetus and other late works w h i c h
can i l l u m i n a t e some of the key issues t h a t the Timaeus raises. I shall also
argue t h a t the Timaeus is more of a discussion of how to do science, and
cosmology i n particular, t h a n the statement of a dogmatic position.
As to the dating of Plato's works, I argue t h a t the Timaeus is late and
t h a t i t contributes to an ongoing revision of Plato's thought as p a r t of a
coherent programme w i t h other late works. I would shy away from at­
t e m p t i n g to establish any precise sequence. P a r t l y this is because the
evidence is inconclusive, and p a r t l y because I suspect there is no such
precise sequence. I have yet to see any convincing argument t h a t Plato
wrote his works sequentially and discretely, t h a t he began one work,
 Introduction 3
worked on i t alone, finished t h a t w o r k and only t h e n began the next one.
Indeed I find this rather implausible, and more implausible for the later
1

works, w h i c h are arguably the result of long meditation on the early and
middle period works, and which i n places deliberately evoke and reflect
upon t h e m . F u r t h e r difficulties here are the nature of 'publication' i n the
ancient w o r l d (did Plato revise works once he released t h e m , i f so w h i c h
version do we have?), and the fact t h a t we know l i t t l e of how Plato's works
may have been used i n the academy (were they revised i n the l i g h t of
debate/teaching?). There is also the problem of Platonic anonymity. Plato
wrote dialogues rather t h a n treatises, and i t is not clear to w h a t extent the
views he gives to his characters are expressions of his own views. I do not
believe t h a t we ought give up discussing the development of Plato's
thought, b u t we do need to realise l i m i t s of our conjectures. I do have an
argument t h a t the development of Plato's cosmology would indicate t h a t
the Timaeus, Philebus, Laws and Epinomis form the later stage of his
thought i n this respect, although t h a t needs to be treated w i t h due caution.
I make use of the Epinomis, so a word on authenticity. I accept t h a t the
Laws was 'on the wax' when Plato died and was subsequently 'published'
(and edited?) by P h i l i p of Opus, who also 'published' (and edited/wrote?)
the Epinomis. 2
Quite possibly he had a greater hand i n the creation of the
Epinomis. W i t h due caution, I take i t to be at least indicative of w h a t a
close associate of Plato thought ought to be an appendix to the Laws.
Unless otherwise stated, a l l references to Socrates are to the character
of Plato's dialogues and not the historical figure. The relation between the
two is unclear, and w i l l not be a major concern of this book, though I am
probably more sceptical t h a n most about identifying the opinions and life
history of the Socrates of the dialogues w i t h the historical figure. 3

I . T h e development of Plato's v i e w s
I argue t h a t there is significant development i n Plato's cosmological ideas,
and i t is certainly w o r t h looking i n detail at the models offered i n various
works. The major issue here is whether there is a critical difference
between the views offered by one group of works, the Phaedo, the Republic
(both i n the discussion of astronomy i n Book V I I and i n the M y t h of E r ) ,
and the Politicus, and another group comprising of the Timaeus, Philebus,
Laws and Epinomis. I n the first group the visible heavens are not amen­
able to precise mathematical description and there is a gradual degenera­
t i o n of the cosmos. I n the second, we can have precise mathematical
t r e a t m e n t of the visible heavens and the cosmos is stable. Along w i t h this
cosmological stability, there also appears to be a change i n Plato's attitude
to inevitable political, social and moral decay. Exactly how widespread the
idea t h a t the world, humans and society were caught up i n some slow b u t
inevitable decay was among authors (e.g. Homer, Hesiod), philosophers
(e.g. Empedocles) and the Greek public i n general is difficult to judge, b u t
4 Plato's Philosophy of Science
we may fairly say t h a t i t was a reasonably common notion, and i t is w o r t h
noting Plato's allusions to Hesiod i n his description of political decay at
Republic 546e ff. Later Plato though, I argue, is unequivocal. The cosmos
is entirely stable, and there is no one and nothing to blame b u t ourselves
for political, social and m o r a l decay. Behind this move there may be some
i m p o r t a n t metaphysical and cosmological realignments, and there are
considerable implications for the Timaeus i f at least some physical entities
can behave i n an entirely regular manner. I shall also be concerned w i t h
ways i n which the technai (crafts, arts and skills ranging from cobbling and
b u i l d i n g to astronomy and medicine) and Plato's conception of the good
become more mathematical i n t h e i r orientation. The more Plato's concep­
4

t i o n of the good rests on ideas of unity, mathematical simplicity and

elegance and less on ethical and aesthetical ideas, the closer he w i l l be to
some modern realist positions i n the philosophy of science. B o t h would
agree t h a t we need to employ criteria beyond t h a t of empirical adequacy
i n order to determine w h i c h theory to adopt. Modifications to Plato's views
on these issues have considerable implications for his philosophy of sci­
ence, and later period Plato may well be i m m u n e to many of the criticisms
aimed at a middle period work such as the Republic.
Plato's position i n the development of Greek cosmology is slightly
complex, as there is significant development w i t h i n his own thought. Plato
gives us one of the first accounts of a spherical, unsupported earth i n the
Phaedo, and so, to use Furley's t e r m , is i m p o r t a n t i n the development of
the centrifocal cosmos. Plato t h e n considers the same sort of questions
applied to the cosmos. He moves from a model i n the Republic w h i c h has
external bonds, a pivot on which i t turns and where celestial motion
degenerates to one w h i c h is entirely self-sufficient, stable and requires no
external bonds or pivots i n the Timaeus. This, arguably, is significant,
parallels Presocratic debates about the nature of the earth, and happens
very quickly w i t h i n Plato. Plato's contribution to Greek cosmology i n this
respect may be substantially underrated. Another matter t h a t w i l l require
careful consideration is the fact t h a t i n the Timaeus Plato introduces the
idea t h a t the celestial bodies are animate, ensouled, i m m o r t a l deities, an
idea w h i c h differs significantly from the Republic, and from the Politicus
where the cosmos is specifically animate b u t m o r t a l when god (Cronus)
relinquishes control. W h y does Plato w i s h to do this, and w h a t sort of step
is this i n the history of Greek cosmology? One more general point t h a t I
w i l l emphasise is t h a t Greek cosmology i n general was considerably
affected by, and indeed perhaps shaped by the absence of a conception of
gravity and indeed of force i n general. Plato introduces a like-to-like
principle operating at the level of the elements (earth, water, air, fire) i n
the Timaeus. I t is highly i m p o r t a n t to understand w h a t this principle can
and cannot do. I t w i l l account for some terrestrial gravitational effects, b u t
w i l l not help w i t h the motions of the celestial bodies or w i t h cosmology
more generally. I t is also v i t a l for understanding Plato's cosmogony to
 Introduction 5
analyse very carefully w h a t this principle is and is not capable of i n the
chaos of the pre-cosmos. Plato's conception of the pre-cosmos and its
t r a n s i t i o n to a cosmos w i l l be an i m p o r t a n t issue i n Plato's attack on the
atomists and i n understanding his reasons for advocating teleology
The key development i n astronomy w i t h i n Plato is the i n t r o d u c t i o n of
regular circular motions for the sun, moon and planets offset from those
of the fixed stars i n such a way as to begin to represent t h e i r motion along
the ecliptic and t h r o u g h the zodiac. This we find i n the Timaeus and
subsequent works b u t not i n previous works. Thus the planets are not
really 'wanderers' (or as the Babylonians referred to them, bibbu, 'sheep')
but move i n a regular manner. I f so, t h e n i t is Plato who provides the
theoretical basis for concentric (or homocentric) sphere astronomy and as
such facilitates an enormous leap forward for Greek astronomy. We w i l l
need to discuss the respective contributions of Plato and Eudoxus care­
fully, as this is a matter of some controversy, as is the nature of the
challenge w h i c h Plato is alleged to have issued to those engaged i n these
studies as reported by Simplicius:

Which hypotheses of regular, circular and ordered motion are able to save
the phenomena of the planets? 5

This I w i l l refer to as the Platonic challenge. A n i m p o r t a n t issue here is

whether Plato considered the model of the Timaeus to be a prototype
embodying the i m p o r t a n t principle of regular and ordered circular motion
(or combinations of such motions) for each of the celestial bodies, or
whether i t is supposed to be a dogmatic and finished model. T h a t w i l l form
p a r t of a larger debate on exactly how we should treat the Timaeus.

I I . W h y teleology?
A m a i n area of criticism of Plato has been his use of teleology. Indeed, at
times the Timaeus has been described as a 'teleologist's manifesto'. A
standard view from the history of science would have i t t h a t only w i t h the
revival of the views of the ancient atomists by such seventeenth-century
t h i n k e r s as Descartes and Gassendi d i d science free itself from the restric­
tive tentacles of the teleology of Plato and Aristotle. Since t h e n science has
progressed rapidly, relying on a mechanistic conception of the w o r l d i n
w h i c h there is no place for teleology. Even Vlastos, who tries to see good i n
the Timaeus where he can, criticises Plato severely on this matter, and
6

speaks of:

The retrograde turn which Plato gives to cosmological inquiry when he

converts so blatantly preconceptions of value into allegations of fact. 7
6 Plato's Philosophy of Science
Vlastos is very positive about the nature of the Platonic challenge, b u t
comments:

What emboldened Plato to start that tradition was a metaphysical fairy

tale.
8

W h y t h o u g h does Plato feel t h a t he requires this teleology? There are

several answers here t h a t are uncomplimentary to Plato, not least t h a t
this is some form of overspill from Plato's programme i n ethics. G r a h a m
comments:

Plato's demand for teleological explanation in natural philosophy has roots

in Socratic ethics. 9

Perhaps t h o u g h Plato has some independent philosophical reasons about

the nature of enquiry, cosmology and regular behaviour. These are difficul­
ties w h i c h face modern philosophy of science too, and Plato's strategy, and
perhaps his answers too may well have some modern relevance. Zeyl
comments:

The metaphysics of Plato's middle period, for example, is now generally

appreciated as a serious and largely coherent attempt (whether mistaken or
not) to answer genuine and important questions about the nature of knowl­
edge and reality.10

One theme of this book is to attempt to show t h a t Plato's teleological

approach to the investigation of nature is a serious and coherent attempt
to answer genuine and interesting problems i n the philosophy of science.
I n modern philosophy of science there is a problem k n o w n as underde-
t e r m i n a t i o n . I n short, i n order to determine which theory to adopt the data
(however good) are insufficient and non-empirical criteria must be em­
ployed. I argue t h a t Plato recognised some of the difficulties here, albeit i n
a rather more general form, i n Meno's paradox. Plato t h e n develops a
solution involving teleology, but one which nevertheless contains certain
i m p o r t a n t affinities w i t h modern realist attempts to solve the underdeter-
m i n a t i o n problem by applying (at various stages of Plato's development)
criteria such as beauty, simplicity and unity. This is i m p o r t a n t i n itself, and
also for understanding how Plato goes about solving certain problems i n
astronomy and cosmology i n the Timaeus, especially those problems where
he argues for a single solution from an indefinite field of possibilities.
I also argue t h a t Plato uses teleology for a second epistemological
problem. I n the Socrates' dream passage of the Theaetetus, he argues t h a t
composite entities can be analysed into t h e i r parts, and so we can have
accounts of them. However, t h a t which is incomposite cannot be given the
same sort of account, and on this conception of knowledge is unknowable.
E m p l o y i n g one of his favourite analogies, Plato argues t h a t we m i g h t have
 Introduction 7
an account of syllables ( i n terms of the letters w h i c h constitute them) b u t
not of letters. L i k e Meno's paradox, this is presented quite generally, b u t
one problem for science here m i g h t well be w h a t sort of account can we
have of the u l t i m a t e parts of the physical world? Clearly we cannot account
for t h e m or analyse t h e m i n terms of f u r t h e r physical parts. The
Theaetetus is happy j u s t to present the problems. I n the Timaeus, however,
we are given the theory of geometrical atomism, couched i n the same
letters and syllables terminology. Here though, at the end of analysis by
decomposition (into atoms, into atomic planes, into the basic triangles) we
have further teleological accounts of the nature of the basic triangles.
Clearly this is a contentious area and we w i l l need to look carefully at the
nature of geometrical atomism, the receptacle and the sort of explanations
envisaged.
Plato also has cosmological reasons for teleology A central question of
cosmology is w h y the w o r l d about us exhibits so much order. The idea t h a t
there is one cosmos, and t h a t cosmos organises itself, is i n fact remarkably
rare i n the history of cosmology, and, I argue, is absent from ancient Greek
thought altogether. W h e n this idea has been rejected, there tend to be two
different types of answer. One can suppose there to be indefinitely many
differing worlds/universes, such t h a t the order of our world/universe is
explained as an inevitable instance among an indefinite range of possibili­
ties. So the ancient atomists, and so the influential modern view t h a t there
may be many universes (the modern t e r m is a 'multiverse'). I f one feels
t h a t a many worlds theory is ontologically rather profligate, or fails to be
a proper explanation ( i n m y own view i t fails on both counts; I argue t h a t
is Plato's view too), one m i g h t argue for a unique cosmos w i t h some extra
explanatory principles i n play. So Plato, and indeed Aristotle, so various
versions of the modern cosmological anthropic principle, and so Christian­
ity. The latter is very i m p o r t a n t historically The seventeenth century does
not simply revive the views of Leucippus and Democritus, b u t transforms
t h e m w i t h i n a C h r i s t i a n context. The problems w h i c h Plato highlights and
solves w i t h teleology are then solved by the existence and actions of a
C h r i s t i a n god, who w i l l n a t u r a l l y act for the best. As we shall see, there
are certain recurring types of problem i n cosmogony and cosmology and
certain recurring patterns of strategies for t h e i r solution, r i g h t down to the
current day. Plato, I argue then, investigates a perfectly reasonable option
i n the face of the perceived difficulties of a materialist account of a u n i t a r y
cosmos or the m u l t i p l e worlds account of the atomists. I also contend t h a t
Plato, contra the atomists and other physiologoi, argues t h a t not only is
there a unique cosmos b u t also t h a t there are a small and well defined
number of types of u l t i m a t e particle and well defined, permanent biologi­
cal species.
I f this is so, t h e n we can say some rather more positive things about
Plato's teleology. First, i n the context of fourth-century Greece, this is a
plausible and viable project. I t is by no means clear i n t h a t context t h a t
8 Plato's Philosophy of Science
the atomist/physiologoi approach is correct, (indeed i t would appear t h a t
there were major flaws i n i t ) , or t h a t Plato's approach is incorrect. W h i l e
this at times has been rather tentatively and grudgingly recognised, one 11

of the aims of this book w i l l be to argue t h a t the ideas of the atomists and
physiologoi were rather less plausible and viable t h a n is generally sup­
posed, those of Plato rather more so. Secondly, Plato is addressing serious
problems i n the philosophy of science, problems which are s t i l l under
debate today. W h i l e modern realists would see Plato as overly teleological,
they would also recognise many structural similarities w i t h t h e i r own
views i n the way t h a t he attempts to solve these difficulties.

I I I . P l a t o a n d scientific r e a l i s m
I argue t h a t the most interesting way of looking at Plato is as someone who
is struggling w i t h the foundations of scientific realism. The Dictionary of
the History of Science defines realism as follows:

Scientific realism is the thesis that the objects of scientific knowledge exist
and act independently of knowledge of them. More generally 'realism' asserts
the existence of some disputed kind of object (e.g. universals, material
objects, scientific laws, propositions, numbers, probabilities). ' 1 2

Or more bluntly, as Musgrave puts i t :

Realists think that scientific theories are true or false descriptions of real­
ity.
13

Such realism may be contrasted w i t h nominalism (either specifically

denying the existence of universals or more generally denying the exist­
ence of disputed objects) and i n s t r u m e n t a l i s m , where theories are i n s t r u ­
ments for calculation b u t no question arises as to whether such theories
are true or false, as many means of calculation may be devised to produce
the same result (see below). 14

Realist positions require t h a t we make and are able to justify certain

assumptions about the nature of the world and the nature of investigation.
Whether these assumptions can u l t i m a t e l y be justified is not a question I
shall be concerned w i t h here, nor shall I argue t h a t Plato was himself a
realist, as such labels are unhelpful when dealing w i t h the ancients. There
are significant differences between Plato's views and those of modern
realism, and ancient debates around these issues are not set up i n quite
the same way nor conducted i n the same terms. Rather, I shall argue t h a t
Plato is concerned w i t h exploring some of the assumptions t h a t are
required to get realism off the ground, and is perhaps the first person to
do so i n a comprehensive manner. The point of this approach is not to
assimilate Plato to modern philosophy of science, but rather to narrow
w h a t some commentators have argued is a y a w n i n g chasm between the
 Introduction 9
two, and to t r y to understand w h y Plato approaches certain questions i n
w h a t m i g h t otherwise seem to be an inexplicable manner. If, i n a positivist
frame of m i n d , reckoning science to be a matter only of empirically
verifiable propositions, we dismiss Plato's discussions of some of the
central questions for a realist view of science as irrelevant or metaphysi­
cal, we risk missing much of philosophical and historical interest.
I f i t is the case t h a t Plato is a proto-realist, t h e n i t cuts against view
t h a t Plato was a (proto-) instrumentalist. I n s t r u m e n t a l i s m can be defined
i n the following way:

Instrumentalists assume that theories have the status of instruments, tools,

or calculating devices in relation to observation statements. In this view i t
is assumed that theories can be used to relate and systematise observation
statements and to derive some sets of observation statements (predictions)
from other sets (data): but no question of the truth or reference of the
theories themselves arises. 15

D u h e m has argued t h a t Plato began a t r a d i t i o n of i n s t r u m e n t a l i s m i n

astronomy. H i s view is that:
16

The hypotheses of astronomy can be viewed as mathematical fictions which

the geometer combines for the purposes of making the celestial motions
accessible to his calculations. 17

Thus when Plato issues the challenge of 'Which hypotheses of regular,

circular and ordered motion are able to save the phenomena of the plan­
ets?', D u h e m asserts t h a t this is an instrumentalist move which begins a
t r a d i t i o n whose sole a i m is to save the phenomena. 1 shall argue t h a t this
18

view is quite wrong, at least as far as Plato is concerned.

Another issue on which I believe Plato comes down very f i r m l y on the
realist side is t h a t of laws and regularities. One problem here is t h a t Plato,
and indeed the Greeks i n general are lacking a proper conception of
physical law. A g a i n I t h i n k we can see Plato struggling w i t h the founda­
tions of a realist account here as he has, at various times, three different
accounts of w h a t generates regular behaviour. A n added complication here
is t h a t certainly as far as the heavens are concerned, Plato's views on
whether there is perfectly regular behaviour change between the Republic
and the Timaeus. I t is perhaps i m p o r t a n t to note here t h a t Vlastos believes
that:

Most of us have been brought up to think that the laws of nature are in the
last analysis radical contingencies - de facto uniformities which we must
either exhibit as special cases of still more general de facto uniformities or
else accept as things for which no further reason can be given. 19

This is perhaps p a r t l y w h y Vlastos sees Plato's efforts as a 'metaphysical

fairy tale'. B u t modern realists, and i n m y view Plato, would disagree w i t h
10 Plato's Philosophy of Science
Vlastos on n a t u r a l law. I f we see Plato as someone who is struggling w i t h
the foundations of realism though, we can begin to appreciate w h y Plato
requires teleology. A n i m p o r t a n t issue here is w h y the heavens move i n a
regular and orderly manner. Plato's answer is to a t t r i b u t e life and i n t e l l i ­
gence to the heavenly bodies and to the cosmos itself. F r o m the modern
point of view, the lines of criticism here appear evident, and Vlastos is
critical of Plato's 'psychokinetic' forces. Yet this too bears some deeper
analysis. A further effect of the lack of a conception of gravity or force i n
Greek thought is t h a t they struggle to give a mechanical account of the
motions of the heavens. A l l i e d to t h a t is the lack of a conception of physical
law. C i v i l law requires people to use their intelligence to act for the good.
So Plato employs the alternative of biological and civil law analogies for
both the motions of the heavens and for the cosmos as a whole. Intelligence
exhibits regularity, and w i l l choose the best, souls are the source of motion,
l i v i n g creatures have a coherence and organisation, so perhaps the heav­
ens have intelligent souls and the cosmos is a l i v i n g creature. I t is
i m p o r t a n t to note t h a t this is a l l Plato requires and a l l he supposes. I shall
argue he strips away a l l superfluous animistic features of these analogies
to be left w i t h motion, intelligence and coherence. I f so, t h e n these expla­
nations look rather different, and again need to be looked at i n the l i g h t of
Plato struggling w i t h some of the problems of giving a realist account. I t
is not enough for Plato t h a t the heavens move i n a regular manner, or t h a t
the cosmos has coherence and organisation, there must be real reasons
w h y these things are so. So perhaps Plato has nomological reasons for his
teleology as well as epistemological and cosmological reasons. Recognition
of this is i m p o r t a n t i n understanding w h y he approaches problems i n the
investigation of nature i n a certain manner, and for appreciating the full
i m p o r t of his criticisms of the atomists and the physiologoi. I t also allows
us to examine w h a t resources were available to Plato i n answering these
questions, and how t h a t may have helped to shape his replies. A further
issue to be discussed here is whether i t is evident to the Greeks of the
fourth century t h a t one should use only mechanical and not biological or
other metaphors to describe the cosmos. I argue t h a t i t is not, t h a t this is
a piece of hindsight available only after the seventeenth century (if at all),
and t h a t biological metaphors may have appeared at least as plausible and
viable as mechanical ones.

IV. O b s e r v a t i o n a n d e x p e r i m e n t
Perhaps the most common complaint against Plato has been either t h a t he
denigrated observation or even t h a t he banned i t entirely from the sort of
enquiries he envisaged. K n o r r comments:

In view of the subordination of the phenomenal and the empirical to the

 Introduction 11
abstract and ideal, one would not expect Plato's thought to be the foundation
for a robust programme of study. 20

The evidence for such a view is supposed to be strongest i n the Republic,

and at 530b i n particular, where Plato is often interpreted as saying t h a t
we should treat astronomy i n the same manner as we do geometry and so
we should downplay the importance of, or even ignore observations of the
visible heavens, i n the same way t h a t geometers treat diagrams as imper­
fect illustrations of the true realities. So Hetherington comments:

Plato's disdain for observation, his near contempt for observation, is readily
apparent in his comments on astronomy. 21

However, I argue t h a t the allegories of line and cave suggest a m u c h more

dynamic approach to investigation, one t h a t begins w i t h the w o r l d about
us and involves the i n t e r r e l a t i o n of sensibles and intelligibles, the inves­
tigation of each being necessary but not sufficient for us to master a given
discipline. One i m p o r t a n t implication of this is that, given a t w o worlds
(TW) view of the Republic, there is no demarcation for investigation
between the two realms. I take the T W interpretation of Plato to be the
idea t h a t four distinctions are co-extensive, and divide w h a t there is into
two disjoint sets. O n the one hand there is t h a t which is intelligible, stable,
knowable and incorporeal, while on the other is t h a t which is sensible,
changeable, opinable and physical. The first set of entities, the forms,
are said to be, w h i l e the second, comprising the n a t u r a l w o r l d , are i n a
state of becoming.
I f there is no demarcation for investigation between the two worlds,
t h e n this serves to undermine the notion t h a t Plato's science deals solely
w i t h intelligible objects and so eschews observation, and the view t h a t i t
deals solely w i t h physical entities and so can a i m no higher t h a n opinion.
I argue t h a t the passage at Republic 529d ff. illuminates a distinction
between doing and teaching astronomy, and assumes t h a t doing astro­
nomy necessarily involves an empirical investigation, or at least an
acquaintance w i t h the data. I t would seem from comments i n the Epino-
mis a n d Timaeus t h a t Plato was indeed acquainted w i t h E g y p t i a n
astronomical data. I also argue t h a t Plato's middle period attitude to the
regular behaviour of physical bodies does not entail t h a t he would have
been dismissive of detailed observational work.
I f these arguments are correct, we need to assess the Republic's views
on investigation (as opposed to the education of the guardians) i n terms of
a necessary i n i t i a l empirical approach followed by a move to the consid­
eration of the relevant intelligibles as part of the same discipline. This
gives i t a potentially much more reasonable methodology, observation
followed by abstraction and mathematisation, t h a n has been allowed for
i n some of the literature. We m i g h t be critical though of the way i n w h i c h
12 Plato's Philosophy of Science
hypotheses are subjects for discussion, not empirical testing, even i f they
are i n i t i a l l y based on observations. One m i g h t w e l l argue though t h a t
w h i l e Plato does not at any stage ban observation from investigation, his
attitude to observation gradually becomes more positive, and t h a t is
perhaps shown by his comments on sight and astronomy at Timaeus 47a
ff. This approach helps to solve an i m p o r t a n t problem, w h i c h Musgrave
formulates as follows:

How are we to reconcile a Plato concerned to save the astronomical appear­

ances with the anti-empiricist Plato of philosophical legend? 22

I hope to show t h a t Plato's alleged anti-empiricism i n the sciences has been

exaggerated. There is also the much quoted passage of Timaeus 68d where
Plato has Timaeus say, i n the context of his theory of colours, t h a t i t w o u l d
be useless to make any experimental test of 'these things'. I f these things
are the way t h a t mixes of pigments make colours, t h e n the position looks
odd and perhaps damaging. However, i f the reference is to the proportions
of the various types of fire particle w h i c h constitute a colour, we get a m u c h
better sense for the passage and the results are sensible and mundane.
This i n t e r p r e t a t i o n also accords w i t h Plato's consistent conception of the
technai i n other works where one can learn some basic things by t r i a l and
error, and the supposed attack on experiment is unparalleled elsewhere i n
the Timaeus or Plato. The nature of the technai w i l l also be an i m p o r t a n t
issue, p a r t i c u l a r l y as Plato draws a distinction between w h a t can be learnt
by empirical knack and a form of t r i a l and error, b u t is s t i l l necessary for
further investigation, and the proper analysis of the subject m a t t e r along
w i t h the use of mathematics. I t is difficult to give a precise t r a n s l a t i o n of
Plato's use oftechne as the E n g l i s h 'art' or 'skill' do not give a sharp enough
contrast between k n o w i n g the nature of something and merely h a v i n g
some experience, while 'science' is rather presumptive, not least because
the technai are broader t h a n the modern connotation of science, though
the German 'Wissenschaft' m i g h t be closer.
Does Plato believe t h a t w o r t h w h i l e results w i l l come from the investi­
gation of nature? I n the Timaeus, the answer to this has to be yes. A n
analogy is developed between the w o r l d soul and our own, where the
revolutions of the w o r l d soul are stable and clear where ours are chaotic
and confused. One key way of stabilising these revolutions and so becom­
ing like god is by doing astronomy, and more generally by gaining control
over our perceptions and emotions. W h a t sort of account we can have of
the physical world? Even i f we take the epistemology of Timaeus 27c-29d
at face value, i t is s t i l l possible to make out a good case t h a t the investiga­
t i o n of nature w i l l produce some w o r t h w h i l e results. As reason has w o n at
least a p a r t i a l victory over necessity, the w o r l d has some order and
comprehensibility to i t . While we may only have opinion of the physical
w o r l d , Timaeus often claims the account to be the best possible, and of
 Introduction 13
course the physical w o r l d itself is the best possible. Timaeus 68e ff. also
tells us t h a t we must seek the necessary explanations (aitiai) for the sake
of the d i v i n e , for w i t h o u t the former we w i l l not be able to come to know
23

the latter, so the investigation of the physical w o r l d has some purpose and
w o r t h w h i l e result. The fact t h a t Plato attempts a thoroughgoing teleologi-
cal treatment of the physical realm should convince us of his seriousness
i n this matter. So too the dynamics of ascent and descent of the line
allegory would suggest some degree of seriousness and o p t i m i s m towards
the physical world. Later i n this book I w i l l argue t h a t the Timaeus has
rather more to say on the subject of epistemology and t h a t we should at
least be suspicious of a t t r i b u t i n g the views of 27c-29d directly to Plato. I f
the heavens are stable, and amenable to precise mathematical description,
t h e n i n astronomy at least some more optimism may be justified. Clearly
the status of the likely, mythical story of the Timaeus w i l l be an i m p o r t a n t
matter here. I t is i m p o r t a n t to point out though t h a t a good deal of the
astronomy and cosmology of the Timaeus is repeated i n the Laws and
Epinomis w i t h o u t the likely account qualifiers.

V. G e o m e t r i c a l a t o m i s m
The view t h a t regular and stable celestial motion is fundamental i n the
Timaeus m i g h t be thought to r u n contrary to some other themes i n the
Timaeus. One m i g h t argue t h a t because of the action of the 'wandering
cause', and because reason only scores a p a r t i a l victory over necessity,
there could never be completely regular behaviour. I argue though t h a t
there is a perfectly viable reading of the relation between reason and
necessity w h i c h allows such regular behaviour. One m i g h t also question
how the heavens can be stable given w h a t the Timaeus has to say about
flux. Here we come to w h a t w i l l be a central issue i n the l a t t e r part of his
book, for I w i l l argue t h a t there is a reading of the Timaeus w h i c h allows
for stability at a fundamental level.
Plato puts forward the theory of geometrical atomism. Clearly these
atoms are not stable, as they can be decomposed into two-dimensional
faces, and the faces i n t u r n can be decomposed into the two basic sorts of
triangle. I argue t h a t there is a perfectly viable reading of the Timaeus
whereby these basic triangles do not undergo any change other t h a n
bonding w i t h each other i n to complexes and being dissolved from com­
plexes. I f so, t h e n there is a fundamental level of stability, and v a r y i n g
levels of stability of complexes of the basic items, from the most transient
of phenomena t h r o u g h to, as I argue Plato makes clear, the eternal
stability of the heavens. I f t h a t is so, t h e n there are further i m p o r t a n t
issues to investigate. First, i f the basic triangles do have this sort of
stability, how does this affect certain other passages i n the Timaeus, most
notably those to do w i t h the receptacle? I argue t h a t i t is quite possible to
get viable readings of such passages which f i t w i t h the idea of the stability
14 Plato's Philosophy of Science
of the basic triangles. W h a t though of the idea of a radical Heraclitean flux
as mentioned by the Theaetetus and Cratylus? I t is by no means clear t h a t
is Plato's position i n those works, and the geometrical atomism of the
Timaeus may be an attempt to account for the apparent flux of the
phenomenal w o r l d (and the apparent stability of the heavens) i n terms of
some basic and stable items w h i c h w i l l allow a coherent theory of the w o r l d
about us. W h i l e the phenomena may be i n a state of flux, the imperceptible
basic entities of geometrical atomism may not be. As indicated earlier, we
may not be able to give an account of the basic entities of geometrical
atomism i n terms of anything more basic, but we can have a teleological
account of them.
I take a fairly radical line on the question of the relation of geometrical
atomism to the receptacle. Most literature on the question of the u l t i m a t e
constituents of the w o r l d i n the Timaeus either treats the receptacle
passages (48e-53b) as standing alone, or attempts to interpret geometrical
atomism i n terms of the results of an analysis of these passages. I believe
interesting results may be obtained by a different approach, t h a t of looking
first at the possible advantages of geometrical atomism and t h e n r e t h i n k ­
ing the passages on the receptacle i n these terms. I argue t h a t there are
paradoxes and other difficulties w i t h the receptacle w h i c h can be solved by
paying close attention to the way i n which geometrical atomism is intro­
duced and subsequently employed. I t is notable t h a t Plato expends
considerable t i m e and effort i n explaining the structure of the world, our
bodies and our perceptions i n terms of geometrical atomism.
One theme of this book w i l l be Plato's criticism of the atomists and
physiologoi either for being indeterminate i n t h e i r views or for not being
able to justify t h e i r views properly. The atomists, for instance, are not
determinate on the issue of the shapes of fundamental particles and do not
justify t h e i r claim t h a t a vortex w i l l be formed. Other theories (perhaps
Empedocles) postulate basic entities which we can see changing and so are
not stable. The task of geometrical atomism i n alliance w i t h teleology may
be to postulate stability at a level below t h a t of h u m a n perception, to be
able to justify the nature of the fundamental particles and to be able to
account for how the cosmos gained its order w i t h o u t recourse to chance and
m u l t i p l e cosmoi.
This may bear on the seriousness of the Timaeus. Plato clearly has
many philosophical objections to the atomists and the physiologoi. The
Timaeus may set out to show t h a t a teleological geometrical atomism is
both necessary and adequate for the explanation of physical phenomena.
I f so, the Timaeus would t h e n be an i m p o r t a n t w o r k where Plato objects to
and seeks to replace the views of his predecessors.
There are of course other things to say about the nature of geometrical
atomism. Vlastos is quite enthusiastic about the possibilities for quantifi­
cation and establishing, at least i n outline, equations for changes. I would
prefer to say t h a t Plato geometrises matter (no small achievement i n itself)
 Introduction 15
and makes i t amenable to mathematical treatment w i t h o u t ever exploit­
i n g t h a t w i t h a quantified physics or chemistry or indicating t h a t such an
exploitation should be made. I t is often said t h a t i t is the atomism of
Leucippus and Democritus t h a t such seventeenth-century t h i n k e r s as
Descartes, Gassendi and Boyle revive. T h a t stands i n need of considerable
qualification, however. As noted, there are several issues on w h i c h the
ancient atomists are either indeterminate or produce no justification for
t h e i r views. Plato employs teleology to come to a satisfactory resolution.
Those who revive ancient atomism i n the seventeenth century do not
accept i n d e t e r m i n i s m or a lack of justification. Rather, they invoke (the
goodness of) a C h r i s t i a n god to solve these problems. Clearly there w i l l be
similarities and differences w i t h Plato's means of dealing w i t h these
problems, b u t i n employing such solutions, the seventeenth century dis­
tances itself from ancient atomism.
I t is also w o r t h discussing how Plato's atomism is rather more like
modern atomism t h a n t h a t of Leucippus and Democritus. Plato is strong
on the idea t h a t there are a small number of types of u l t i m a t e particles,
w h i c h are mathematically well defined, against the indefinite number of
shapes and sizes of the atomists. The discussion of geometrical atomism
stresses t h a t matter has deep structure i n the sense of u l t i m a t e particles
forming structures w h i c h i n t u r n themselves form structures and so on.
A l t h o u g h Plato is not specific about how bonding between particles occurs,
he is r i g h t t h a t this is not an accidental matter based on mechanical
interaction but happens i n a specific and well defined manner. This of
course is a somewhat Whiggish approach to assessing Plato's geometrical
atomism, b u t given the praise t h a t the ancient atomists have received
precisely along those lines i t is w o r t h pointing out w h a t we m i g h t be able
to say about Plato too. I f we are going to give the ancient atomists credit
for i n this manner, t h e n we must also give credit to Plato. Perhaps more
importantly, Plato's atomism, like his cosmology, is perfectly viable i n the
context of fourth-century Greece. I t is aimed at solving certain problems
w i t h the accounts of the atomists and physiologoi, and is no less plausible
as an account of the physical world.
This page intentionally left blank
 1

Cosmology, Cosmogony
and Teleology

The task of this chapter w i l l be to argue t h a t Plato had reasons concerning

cosmology and cosmogony for employing teleology. I m p o r t a n t here is a
question w h i c h recurs throughout the history of science, indeed r i g h t down
to the present day. Can one explain the apparent order of the cosmos solely
i n terms of matter and motion, or matter, motion and forces? Clearly
Plato's answer to this is no, and we shall need to look at precisely w h y he
adopts t h a t position, and whether this is a reasonable position to adopt i n
the context of fourth-century Greece. As the modern answer to this ques­
t i o n is also no (though for different reasons) i t is also useful to compare the
ancient and modern debates. As we shall see, at least thematically there
are sets of recurring answers too, and some interesting parallels between
modern and ancient answers to this question. This is i m p o r t a n t as I w a n t
to argue for two m a i n points i n Plato's favour here. First, the strategy
w h i c h Plato uses to approach certain questions i n cosmology and cosmog­
ony is s t i l l relevant to modern debates. Secondly, not only does Plato have
i m p o r t a n t criticisms of the atomists and other physiologoi, b u t his strategy
is at least as plausible and viable as these alternatives i n the context of
fourth-century Greece.

I . P l a t o a n d m o d e r n cosmology
A comparison between a modern view of the evolution of the cosmos and
Plato's view of the ordering of p r i m o r d i a l chaos is instructive. According to
modern cosmology, the universe began w i t h the b i g bang around 15 b i l l i o n
years ago. After the i n i t i a l expansion there is a period of r a p i d inflation,
1

followed by a calmer period of expansion. A t this t i m e there are the

fundamental particles and there is a great deal of radiation, b u t none of
the elements have formed as yet. This 'chaos' gradually sorts itself as the
universe expands and cools. F i r s t hydrogen and h e l i u m nuclei form, and
t h e n m a t t e r and r a d i a t i o n decouple, w i t h electrons b i n d i n g to the nuclei
of the l i g h t elements to form the first atoms. I t is from this period, about
300,000 years after the b i g bang, t h a t we can find evidence of the cosmic
background radiation. While this is i n remarkable accord w i t h predictions,
18 Plato's Philosophy of Science
there are slight inhomogeneities. These small variations i n the intensity
of the cosmic background radiation, 'ripples', 'wrinkles i n time' or whatever
one wishes to call them, are the seeds for the future development of the
cosmos. Clusters of matter form from these ripples and eventually create
stars and galaxies. Stars b u r n their fuel of hydrogen and h e l i u m by
nuclear fusion creating the even-numbered elements up to and including
iron, and w h e n these stars go supernova the other odd-numbered elements
and the elements heavier t h a n i r o n are formed. O u r solar system t h e n
forms out of the detritus of such stars. W h e n sufficiently complex organic
chemicals have formed, life gets under way and evolves, w h i c h brings us
to the present day.
A t least at first examination (as we shall see, the situation is somewhat
more complex), this account is one t h a t relies solely upon 'chance and
necessity', as Plato would p u t i t , and requires no other factor to account for
the current order of the cosmos. I t is critically dependent upon the attrac­
tive force of gravity w h i c h leads to an accretion of order around the i n i t i a l
inhomogeneities. Plato's view of the t r a n s i t i o n from chaos to order is of
course very different. I n places his view of matter i n motion prior to the
ordering of the cosmos is rather like our modern conception of the motions
of gas molecules i f we assume t h a t there are no attractive or repulsive
forces between them. The motions of the molecules are random, or as Plato
puts i t at Timaeus 30a:

For god desired that all things should be good, and nothing paltry (phlauron)
as far as was possible, he took over all that was visible, which was not at rest
but in discordant and disorderly motion (alia kinoumenon plemmelos kai
ataktos), and led it into order out of disorder, judging the former to be entirely
better than the latter.

The key assumption for Plato from a cosmological perspective is t h a t chaos

is non-progressive. Left to its own devices chaos w i l l continue to be chaos
and order w i l l not grow out of i t . I f some order should come about by
chance, i t is likely to dissolve immediately rather t h a n breed further order
as i t does i n the modern view of the evolution of the cosmos. Therefore
something else is required to explain how the cosmos acquired t h a t order.
I take t h a t to be a perfectly plausible position to adopt i n an ancient
context. This needs to be distinguished from the modern view, where i t is
possible for an ordered state to come about w i t h randomly m o v i n g p a r t i ­
cles and no forces, b u t highly unlikely. Indeed as there are vastly more
disordered states t h a n ordered ones, the system w i l l tend towards maxi­
m u m disorder. Plato's view of chaos would seem to be t h a t i t neither
progresses to greater order nor regresses to greater disorder.
The Timaeus gives us a like-to-like principle, whereby s i m i l a r sorts of
entities, p r i m a r i l y the elements of earth, water, air and fire congregate
together. So at Timaeus 63c we are told that:
 1. Cosmology, Cosmogony and Teleology 19
Standing on the earth, and separating off types of earth, and sometimes
earth itself, i f we draw them into the unlike air by force and contrary to
nature, both cling (antechomena) to their kind. The smaller is more readily
forced into the dissimilar than the larger.

This is repeated i n a more generalised form at 63e:

The thing to be recognised in all of these cases is that it is the motion of each
towards its own kind that makes the moving thing heavy, and the place into
which it moves 'down'. 2

I n addition to the like-to-like principle, there are two sources of motion. A t

Timaeus 52d f f , we are told t h a t i n the pre-cosmos chaos the receptacle,
because i t is filled w i t h powers t h a t are not balanced or evenly distributed,
is itself shaken by these things and by this motion shakes t h e m i n t u r n . I t
thus acts like a w i n n o w i n g basket (plokanon), 3
and j u s t as the process of
w i n n o w i n g sorts the corn from the chaff, so the four elements are sepa­
rated out, the most u n l i k e furthest away and the most like closest to­
gether. A t Timaeus 57c and 88de we are told t h a t this shaking continues
4

i n the cosmos. A t Timaeus 58a ff. we are told of the compressive effect
w h i c h the rotation of the cosmos has on the elements, m a k i n g t h e m
intermingle and denying the possibility of a void on a large scale. W h i l e 5

the like-to-like principle and the sources of motion w i l l be used to account

for t e r r e s t r i a l gravitational effects, this is very far from a n y t h i n g like a
modern theory of gravitation. There is no universal attraction of m a t t e r to
all other matter. E a r t h w i l l congregate w i t h earth, b u t not the other
elements. N o r is there any force which operates at a distance. S i m i l a r l y at
Timaeus 80bc electricity and magnetism are explained i n a general man­
ner as due to contact action and m u t u a l replacement, and there is an
outright denial t h a t any attraction (holke, 80c3) is involved. Thus i t is wise
to refer to a like-to-like principle for Plato rather t h a n to forces.
The i m p o r t a n t m a t t e r here for cosmogony is t h a t a combination of the
shaking of the receptacle and the like-to-like principle w i l l not produce a
coherent, well ordered cosmos b u t w i l l simply produce a sorting of the
elements. The idea t h a t some physical principle, on its own and undirected
could produce the sort of order t h a t the cosmos currently displays is an
alien one to Plato. T h a t is quite reasonable given usual analogues for these
processes. W i n n o w i n g corn does not produce any interesting order, b u t
merely the separation and congregation of the entities involved. There is
also an i m p o r t a n t point to make about the modern view, w h i c h I shall
expand upon below. This is t h a t gravity is able to create the sort of order
we see i n the our cosmos only because of an extraordinarily delicate (and
extremely fortuitous?) balance between itself and the other fundamental
forces, allied to a balance w i t h other fundamental constants. W i t h o u t this
balance, the cosmos would collapse very quickly (gravity too strong), or
m a t t e r would disperse too rapidly i n the expanding universe for a n y t h i n g
20 Plato's Philosophy of Science
interesting to form (gravity too weak). Only i n w h a t appear to be highly
6

exceptional circumstances do we get order from the i n i t i a l chaos w i t h

forces alone. As we have no reason for w h y there is this balance, one m i g h t
7

argue t h a t this merely shifts matters back one stage. Plato m i g h t w e l l ask,
as many modern cosmologists have, w h y does gravity have this apparently
fortuitous balance w i t h the rest of nature? Finally, the following passage
8

at Laws 889b seems critical to Plato's cosmogony and his criticism of the
atomists:

Let me put it more clearly. Fire, water, earth and air all exist due to nature
and chance they say, and none to techne, and the bodies which come after
these, earth, sun, moon and stars, came into being because of these entirely
soulless entities. Each being moved by chance, according to the power each
has, they somehow fell together in a fitting and harmonious manner, hot
with cold or dry with moist or hard with soft, all of the forced blendings
happening by the mixing of opposites according to chance. I n this way and
by these means the heavens and all that pertains to them have come into
being and all of the animals and plants, all of the seasons having been
created from these things, not by intelligence, they say, nor by some god nor
some techne, but as we say, through nature and chance.

A ' f i t t i n g and harmonious' ordering w i l l not be generated by like to like

principles, as t h a t ordering has a m i x t u r e of types, and indeed, as Plato
emphasises here, of opposites. The only way t h a t Plato's opponents here
can explain how such an ordering came about is by chance, and t h a t w i t h
opposites looks extremely unlikely. Here again we see a problem generated
by a lack of a conception of gravity. While gravity w i l l b r i n g together things
t h a t (for the Greeks) are unlike, principles such as like-to-like w i l l not.
Plato's comment here t h a t the elements require some techne leads into
another i m p o r t a n t area for Plato's cosmogony.

Fig. 1. Random motion With no

forces acting between particles in
random motion, the modern view is
that they will tend towards the most
disordered states, Plato's view is that
chaos will remain as chaos.
 1. Cosmology, Cosmogony and Teleology 21

Fig. 2. Random motion to order (Greek cosmos) With no forces acting it is

highly unlikely that an ordered state such as the Greek cosmos should come into
being, even more unlikely that such a cosmos will stay ordered. One might find i t
even more unlikely that humans could be produced out of this. According to
modern cosmology though, gravity in balance with other forces can produce this
order.

Fig. 3. Random motion to concentric spheres. Like-to-like principles on their

own will not produce interesting order, but merely clumps or spheres of like
entities. Any mixing can happen only by chance, and the likelihood of that
producing the sort of order we see in the cosmos, is, according to Plato, extremely
remote to say the least. Plato's position appears a perfectly reasonable one, given
the lack of sophistication of Greek conceptions of gravity and force, and the
extraordinarily delicate balance of gravity with the rest of nature that is required
to produce our cosmos.
22 Plato's Philosophy of Science

Fig. 4. Modern cosmology The question for modern cosmology, on the other
hand, is why, after the big bang, the universe does not recollapse to an uninterest­
ing black hole (left), nor expand so fast that nothing of interest forms as matter is
too diffuse (right), but expands at just the right rate to form clusters, galaxies,
stars and planets and then life (middle).

I I . The chaos of the elements

Plato's vision of chaos is more radical t h a n I have suggested so far. The
implications of this are not generally recognised, b u t are highly i m p o r t a n t
for Plato's cosmology and cosmogony. I t is not only the motions and
arrangement of whatever Plato t h i n k s exists prior to the ordering of the
cosmos w h i c h are chaotic, b u t the elements themselves do not have order.
A t 53a, Timaeus tells us:

Before this, all things were lacking in reason and measure (alogos kai
ametros). When the task of ordering the universe was attended to, fire, water
earth and air all possessed traces of their own nature, but were so disposed
as would be likely in the absence of god, and such being their nature at that
time, god then firstly fashioned them by means of shapes and numbers
(eidesi te kai arithmos). They were organised by god to be as fair and
beautiful as possible, a state which they were not previously in.

We are given a reprise w i t h some further comments on n a m i n g slightly

later at Timaeus 69c:

As was said at the beginning, all these things were in a state of disorder until
God created in them commensurability with respect to themselves and each
 1. Cosmology, Cosmogony and Teleology 23
other, so far as i t was possible for them to have proportion and coramen-
surability. For at that time nothing partook of these things except by chance,
nor was there anything deserving to be named as we now name things, such
as fire and water and the others. But all these he first ordered, and then out
of them he constructed the universe.

Not only are the elements not properly ordered, but the elemental chaos
is non-progressive as well. I shall refer to this sort of order as the intrinsic
order of the elements, and the arrangement of the elements w i t h respect
to each other as the extrinsic order. W i t h the modern conception of the b i g
bang there is indeed an early phase before the elements i n the sense of
members of the periodic table are formed, this stage is neither w i t h o u t
reason nor w i t h o u t measure. This lack of intrinsic order has some impor­
t a n t consequences for the application of a like to like principle to the
p r i m o r d i a l chaos. Such a principle w i l l not w o r k to produce any sort of
ordering i n the p r i m o r d i a l chaos as we do not have elements. I f we
interpret 63c-e more strongly as a general like-to-like principle, s t i l l this
w i l l fail to produce any sort of ordering unless there is some likeness i n the
i n i t i a l chaos. Certainly Plato is aware of the idea of a chaos of complete
unlikeness. The Politicus m y t h gives us a vision of the deteriorating
cosmos, and at 273e a god who eventually intervenes because he is:

concerned that i t should not be storm driven by confusion and broken up into
an endless sea of unlikeness. 9

One m i g h t argue though t h a t this somewhat overstates Plato's position on

elemental chaos. Just prior to Timaeus 53ab, we are t o l d t h a t the recepta­
cle has a w i n n o w i n g effect, such t h a t the four kinds of element are
separated out, the most dissimilar farthest apart, the most s i m i l a r near­
est. So at 53ab, while the elements only have traces of t h e i r nature,
perhaps they have sufficient traces for the w i n n o w i n g effect to take place, 10

otherwise the passage at 53a ff. looks strange. However, Timaeus 53a is
quite specific t h a t i t is the four kinds t h a t are sorted:

In the same way at that time the four kinds were shaken by the recipient,
whose motion was like that of an instrument for shaking, and separated the
most unlike kinds furthest from each other and forced together the most like
kinds, so that the different kinds had different regions, even before the
universe was organised out of them.

However, i f this w i n n o w i n g effect does take place, there seems to be no

l i m i t on i t so eventually we would reach a situation where the four
elements are separated out. There is no m e n t i o n of this w i n n o w i n g effect
or its results at Timaeus 30a or 69c, both passages seeming to indicate t h a t
there is ordinary chaotic motion. There would appear to be several incon­
sistencies i n Plato's vision of chaos, or perhaps he presents us w i t h
24 Plato's Philosophy of Science
differing positions. This leads us to the question of how we should treat the
cosmogony of the Timaeus, the usual options being a l i t e r a l or a metaphori­
cal reading. I w a n t to suggest an alternative to this dichotomy, w h i c h
11

treats these passages as a framework for t h i n k i n g about cosmological

issues, i n particular cosmogony and some epistemological issues related to
the order of the cosmos.
The l i t e r a l i n t e r p r e t a t i o n would have i t t h a t the Timaeus gives us
Plato's view of w h a t the p r i m o r d i a l chaos was like, and the emphasis t h e n
appears to be on cosmogony, and the severity of the problem of producing
an ordered cosmos from the i n i t i a l state. I n order to get the current cosmos,
some teleological ordering is required. I n this one can see the depth of
Plato's objections to the atomists i n cosmogonical terms. F r o m the i n i t i a l
chaos, there has to be an intrinsic ordering of the elements by chance,
w h i c h seems radically improbable, added to which there m u s t be an
extrinsic ordering by chance, as a like-to-like principle is insufficient,
w h i c h also seems radically improbable. Taken together, i t seems u t t e r l y
implausible t h a t both the intrinsic and the extrinsic ordering of the cosmos
should come about by chance alone.
Plato's actual position on the p r i m o r d i a l chaos is difficult to n a i l down
though. T h a t m i g h t be due to some indeterminacy or inconsistency i n the
w r i t i n g of the Timaeus, or may be deliberate. Perhaps he is looking for
something more general here, which one m i g h t formulate as follows. I f the
current cosmos has come into being from a less ordered state, one t h a t is
to some extent chaotic, t h e n t h a t cannot be explained solely i n physical
terms, whatever the extent of the disorder. This may be the point of the
unquantified comment on 'traces' of order. I t covers many positions from
no order onwards. I f so, Plato has a rather stronger and more flexible
position against the atomists. They would otherwise have the option of
simply pointing to a disagreement over the nature of the p r i m o r d i a l chaos,
t h e i r view being less radical t h a n Plato's and so perhaps being more likely
to result i n an ordered cosmos. So perhaps Plato is not stating a specific
cosmology, one which he would appear to be confused about, b u t is rather
enunciating a general cosmological principle about the t r a n s i t i o n from any
more chaotic state to the current cosmos. I f so, the literal/metaphorical
question about his cosmogony may not be appropriate.

I I I . C h a o s a n d cosmogony
A second approach to the cosmogonical passages, complimentary to the
first, is to take t h e m as a counter-factual analysis. A n i m p o r t a n t example
of such an approach i n Plato is Laws 897b. Here, i f the cosmos moved i n a
'mad and disorderly' manner t h e n we would conclude a bad soul was i n
charge. I t does not though, i n fact i t moves i n an entirely r a t i o n a l manner
so we conclude a good soul is i n charge. The emphasis i n the Timaeus, i f i t
too can be interpreted as giving a counter-factual analysis, t h e n appears
 1. Cosmology, Cosmogony and Teleology 25
to be more epistemological. I f the cosmos has no order, or only p a r t i a l
order, and no teleology is involved, w h a t would i t be like? I f the answer to
t h a t is n o t h i n g like our current cosmos, but a chaos, and a non-progressive
chaos at that, t h e n we have the problem of w h a t order we ought to assume
our cosmos has, and w h y Plato's analysis here is thoroughgoing. The
reason w h y the resulting chaos is so radical is t h a t Plato questions any
assumption of order at a l l . T h a t may well be intended to be a critique of
the atomists and the physiologoi, as t h e i r conceptions of chaos are by no
means as radical as Plato's. Can they justify assumptions of aspects of
order w i t h i n t h e i r ideas of chaos?
Their view t h a t the vortex i n alliance w i t h a like-to-like principle w i l l
produce order seems unjustified, as any m i x i n g of unlikes, as is critical for
an ordered cosmos, w i l l come about only by chance, as the Laws empha­
sises, and the like-to-like principle w i l l produce only a clumping of the
elements, as the Timaeus emphasises.
The atomists may assume there to be a l l shapes and sizes of atoms, b u t
they also assume t h a t they are intrinsically ordered such t h a t they have
determinate shapes and sizes. A radical chaos m i g h t lack such an intrinsic
ordering. I f there were no intrinsic order of the elements at all, t h e n there
would be a complete (and non-progressive) chaos, as a like-to-like principle
would be unable to function. How does one justify the presumption of
intrinsic order at all, and of any specific type of intrinsic order?
Plato's questioning of assumptions of order may go deeper s t i l l . The use
of the analogy of the w i n n o w i n g basket would seem to pick up directly on
a theme i n Democritus. The relevant passage from Sextus E m p i r i c u s is as
follows:

There is an ancient opinion which, as I have said, has long been current
among the natural scientists to the effect that like recognises like. Democri­
tus is thought to have produced confirmation of this opinion and Plato to
have touched on i t in his Timaeus. Democritus bases his argument on both
animate and inanimate things. For animals, he says, congregate with ani­
mals of the same kind - doves with doves, cranes with cranes, and so with
the other irrational animals. Similarly in the case of inanimate things, as we
can see from seeds that are being riddled and from pebbles on the sea-shore.
For in the one case the whirling (ton dinon) of the sieve (tou koskinou)
separately arranges lentils with lentils, barley with barley, wheat with
wheat; and in the other case, by the motion of the waves, oval pebbles are
forced into the same place as oval pebbles, and round pebbles as round as
pebbles, as though the similarity in things contains some sort of force for
collecting things together. That is Democritus' view. 12

Plato of course would deny t h a t we can get an ordered cosmos from a

like-to-like principle and a vortex. There may be another w o r r y though. A t
Timaeus 52d ff. the motions of the four kinds are chaotic, they t r a n s m i t
this to the receptacle w h i c h t h e n shakes the kinds. I f we are supposing
there to be no order, t h e n there is no reason for the receptacle to move i n
26 Plato's Philosophy of Science
an orderly manner. This may be an i m p l i c i t criticism of the atomists'
assumption t h a t a vortex would form. W h y a vortex as opposed to some
other form of motion? The atomists ought not to help themselves to the
idea of a vortex w i t h o u t any justification. This is perhaps w h y Plato
chooses the analogy of the winnowing-basket for the receptacle, w h i c h is
shaken i n a fairly random way, as opposed to a sieve w h i c h is w h i r l e d i n a
specific manner. This adds another layer of i m p l a u s i b i l i t y to the atomists,
account. W h y should we believe t h a t a vortex, as opposed to a n y t h i n g else,
should form i n the i n i t i a l stages? I t is interesting t h a t Aristotle too is
critical of the atomists on these points. A t Physics 196a24 ff. he attacks
those who make the i n i t i a l rotation a spontaneous event b u t do not give
this any serious consideration.
There is t h e n an epistemological problem attendant on cosmology. We
need to assume t h a t the w o r l d is ordered i n some fashion so as to be able
to explain the current state of the cosmos. W h a t order ought we to suppose
though, and w h y t h i s rather t h a n any other? Plato's answer to t h i s is clear.
We should be guided by the principle t h a t the w o r l d has been constructed
for the best. W h y Plato believed t h a t to be the appropriate way to address
such problems is a question I shall leave to Chapter 3. I n w h a t follows i n
this chapter I w i l l point out t h a t where Plato is faced w i t h an indefinite
field of possibilities, he gives a teleological account of his choice of one, or
a small number of those possibilities. T h a t is so for the extrinsic order of
the elements (there is one cosmos and t h a t is as well ordered as possible),
and for t h e i r i n t r i n s i c order as well (to account for the m u l t i p l i c i t y of
phenomena we need to assume two basic entities, the half-equilateral and
the half-square triangles of Timaeus 53d ff.).
Plato's concerns t h e n are twofold. There is a concern w i t h w h a t order
we need to assume to be able to account for the current order of the cosmos.
There is also a cosmogonical concern about how the current cosmos could
have come about from any previous chaotic state. Physical processes are
inadequate to explain this, so there must have been some form of teleologi­
cal ordering. For Plato, the physical universe is incapable of ordering itself.
I t is clear t h e n t h a t Plato has cosmogonical reasons for advocating teleo­
logy.

IV. A t o m i s t s , physiologoi a n d c o s m o i
I t is i m p o r t a n t to recognise t h a t Plato was far from being alone i n the view
t h a t matter, motion and forces/principles cannot account for the apparent
order of t h i s cosmos. Obviously there are other ancient teleologists, b u t so
too the ancient atomists agree on this point, even i f they have a different
means of resolving the difficulty. Leucippus and Democritus, as is w e l l
k n o w n , eschew any form of teleology. Fragment 2 tells us:

Nothing happens randomly, but everything from reason and necessity.

 1. Cosmology, Cosmogony and Teleology 27
For Leucippus at least the development of the universe begins w h e n many
bodies of all shapes move i n the void and form a vortex. These bodies t h e n
13

separate like to like, and when the number of bodies prevents e q u i l i b r i u m

being m a i n t a i n e d i n the vortex, the smaller bodies move into the void
while the rest stay together, become entangled and create the first spheri­
cal structure. A l l this happens by necessity, as:

Everything happens according to necessity, for the explanation of the genesis

of all things is the vortex, which he calls necessity. 14

Leucippus and Democritus believe there to be many, i n fact an indefinite

number of, cosmoi. Certainly t h a t is so synchronically:

Democritus holds the same view as Leucippus ... There are innumerable
cosmoi, which differ in size. In some of these there is no sun or moon, in some
they are larger than ours and in some more numerous. The spaces between
cosmoi are not equal, in places there are more and in others less, some are
growing, some are in their prime, some declining, some are coming to be and
others failing. They are destroyed by falling into each other. There are cosmoi
bereft of animals and vegetation and all moisture. 15

This is almost certainly so diachronically as well:

Leucippus states that the whole is indefinite ... some of it is full and some
void ... hence an indefinite number of worlds come from these and are
dissolved into these. 16

I f there are an indefinite number of cosmoi, t h e n one would expect some

of t h e m to be such as to dissolve back to t h e i r original elements, and i f the
original vortex is indefinite, t h e n more cosmoi w i l l always being made.
Whatever principles Leucippus and Democritus use for the workings of
necessity to explain the development of the cosmos, i t is clear they are not
strong enough to generate this cosmos uniquely out of the i n i t i a l chaos.
Nor does Anaxagoras believe t h a t matter and motion alone, or i n
alliance w i t h some principles or forces are able to generate our cosmos
uniquely out of the p r i m o r d i a l chaos. Rather he believes t h a t nous, m i n d
or intelligence, plays the critical role i n the ordering of the cosmos:

Nous is the finest and purest of all things, it has all knowledge concerning
everything and i t has the greatest power. Nous controls all things that have
life, both the greater and the smaller. Nous controlled the revolution of the
whole, such that i t revolved in the beginning. At first i t revolved in a small
region, but now i t revolves in a greater, and will revolve in a greater still. The
things which are mixed and separated and divided are all known by mind.
Nous ordered (diekosmese) them all, this revolution in which the stars and
the sun and the moon and the air and the aether which are being separated
off.
17
28 Plato's Philosophy of Science
Plato's objection is well known, t h a t Anaxagoras makes no use of his nous
i n explaining the nature of the cosmos. The first problem here is t h a t
18

nous does not appear to have any ongoing role i n the ordering of the
cosmos:

After mind initiated motion, it began to withdraw from all that was moved,
and all that Mind moved was divided. And as this motion and division went
on the revolution caused i t to be divided much more. 19

Anaxagoras' position has some interesting similarities w i t h those adopted

by some mechanical philosophers i n the seventeenth century. There a
C h r i s t i a n God created the w o r l d i n a ready ordered state ( i n line w i t h the
book of Genesis) and t h e n took no further part i n the r u n n i n g of the
universe. Here nous provides the i n i t i a l impetus and ordering before
w i t h d r a w i n g , and i t is interesting to note t h a t like a C h r i s t i a n God, nous
knows a l l and is a l l powerful. The critical difference here is t h a t i n the
20

seventeenth-century conception God also creates a framework of physical

law and forces to which matter is subject. Given the i n i t i a l ordering, these
forces and laws are supposed to be sufficient to explain the ongoing order
of the world. As i n the ancient world, there was a religious debate around
this idea. N e w t o n and his supporters argued t h a t removing god from any
p a r t i n the r u n n i n g of the w o r l d would lead to atheism. Leibniz on the
other h a n d argued t h a t to suggest t h a t God was an incompetent craftsman
and would produce a mechanism t h a t needed 'winding up and cleaning'
was derogatory of the power of God and would lead to atheism. 21

I t is at least a plausible view t h a t w i t h o u t the direction of intelligence

and w i t h o u t some framework of laws and forces the cosmos w i l l not be able
to keep its i n i t i a l order, especially i n a Greek context. Some have t a k e n the
view t h a t even w i t h such a framework the cosmos w i l l not be able to
m a i n t a i n order. Newton, for instance, believed t h a t w i t h o u t the active
i n t e r v e n t i o n of God, the stars would move into one giant fireball under
m u t u a l gravitational influence, and t h a t the gravitational attraction of
22

the planets for each other would gradually disorder the solar system u n t i l
i t was no longer a system. 23

Anaxagoras though provides no such framework for the explanation of

why the cosmos has continuing order. I t is not the ongoing w o r k of nous,
yet nor does he give anyone i n an ancient context sufficient grounds to
believe t h a t m a t t e r w i l l be able to sustain or develop the i n i t i a l ordering.
I n p a r t i c u l a r while Anaxagoras allows t h a t nous controls l i v i n g things, he
does not make the cosmos itself a l i v i n g t h i n g controlled by nous. The
Timaeus provides a rather different picture, as after the i n i t i a l ordering
the cosmos has a soul, and so do the celestial bodies, w h i c h is w h y they
move i n an intelligent, ordered and regular manner. T h a t of course is i n
particular contrast to Anaxagoras' view, of which Plato was w e l l aware,
t h a t the sun and the stars were merely hot stones. I f nous w i t h d r a w s its
24
 1. Cosmology, Cosmogony and Teleology 29
influence, and there is no framework for the maintenance of cosmic order,
t h e n either t h a t ongoing order is inexplicable for Anaxagoras, or he needs
to rely on the answer of Leucippus and Democritus. I f Anaxagoras recog­
nised t h i s difficulty, i t may have given h i m an i m p e t u s towards a
multiple-worlds view. The evidence for this is cited by Simplicius:
25

We must suppose that men and the other animals which have life have been
put together just as we have, and that that they have sun and moon and so
forth, just as we have ... the separation would have happened not only with
us, but with others as well. 26

Simplicius (quite r i g h t l y ) has his doubts about how far we can push this
evidence i n the direction of m u l t i p l e worlds for Anaxagoras. A l t h o u g h not
t a k e n up by Simplicius, the point about separation here seems significant,
especially i f we take i t i n conjunction w i t h fragment 12 above w h i c h m i g h t
be thought to be suggesting t h a t separation is a continuing process. I f so
cosmoi m i g h t be continually created, and Leucippus and Democritus at
least take i t as straightforward t h a t many cosmoi separate off from the
vortex. One m i g h t add t h a t fragment 2 m i g h t well suggest t h a t the vortex
is indefinite i n extent, w h i c h could quite possibly support an ongoing
process of the separating off of cosmoi:

Air and aether are being separated off from the great surround, and this
surround is indefinite in extent. 27

There is no suggestion t h a t separating off stops. Against this one m u s t say

t h a t most of the more reliable secondary evidence is against m u l t i p l e
cosmoi for Anaxagoras. Simplicius tells us that:

Those who say that the cosmos is unique, like Anaxagoras and Empedocles,
have no need of an infinity of elements, so that Empedocles does better to
make them limited. 28

I shall come back to the question of an i n f i n i t y of elements for Anaxagoras

later w h e n we look at how Plato's geometrical atomism relates to these
questions. I t is unclear from the r e m a i n i n g fragments of Anaxagoras'
w r i t i n g s i f he believed nous to be sufficient to generate our cosmos
uniquely.
Anaxagoras' nous t h e n cannot explain the ongoing order of the cosmos.
There is though a second difficulty, about explaining the i n i t i a l ordering of
the cosmos. W h y does nous choose this i n i t i a l ordering rather t h a n any
other state, as surely there were other possible arrangements t h a n the
vortex? I f there were not, nous would not be required as an arranger,
though i t m i g h t be required as an i n i t i a t o r of motion, though one m i g h t
again ask, w h y this motion rather t h a n any other motion? Unless there is
some means of differentiating between the possible i n i t i a l states, t h e n any
30 Plato's Philosophy of Science
choice t h a t nous makes seems to be radically underdetermined. I f so one
m i g h t say t h a t the i n i t i a l ordering is not an act of intelligence at a l l , i f i t
is not based on any rational criteria. Yet Anaxagoras says n o t h i n g about
this being the best cosmos. Again, one m i g h t say t h a t Anaxagoras is
m a k i n g no use of his nous. The Timaeus again produces a very sharp
contrast to this. The demiurge produces this cosmos because i t is the best
cosmos, and so has to be the choice of intelligence.
Anaxagoras t h e n agrees t h a t matter and motion alone are insufficient
to account for the i n i t i a l ordering of the cosmos. L i k e Plato b u t u n l i k e
Leucippus and Democritus, he considers t h a t some intelligence m u s t be
involved i n this i n i t i a l ordering. I w i l l postpone discussion of Empedocles
for the moment, as his case is somewhat more complex. W h i l e he posits a
unique cosmos, despite allowing some role for chance i n physical matters,
he feels the need to m u l t i p l y biological entities to give a plausible zoogony
where he allows a considerable role for chance.
W i t h the m u l t i p l e cosmoi of Leucippus and Democritus i n m i n d , we
m i g h t begin to consider the question of parsimony i n ancient cosmological
thought. One of the objections levelled against teleological accounts is t h a t
they are ontologically profligate. I f we can produce an explanation i n terms
of matter, motion and perhaps forces alone, w h a t need have we of the
entities t h a t would be needed to underpin a teleological account? The
problem for the physiologoi though is t h a t they cannot produce a plausible
account i n these terms and need to m u l t i p l y entities so t h a t there are
indefinite cosmoi. As our cosmos is entirely separate from the other cosmoi,
and we have no evidence for t h e i r existence, they are purely theoretical
entities. The alternative, as t a k e n by Plato and indeed Aristotle is to posit
one cosmos and some teleological principles. Plato may well feel t h a t the
sort of teleology he advocates to be more parsimonious t h e n the physiolo­
goi. One m i g h t ask: H o w m u c h more profligate of entities can you get t h a n
postulating an indefinite number of cosmoi? As we shall see, a factor t h a t
may t i l t this i n the teleologist's direction is t h a t this sort question about
the order of the w o r l d about us and these sorts of answers w i l l recur over
the number of species and the number of sizes and shapes of atoms. W h i l e
the physiologoi again m u l t i p l y entities, Plato's account is parsimonious
over the number of species and the number of sizes and shapes of the
u l t i m a t e particles while using the same teleological principles.

V. R e c u r r i n g cosmological p r o b l e m s
E a r l i e r I suggested t h a t the question of whether one can explain the
apparent order of the cosmos solely i n terms of matter and motion, or
matter, motion and forces is a recurring theme i n the history of science.
Certainly i t occurred i n the seventeenth century i n relation to the me­
chanical philosophy. The basic principle of the mechanical philosophy was
t h a t a l l physical phenomena could be analysed i n terms of the mechanical
 1. Cosmology, Cosmogony and Teleology 31
interactions of m a t t e r i n motion. M a t t e r was simply extension, and h a d no
active properties whatsoever. Thus m a t t e r had no inherent attractive or
repulsive forces, and m a t t e r interacted by contact action only. Phenomena
such as gravity and magnetism were reduced to patterns of m a t t e r i n
motion (vortices of m a t t e r give gravity, the motion of screw-shaped p a r t i ­
cles gives magnetism), b u t are not inherent properties of matter. So too
(for Descartes at least) l i g h t and colour could be reduced to pressure waves
and particle spin. As the mechanical philosophy was formulated i n a
C h r i s t i a n context, there was not a problem about the i n i t i a l organisation
of the cosmos. According to the book of Genesis, God created the universe
i n a ready ordered state. Descartes actually believes t h a t the principles he
formulates w o u l d allow the generation of the current cosmos from a
p r i m o r d i a l chaos:

It may be possible to start from primeval chaos as described by the poets, i.e.
total confusion in all parts of the universe and deduce from it, in accordance
with the laws of nature, the precise organisation now to be found in things. 29

However, he backs off from this position because:

The sun and the earth and moon and stars thus existed in the beginning,
and, what is more, the earth contained not just the seeds of plants but the
plants themselves; and Adam and Eve were not born as babies but were
created as fully grown people. This is the doctrine of the Christian faith, and
our natural reason convinces us that it was so. 30

W h i l e Descartes distinguished between res extensa, the extended sub­

stance, a n d res cogitans, t h e t h i n k i n g substance, a n d proposed a
mind/body dualism for humans although not for animals, Hobbes proposed
a rather more radical m a t e r i a l i s m t h a t would reduce m i n d to m a t t e r and
motion as well. Against the mechanical philosophy, the Cambridge Platon-
ists (More, Cudworth, Barrow) argued t h a t m a t t e r and motion alone could
not account for a l l the phenomena we see around us. Certainly there were
religious motivations for this reaction, particularly i n respect of Hobbes,
b u t the Cambridge Platonists also focused on w h a t they perceived as
weaknesses i n the mechanists account. I n particular, the mechanical
philosophy had problems providing plausible accounts for gravity, magnet­
i s m and cohesion. Thus we find More m a k i n g comments like:

The phenomena of the world cannot be solved mechanically, but there is the
necessity of the assistance of a substance distinct from matter, that is, of a
spirit incorporeal.

So too he says that:

In all nature there can be nothing more certain than or well tested than that
the phenomenon of gravity is repugnant to mechanical laws; and further,
32 Plato's Philosophy of Science
that its explanation cannot be resolved into causes purely mechanical and
corporeal; but that it is necessary here to admit certain additional causes
which are immaterial and incorporeal.

The seventeenth-century debate was conclusively won by the mechanical

philosophy. T h a t has h a d significant implications for the way i n w h i c h we
explain physical phenomena, and has coloured our t h i n k i n g about the
cosmos r i g h t down to the t w e n t i e t h century. The dominant seventeenth-
century metaphors were those of clockwork and mechanism. Descartes,
original conception of forces being reduced to contact action was softened
by the inclusion of forces acting at a distance (the N e w t o n i a n conception
of gravity), while r e t a i n i n g the general mechanical conception of the
world. 31

However, these matters do not stand s t i l l and our increased knowledge

and understanding of the cosmos has resulted i n a recurrence of this
debate i n the t w e n t i e t h century, i n a slightly different form. As outlined
above, for modern cosmology the explanation of the development of the
cosmos from the p r i m o r d i a l chaos of the b i g bang i n terms of matter,
energy and k n o w n forces alone is not a problem, at least i n outline. We
have pieced together a series of events w i t h i n a reasonable timescale
w h i c h would lead from the big bang to the cosmos as we know i t today, and
indeed have spectacular confirmation for some parts of this account.
There is a problem though i n explaining w h y the fundamental constants
of nature have the values t h a t they do. Examples here are the gravita­
tional constant, the velocity of light, Planck's constant, the electro-weak
force, the nuclear strong force, etc. We are aware of no constraints or
determinants on these values, and there is no apparent reason for t h e m
being at t h e i r observed levels. W h a t is more, some argue, i t is extremely
fortuitous t h a t they are at these levels, as i n some cases very small
deviations would upset some very fine balances and the universe would no
longer be able to create or support life forms like our own. On the grand
scale, w i t h o u t this 'fine tuning', the universe m i g h t expand too quickly,
thus preventing the formation of galaxies, stars and planets, or recollapse
long before the life cycle of stars had created the conditions necessary for
life. The balance of electro-weak, nuclear strong and gravitational forces
is critical not only to the stability of atoms, b u t also to the conditions
required for t h e i r generation, the manner of t h e i r generation and t h e i r
relative abundance w h e n generated. I t is also critical for the life cycle and
nature of stars. Figures for the degree of this fine t u n i n g are typically of
the order of one part i n 10 . Thus our universe seems very u n l i k e l y indeed.
50

So we m i g h t w e l l ask, w h y is the universe like this? W h y this universe and

not others? Typically, though perhaps not exclusively, there are three types
of proposed answers.
First, one can posit some more explanatory principles. God is an obvious
one, and one w h i c h soon involves some teleology. God selects the values of
 1. Cosmology, Cosmogony and Teleology 33
the fundamental constants i n order t h a t life forms of our type can later
come about i n the universe, and perhaps even i n order t h a t the passion of
Christ can be enacted. Alternatively, another approach is the cosmological
anthropic principle, which has two popular forms:

The Weak Anthropic Principle (WAP)

The observed values of a l l physical and cosmological quantities are not
equally probable b u t they take on values restricted by the requirement
that there exist sites where carbon-based life can evolve and by the require­
ment that the Universe be old enough for i t to already have done so. 32

The Strong Anthropic Principle (SAP)

The Universe m u s t have those properties w h i c h allow life to develop
w i t h i n i t at some stage i n its history. 33

The anthropic principles are supposed to lessen our surprise t h a t we live

i n the sort of universe t h a t we do. Exactly what sort of explanations are
generated by the anthropic principles, and w h a t merits those explanations
have, are a m a t t e r of considerable controversy. Barrow and Tipler are
34

quite explicit about the teleological nature of the Anthropic P r i n c i p l e , 35

and recognise t h a t SAP is more open to teleological i n t e r p r e t a t i o n t h a n

WAP. 36

The second type of answer is t h a t one m i g h t hold t h a t we do not yet

know enough yet about the u l t i m a t e laws of physics. W h e n we have the
'theory of everything' (for which the modern acronym is TOE), perhaps
from a synthesis of q u a n t u m mechanics and general relativity, t h e n i t w i l l
be clear w h y the constants are at these levels and indeed have to be at
these levels. For some people i t is definitive of a T O E t h a t i t should explain
the values of the fundamental constants.
The t h i r d type of answer is to suggest t h a t our universe is only one of
many universes. The modern t e r m here is 'multiverse'. I f there are an
indefinite number of universes, i n s t a n t i a t i n g a l l the possible levels of a l l
the fundamental constants, and a l l possible combinations thereof, t h e n i t
is no surprise t h a t our cosmos, however u n l i k e l y i t may be, exists. This can
w o r k diachronically or synchronically, or both together. W h e n the b i g bang
occurs, i t does not send m a t e r i a l flying out into pre-existing t i m e and
space, b u t creates t i m e and space as well. The straightforward model of
the b i g bang is t h a t from the i n i t i a l instant one space-time e n t i t y is created
w h i c h expands to become the cosmos as we know i t today. Other models
have been postulated, however, where the i n i t i a l conditions are suffi­
ciently extreme such t h a t the i n i t i a l space-time not only warps b u t
separates off into many small bubbles, k n o w n as a q u a n t u m foam or
f r o t h . Each of these bubbles t h e n expands i n the usual way to form a
37
34 Plato's Philosophy of Science
universe like the one we know. I t is possible t h a t the fundamental con­
stants have different values i n these different universes.
There are certain models of the cosmos whereby the universe collapses
after the b i g bang and an expansion phase. One possibility w i t h these
'bounce' models is t h a t w i t h each collapse and subsequent b i g bang the
fundamental constants are reset, such t h a t w i t h a greater number of these
bounces, the more likely i t is t h a t our universe comes into existence. I n
fact, the latest evidence at the t i m e of w r i t i n g from the recession velocities
of supernovae is t h a t the cosmos is expanding a l i t t l e quicker t h a n was
previously thought, and so standard models do not predict recollapse,
though there are of course other models under w h i c h this may be s t i l l a
possibility. The synchronic model suggests t h a t there are an u n l i m i t e d
number of universe i n existence now. One general argument i n favour of
the multiverse is generated from w h a t is k n o w n as the 'Copernican
principle'. This is an anti-anthropocentric principle w h i c h states that:

We do not occupy a privileged position in the universe. 38

The Copernican principle has met w i t h a great deal of success. F i r s t i t was

recognised t h a t the earth is not central and stable i n the cosmos, b u t i n
fact orbits the sun. I t was recognised t h a t our sun was not unique, and
perhaps not our solar system either, b u t t h a t are many suns and perhaps
m a n y planetary systems. I t was recognised t h a t our galaxy was not unique
and t h a t u l t i m a t e l y we do not have any privileged position i n terms of
observing the universe. So perhaps too w h a t we refer to as our universe is
not unique either, and there are other, maybe many other 'universes', and
we i n fact live i n a m u l t i v e r s e .
39

The parallels w i t h ancient debates are evident. Leucippus and Democri-

tus do not feel t h a t they have sufficient principles to generate this cosmos
uniquely from the i n i t i a l chaos, so they postulate an indefinite number of
cosmoi, the ancient analogue to the modern multiverse. Plato and indeed
Aristotle, on the other hand, take up the alternative of positing a u n i t a r y
cosmos and some more explanatory principles. I t is perhaps indicative of
one of the weaknesses of ancient science t h a t no one adopts the second
modern position of a d m i t t i n g t h a t there simply is not yet enough k n o w l ­
edge about the fundamental workings of the cosmos. The problem of
whether the order of the cosmos can be explained i n physical terms alone
is t h e n a recurrent one i n the history of science. W h e n i t is felt t h a t i t
cannot be, there are also typical recurrent lines of approach, i n terms of
m u l t i p l e cosmoi or i n further explanatory principles.

V I . B i o l o g y a n d zoogony
I f there is a problem for the ancients i n this regard for cosmology and
cosmogony, t h e n a fortiori there is a similar problem w i t h biology and
 1. Cosmology, Cosmogony and Teleology 35
zoogony, i f we hold t h a t life forms have an even greater order and complex­
i t y t h a n the physical cosmos. A t t e m p t i n g to explain the body i n terms of
m a t t e r and motion was a job the eighteenth century found impossibly
difficult. F r o m the middle part of the seventeenth century onwards, there
was a programme k n o w n as mechanical biology. This was founded on the
mechanical philosophy, w h i c h had great success i n the physical sciences
d u r i n g the seventeenth century and w h i c h people now sought to extend to
the study of life forms. The two key principles for the study of life forms
were t h a t (1) a l l physical phenomena were subject to a reductive analysis
i n terms of matter, motion and the contact action of matter; (2) animals,
and the h u m a n body up to and including the b r a i n , were entirely physical.
The body was no more t h a n m a t t e r i n motion. A n i m a l s , and h u m a n bodies
too, were no more t h a n machines. Very complex machines, no doubt, b u t
machines a l l the same, and the principles for analysing mechanisms were
to be applied to the body. Some such as Boyle distinguished between
40

watches as the creation of m a n and animals as the creation of god, while

holding t h a t mechanical principles could explain both.
Mechanical biology functioned on two levels. O n a macro level, i t treated
the body as a machine, t r e a t i n g limbs and muscles as levers and ropes, the
circulation of the blood as a hydraulic system driven by a p u m p (the heart),
and the stomach as a device for crushing and pulverising food. O n a micro
level, the u l t i m a t e goal was to explain the functioning of the body i n terms
of the m a t t e r and motion of its constituent parts. As matter was conceived
to be passive, u l t i m a t e l y a l l explanation had to be i n terms of immediate
and proximate causes.
There were, however, certain very i m p o r t a n t phenomena t h a t mechani­
cal biology struggled to explain i n a n y t h i n g like a plausible manner. 41

Some of these were standard ancient objections, some were generated by

the use of the newly discovered microscope. The basic problem for me­
chanical biology was t h a t organisms appeared able to organise themselves
beyond a n y t h i n g t h a t mechanical biology could explain i n terms of matter,
motion and proximate causes. This manifested itself i n the question of
reproduction, where things as (mechanically) s i m i l a r as equine and h u ­
m a n embryos i n things as (mechanically) s i m i l a r as equine and h u m a n
wombs would consistently grow into things as different as horses and
h u m a n s . A further problem lay w i t h the early development of the h u m a n
42

embryo, w h i c h i n its early stages of development looks n o t h i n g like a

h u m a n being at a l l , and seems to develop due to some i n t e r n a l dynamic
rather t h a n being determined by proximate causes. The ability of some
animals to regenerate significant body parts (such as some crabs) or
reorganise into two after being divided (some worms) or reorganise them­
selves after being t u r n e d inside out (water h y d r a ) , also indicated a
43

greater ability to self-organise t h a n the mechanical philosophy could

allow. A further blow to the mechanical programme was the discovery of
the cell w i t h the new microscopes, as the cell appeared to be a l i v i n g e n t i t y
36 Plato's Philosophy of Science
itself, while the mechanical programme required the parts of the body to
be mechanical.
So mechanism could not account for the organisational abilities of
animals and cells, and t h e i r ability to pass on t h a t organisation and
organisational ability. Even Aristotelian ideas of form and the actualisa-
t i o n of potential seemed to provide a better conceptual framework for
understanding these phenomena t h a n mechanical biology. Thus biology i n
the eighteenth century adopted vitalist ideas. F r o m a modern perspective
the answer to many of these questions is D N A , something t h a t w o u l d have
delighted the mechanists b u t was of course far beyond any conception they
h a d at the time. There are t h e n problems, some well k n o w n i n a n t i q u i t y
w h i c h cannot be solved by the resources available to the mechanical
philosophy i n the eighteenth century. Let us now take this back to the
ancient w o r l d and Plato's views.
There is very l i t t l e information on how Leucippus and Democritus
thought l i v i n g beings came about, b u t we may well presume t h a t this
happened by reason and necessity, and the view a t t r i b u t e d to Democritus
t h a t the first men came from the earth and life-giving moisture w o u l d fit
well. Whether this process is supposed to occur i n a l l cosmoi t h a t can
support life is open to conjecture, but possibly not i f this is merely a m a t t e r
of chance and necessity. There cannot be life i n a l l cosmoi as the physical
conditions i n some cosmoi (some have no sun, some are devoid of water)
w i l l not support life. I f there is an i n f i n i t y of cosmoi though, there m u s t be
some w h i c h support life and some w h i c h have l i v i n g creatures. Whether
other cosmoi would have different forms of life to those on earth, and
whether there would be many non-viable life forms before viable ones
emerged, as w i t h Empedocles, are also questions we simply have enough
information to answer. W i t h Anaxagoras we have a l i t t l e more informa­
tion, as we are told that:

Anaxagoras says that the air contains the seeds of all things and these are
brought down together with water and generate plants. 44

This may well apply to animals too. The difficulty for Anaxagoras'
45

account is t h a t by the t i m e l i v i n g entities start to form, nous has long since

w i t h d r a w n its influence. T h a t leaves us w i t h three alternatives. F i r s t ,
nous i n some way remains and helps the organisation of life. As nous is
active here, as long as i t acts w i t h the best i n m i n d t h e n Plato w o u l d have
no objection. Secondly, despite there being no framework for ongoing order
i n the cosmos, life j u s t happens, w h i c h seems a l i t t l e hopeful to say the
least. Thirdly, Anaxagoras may believe i n m u l t i p l e cosmoi w h e n he could
have an account more i n line w i t h Leucippus and Democritus.
 1. Cosmology, Cosmogony and Teleology 37

VII. Empedocles
Empedocles' account of the formation of l i v i n g things is both fascinating
and fantastic. I n the first stage of generation:

And earth chanced in about equal quantity upon these, Hephaestus [i.e. fire],
rain and gleaming air [aither], anchored in the perfect harbours of Cypris
[i.e. Love], either a little more of it or a little less of it among them. From
these arose blood and the various form of flesh. 46

And kindly earth received in its broad melting-pots two parts of the glitter
of Nestis out of eight, and four of Hephaestus; and they became white bones,
marvellously joined by the gluing of Harmonia. 47

One m i g h t ask here i f the process i n the second passage is the w o r k i n g of

love b u t under some other names, or a different process altogether. One
m i g h t also w o r r y t h a t getting flesh and getting bones is one t h i n g , getting
t h e m to combine together into limbs and other body parts is quite another.
There are further stages to this process of generating animals:

Empedocles believed that the first generation of animals and plants were not
generated complete in all parts, but consisted of parts not joined together,
the second of parts joined together as in a dream, the third of wholes, while
the fourth no longer came from homogenous substances like earth and water,
but by mingling with each other. 48

On the earth there burst forth many faces without necks, arms wandered
bare bereft of shoulders, and eyes wandered needing foreheads. 49

Many sprang up two faced and two breasted, man faced ox progeny, and
conversely ox headed man progeny. 50

A t the f o u r t h stage the creatures are well formed enough to procreate and
so there is no need for these processes to continue, though Empedocles
does not explain w h y they cease. Empedocles does not explain a l l of this
i n terms of the actions of love and strife. He also brings i n an element of
chance:

Wherever everything came together as if it were for the sake of something,

there they survived, having been accidentally brought together artfully, but
where this was not the case, they died and are dying, just as Empedocles says
of his man faced ox progeny. 51

These things fell together, encountering each other by chance, and many
other things were constantly being produced. 52

I t is i m p o r t a n t here t h a t many things other t h a n well formed creatures

are produced. W h i l e Empedocles believes his principles of love and strife
38 Plato's Philosophy of Science
to be sufficient to generate this physical cosmos uniquely, and so has no
need of m u l t i p l e cosmoi, i t is clear t h a t he does not believe t h e m to be
strong enough to generate the current biological species uniquely. Thus we
get a m u l t i p l i c a t i o n of biological entities (as many of these cannot breed
one is w a r y of calling t h e m species) to explain the formation of the current
species. Thus none of the physiologoi give a mechanistic account of the
development of the cosmos t h a t does not m u l t i p l y entities, whether they
be cosmoi or species. Leucippus and Democritus, as far as we know, give
the same sort of account based wholeheartedly on a m u l t i p l i c i t y of cosmoi
to explain both the physical and the biological. Anaxagoras' account uses
nous, or fails to explain, or multiplies worlds like the atomists. Empedo­
cles' account of the actions of love and strife, however we are to construe
those, gives out at a key moment. T h a t is significant, as Empedocles at
least seems to recognise a leap i n complexity from physical to biological
phenomena. I t is also i m p o r t a n t to recognise t h a t chance plays a p a r t i n
the description of the physical world for Empedocles. Aristotle tells us:

Empedocles says that air is not always separated off to the highest region, 53

but in a chance manner - he says in his Cosmogony that I t chanced to run

this way at that time, but often it is otherwise' and he tells us that most of
the parts of animals came about by chance. 54

I f Plato found these accounts radically implausible, t h e n I t h i n k we may

well have some sympathy w i t h h i m . W i t h Aristophanes' speech i n the
Symposium, Plato may indeed be openly laughing at mechanist ideas, i n
particular those of Empedocles. I n the Timaeus, he develops an account of
the demiurge's construction of humans. F i r s t the head is constructed, and
t h e n at Timaeus 44c ff. we are told:

In order that it should not roll around on the ground, with its heights and
depths of every kind, and be at a loss in scaling these things and climbing
out of them, they gave it body as a means of support for ease of travel.

One m i g h t also wonder w h y love produces good order. I f love is a sexual

metaphor (derived from more ancient mythological accounts?) Plato m i g h t
w e l l take the view t h a t sexual hedonism w i l l not produce good (cf. the
Philebus) b u t needs to be guided by intelligence, which brings us back to
the good. A l t h o u g h this is perhaps a more modern concern (cf. m y earlier
comments on g r a v i t y and the cosmos) one m i g h t wonder w h y love has this
power rather t h a n being stronger or weaker, when i t would not achieve a
proper ordering, and w h a t i t is balanced so well against strife. As w i t h the
value of the gravitational constant and its balance w i t h other forces, this
looks fortuitous and unexplained. W h y strife begins to gain on love is also
unexplained, as Aristotle points out i n Metaphysics 1000b 12 ff. Empedo­
cles also appears to make a significant leap, from the formation of bones
and flesh to the formation of well formed biological entities such as arms,
 1. Cosmology, Cosmogony and Teleology 39
legs, necks and eyes. The account of how the biological elements of the
body are organised looks significantly underdone.
There is t h e n a further problem for the physiologoi w i t h zoogony and
biological phenomena. Here we find t h a t even Empedocles now needs to
m u l t i p l y entities i n order to account for biological phenomena. I t is 55

i m p o r t a n t to note t h a t this is a recurrent problem i n the history of science,

or at least i t was u n t i l comparatively recently, as now we have a coherent
theory of evolution and a sophisticated analysis of the role of D N A . I n
contrast to the physiologoi, Plato posits a single cosmos and the m i n i m u m
number of species to account for the phenomena (as indeed does Aristotle).
Certainly Plato gives a heavily teleological account of the formation of
l i v i n g beings. We m i g h t balance t h a t against the need for m u l t i p l e cosmoi
or for m u l t i p l e biological entities, for w h i c h there is l i t t l e or no evidence i n
the ancient world.

V I I I . T h e s h a p e s a n d sizes of atoms
One can see a similar p a t t e r n of argument occurring w i t h the question of
the shapes and sizes of the particles i n ancient atomic theories, and
perhaps w i t h Anaxagoras' theory of m a t t e r too. I f there is structure, i n
terms of groups of atoms forming an object, the size and shape of the atoms
themselves, or w i t h Plato the u l t i m a t e t r i a n g u l a r particles, t h e n t h a t
structure is something t h a t needs to explained. W h y this structure rather
t h a n any other structure? One answer from the atomists is t h a t there are
an infinite number of shapes and sizes to the atoms, so i t is no surprise
t h a t there are these particular atoms which have this particular shape and
size:

Leucippus supposed there to be an infinite number of atoms that are always

in motion and have an infinite number of shapes on the grounds that nothing
is such rather than such (dia to meden mallon toiouton e toiouton einai). 56

I f there are no restrictions on the shapes and sizes of atoms, t h e n one

m i g h t expect there to be some very large atoms, though of course they may
only be i n other cosmoi:

One (Epicurus) supposed all atoms to be very small, and on account of this
imperceptible, the other, Democritus that some atoms are very large. 57

I f one accepts t h a t 'nothing is such rather t h a n such' t h e n there cannot be

an orthodox explanation of w h y one atom differs from another. T h a t there
are an infinite number of worlds is held to explain w h y there are certain
groups of atoms w h i c h make up objects. Infinite number of shapes and
sizes to atoms explains w h y there are these shapes and sizes of atoms.
Aristotle is quite clear on w h a t he believes are the m a i n differences
40 Plato's Philosophy of Science
between Leucippus and Plato at De Generatione et Corruptione 325b20 ff.
B o t h hold t h a t the u l t i m a t e constituents are indivisible and are figures,
but while Leucippus holds t h a t they are solid and infinite i n figure, Plato
holds t h a t they are planes and are definite i n number. For Leucippus
58

coming to be and passing away depends on the contact and motion t h r o u g h

the void of these particles, for Plato on contact only
Whether Anaxagoras posits m u l t i p l e cosmoi or not, Aristotle certainly
considers h i m to posit an i n f i n i t y of principles i n contrast to the four
element theory of Empedocles:

These two differ, however from each other in that Empedocles imagines a
cycle of such changes, Anaxagoras a single series. Anaxagoras again posited
an infinity of principles, namely the homeomerous substances and the
opposites together, while Empedocles posits only the so-called elements. 59

Elsewhere, Aristotle confirms this opinion about Anaxagoras, and sug­

gests t h a t for Empedocles the number of elements is six. One m u s t take
60

Aristotle's evidence here cautiously, as he is a t t e m p t i n g to f i t a l l of his

predecessors into a scheme of those who posited one principle or a l i m i t e d ,
or u n l i m i t e d number of principles. T h a t Anaxagoras is not an atomist also
complicates matters, b u t there is a sense i n w h i c h he places no l i m i t s on
the numbers of principles he uses to explain the physical world, though not
quite i n the sense of Leucippus and Democritus.
Empedocles specifically l i m i t s the elements to four earth, water, air and
fire. I n a later chapter I w i l l argue t h a t Plato objects to these being
elements because they are not basic and stable enough, and I w i l l argue
t h a t several aspects of Plato's geometrical atomism can fruitfully be seen
as an attack on and an attempt to solve some of the perceived problems of
Empedocles' account. While Plato and Empedocles would agree on a small
number of basic constituents of the world, there is the question of w h a t we
should hypothesise about t h e i r nature. W h y these elements rather t h a n
any others? W h a t shape and size of u l t i m a t e particles ought we to hypoth­
esise? Plato of course has a very strong answer to this. Rather t h a n
subscribe to the ou mallon principle of the atomists, which generates an
i n f i n i t y of shapes and sizes of atoms, we have the heavily teleological
account both of the two basic types of triangle and of the way t h a t they
combine to form complex plane figures and t h e n three-dimensional p a r t i ­
cles of earth, water, air and fire. Thus we find Timaeus saying at 53d4-54a6
that:

This we hypothesise (hupotithemetha) as the principle (archen) of fire and of

the other bodies ... but the principles of these which are higher (anothen) are
known only to God and whoever is friendly to him. I t is necessary to give an
account of the nature of the four best (kallista) bodies, different to each other,
with some able to be produced out of the others by dissolution ... We must be
eager then to bring together the best (kallei) four types of body, and to state
 1. Cosmology, Cosmogony and Teleology 41
that we have adequately (hikanos) grasped the nature of these bodies. Of the
two triangles the isosceles has one nature, the scalene an unlimited number.
Of this unlimited number we must select the best (kalliston), if we intend to
begin in the proper manner. If someone has singled out anything better for
the construction of these bodies, his victory will be that of a friend rather
than an enemy. We shall pass over the many (ton pollon) and postulate the
best {kalliston) triangles.

There are a considerable number of echoes of the Phaedo i n this passage.

We have a hypothesis, there is t a l k of something higher (anothen), there
is hope to produce a sufficient (hikanos) account, and of course at a l l times
i t is the best t h a t is hypothesised. One can also see w h y Plato m i g h t well
61

feel t h a t the 'nothing is such rather t h e n such' principle of the atomists is

u t t e r l y inadequate. I n Chapter 3 I shall discuss i n general how Plato
makes choices from an indefinite field of possibilities and how t h a t relates
to the way i n w h i c h he attempts to solve one of the difficulties raised by
Meno's paradox.
Plato's approach to the question of the shape and size of the u l t i m a t e
particles is t h e n consistent w i t h his approach to the questions of the
number of cosmoi and the number of species. Plato uses teleological
principles to m i n i m i s e the number of entities here, and eschews the
multiplicative approach. Once more we have our parsimony choice, this
t i m e between an i n f i n i t y of shapes and sizes of atoms and some teleological
principles. Note t h a t there is no evidence for an abundance of shapes and
sizes of atoms. 62

I would also argue that, apart from being a perfectly plausible line of
investigation i n an ancient context, Plato's approach to the question of the
nature of the u l t i m a t e constituents of the physical w o r l d has been histori­
cally very significant. There are several matters on which some major
trends i n the history of atomism have gone w i t h Plato rather t h a n Leucip­
pus and Democritus or Epicurus. Perhaps due to the fact t h a t the revival
of ancient atomism occurred i n the highly C h r i s t i a n context of the seven­
teenth century, the atomic theory has l i m i t e d itself to one cosmos. Rather
t h a n an infinite number of shapes and sizes of u l t i m a t e particles, the
tendency has been strongly to hypothesise a relatively small number of
types of u l t i m a t e particle w i t h specific shape, usually spherical. F r o m the
end of the seventeenth century onwards we have pictured atoms, and
indeed some sub-atomic particles (certainly electrons, protons and neu­
trons i n some common theories of the atom) as being perfectly spherical.
W h y should we do so w h e n surely choice of shape here is underdetermined
empirically? One influence on this is the supposed existence of a C h r i s t i a n
God who would n a t u r a l l y create the u l t i m a t e particles of his cosmos i n the
best manner.
One m i g h t argue t h e n t h a t Plato's geometrical atomism has been
significantly undervalued i n these respects. There is m u c h more to be said
i n analysing and evaluating geometrical atomism and this is a m a t t e r I
42 Plato's Philosophy of Science
shall r e t u r n to i n a later chapter. I n the meantime, i t is i m p o r t a n t to
recognise t h a t Plato's geometrical atomism is not a metaphysical fancy
done for amusement. I t is an example of a specific type of answer to a
recurrent problem i n cosmology. As I shall argue i n Chapter 3, i t is also
typical of Plato's strategy for dealing w i t h a problem i n the philosophy of
science. Where we have an indefinite field of possibilities to choose from,
Plato begins w i t h hypothesising the best shapes for the particles. I t is also
intended to propose answers to problems t h a t Plato sees w i t h the assorted
element theories of the physiologoi, and as we shall see later, i t is intended
to answer certain problems raised i n later Plato about the nature of the
u l t i m a t e physical constituents of the cosmos and the relations between
them.

I X . A teleological cosmos vs. m u l t i p l e entities

The physiologoi t h e n postulate multiplicities, be they cosmoi, shapes and
sizes of atoms or species. One m i g h t object to this on grounds of parsimony,
b u t I would argue t h a t Plato also objects to this approach on the grounds
t h a t the m u l t i p l i c a t i o n of entities i n this manner is non-explanatory, and
t h a t the supposed cosmoi of the physiologoi are not sufficiently ordered. As
indicated earlier, Plato is not only worried about the arrangement of the
constituents of the cosmos i n relation to each other, he is also worried
about the order t h a t the constituents have themselves. The shapes and
sizes of atoms is unexplained by Leucippus and Democritus who consider
t h e m to be infinite i n variety and number, on the grounds t h a t 'nothing is
such rather t h a n such'. The following passage at Philebus 17e5, while i t
63

is i n the specific context of the analysis of l i m i t e d and u n l i m i t e d , m i g h t

equally w e l l apply to Leucippus and Democritus:

The indefinite plurality of things and in things makes you in each case
indefinite of thought and someone of neither status nor account (ouk ellogi-
mon oude enarithmon), since you have never yet examined the number in
anything. 64

There are several w o r d plays i n the Greek here w h i c h associate allowing

indefinite p l u r a l i t y i n the w o r l d w i t h being indefinite i n thought and
suggest t h a t i f you cannot give an account i n either words or numbers t h e n
you are a person of no account. Similar sentiments, though perhaps less
vigorously expressed, can be found i n the Timaeus as well. The following
passage at Timaeus 55c comes directly after the description of how the two
basic types of triangles combine to form the complex plane triangles and
squares and these t h e n form the three-dimensional elements:

I f in considering all this someone should raise the quite proper (emmelos)
question of whether the number of cosmoi should be said to be unlimited or
limited, he will suppose that the view that they are infinite is that of someone
 1. Cosmology, Cosmogony and Teleology 43
who is indefinite {apeirou) on a matter on which he should be definite
(empeiron), though to raise the question of whether in truth i t is proper to
speak of these as one or five is rather more reasonable. Our view declares i t
to be one, according to the likely account, but another, looking somehow to
other things, might hold a different opinion.

A g a i n we have the play on indefinite i n the w o r l d and of thought, though

here i t could be rendered 'inexperienced on a m a t t e r i n w h i c h he should be
experienced'. Emmelos is also interesting as a contrast to the plemmelos
(discordant) behaviour prior to the ordering of the cosmos i n the critical
passage on the ordering of the cosmos at Timaeus 30a. The passage at
Timaeus 55c has long been considered a rather puzzling one. I f we take 65

the uses of 'cosmos' here to refer simply to worlds, t h e n i t is h a r d to see

w h y Plato should even begin to consider the possibility of five worlds.
However Plato has j u s t had Timaeus describe the organisation of the
elements out of t h e i r component parts. Each element has summetria and
taxis and so m i g h t be considered a cosmos i n itself, and Plato is quite
happy to use cosmeo and its cognates to describe the process of ordering
the elements. The passage can t h e n be read i n the first instance as an
66

objection to the idea t h a t there are u n l i m i t e d shapes and sizes of atoms

rather t h a n u n l i m i t e d worlds, something t h a t Plato has h a d Timaeus
object to previously at 31a ff. Whether we speak of one or five cosmoi t h e n
depends on whether we are referring to the universe or to the elements
from w h i c h i t is organised. T h a t Plato has had Timaeus describe the
construction of five elements prior to this passage as well as after aids this
interpretation, as we t h e n have five cosmoi. T h a t t h e n leaves the puzzle of
the dodecahedron, w h i c h cannot be generated out of the two basic types of
triangle, and is not used as an element as w i t h earth, water, air and fire,
b u t 'God used i t for the whole w r i t i n g a n i m a l figures on it'. Possibly this
is a reference to the zodiac, though i t is h a r d to see how Plato is being
consistent here. 67

For Plato t h e n neither postulating an indefinite number of worlds nor

an indefinite number of shapes and sizes of atoms produces an adequate
explanation of the nature of the world about us. I have a considerable
amount of sympathy w i t h Plato on this matter. To start w i t h , his approach
to the u l t i m a t e particles is not only plausible i n context, b u t as argued i n
the previous section, is the direction t h a t science has t a k e n on this issue
ever since. O n the m a t t e r of an indefinite number of cosmoi, i t may be the
case t h a t j u s t as there are other stars w i t h planetary systems i n our
galaxy, and as there are other galaxies t h a n our own, so too w h a t we call
the universe may be only one of many. One should note though t h a t we
believe there to be a finite amount of matter i n our universe, and a large
b u t finite number of other stars, planets and galaxies. I f there are other
universes i n this sense, w h a t they are like (and whether we can ever k n o w
a n y t h i n g of them) is another matter. I t is only a m a t t e r of conjecture t h a t
44 Plato's Philosophy of Science
they have different values for the fundamental constants to our universe.
I n the absence of evidence, this approach appears to be enormously
profligate of entities. There is also a m a t t e r of general aetiological princi­
ple here. I f I ask w h y something is this like i t is, i n t h a t i t has some
determinate property, does i t explain a n y t h i n g to say there are an indefi­
nite number of these w i t h a l l possible values and this is one of t h e m ? I ' m
68

not convinced t h a t such an explanation generates any degree of under­

standing, and nor does such an approach seem heuristically useful. One
m i g h t also w o r r y about the following difficulty. A multiverse w i l l have a
structure to i t (universes A , B and C are related to each other so) w h i c h
could be different. How do we explain t h a t structure? O n the principle we
employed to explain the order of our universe, are there many m u l -
tiverses? Or is there some principle which explains the order of the
multiverse? I f this is regressive, then I do not see t h a t i t is explanatory. I f
i t is not, we ought only to posit extra worlds w h e n we have concrete
evidence of t h e m , rather t h a n as a means of explanation. One m i g h t t r y to
deny t h a t the multiverse has structure or could be treated as a single
entity, b u t t h a t would seem very odd, and I would certainly side w i t h Plato
on whether indeterminacy i n this sense could ever be explanatory.
There are t w o i m p o r t a n t premises i n play i n some arguments i n the
Timaeus. The first is t h a t w h a t there is constitutes a single, whole e n t i t y
and not an indefinite plurality. We can take t h a t internally, i n t h a t w h a t
there is forms a coherent whole, or externally, i n t h a t our cosmos is a l l
there is. The second is t h a t there is a unique best. Put another way, there
are no ties for first place w h e n we ask w h a t is best, b u t there is a unique
best. As we shall see i n Chapter 3, t h a t is i m p o r t a n t epistemologically, b u t
i t is also i m p o r t a n t cosmologically. There is one, and only one best arrange­
m e n t for t h e cosmos, a n d the demiurge w i l l i n s t a n t i a t e t h a t best
arrangement as he always acts for the best, as is emphasised at the
beginning of Timaeus' m a i n speech at 29d7 ff.
Plato has been accused of a fallacy of division at Timaeus 3 1 a l f f . The 69

passage begins w i t h Timaeus asking whether we ought to speak of one

heaven, or many or an indefinite number. The critical point is Timaeus
31a3-4, where the answer is:

One, i f it is crafted according to a paradigm (hena, eiper kata to paradeigma

dedemiourgemenos estai).

Does Plato illegitimately assume t h a t because the paradigm is unique and

the craftsman copies i t , t h a t the copy w i l l be unique too? We have t h o u g h
a perfectly plausible alternative. The copy is unique because the demiour-
gos w i l l make the best copy of i t and there is one and only one best copy.
The uniqueness of the copy is generated by the nature of the demiourgos
(acts for the best, no reason to do a n y t h i n g else) and not an attempt to copy
the uniqueness of the model.
 1. Cosmology, Cosmogony and Teleology 45
Does Plato consider the cosmoi of the physiologoi to be properly or­
dered? I f order has come about accidentally and there is n o t h i n g w h i c h
70

maintains t h a t order, t h a t may be a very fragile order liable to deteriorate

or even collapse entirely at any moment. The next stage of the develop­
ment of this sort of cosmos may exhibit some disorder. I f i t does not, t h e n
i t looks increasingly fortunate we are l i v i n g i n an i n ordered cosmos (or
one w i t h a great deal of order). T h i n k of this i n terms of our randomly
moving gas molecules w i t h no attractive or repulsive forces. I t is highly
u n l i k e l y t h a t they would come together as eight clumps of the appropriate
size and distribution along w i t h many other clumps arranged i n a spheri­
cal formation around t h e m (i.e. like earth, sun, moon and planets and fixed
stars). I t is even more u n l i k e l y t h a t they w i l l r e t a i n t h a t formation and
manage to execute the motions of the heavens. O f course t h a t is so long
t e r m for the atomists, as t h e i r cosmoi are perishable, b u t to the extent to
w h i c h chance is involved i t also applies from moment to moment. Leucip­
pus and Democritus do have some mechanism for the accretion and
maintenance of order, i n t h e i r conception of atomic bonding and t h e i r
principle t h a t like attracts l i k e , b u t there remains an element of chance
71

for w h i c h they need to postulate m u l t i p l e worlds. One m a t t e r t h a t is

critical to the account of the Timaeus is t h a t i t provides not only for the
i n i t i a l ordering of the cosmos by the demiurge, but also for the ongoing
order of the cosmos. There is another related issue here, w h i c h i n modern
terms is the difference between regularities and laws. I f the order of the
cosmos is (to some extent) accidental then (to t h a t extent) there are not
laws of nature b u t only regularities, i f we are fortunate enough to live i n
a cosmos w h i c h displays these regularities. Later I w i l l argue t h a t regu­
larities are not enough for Plato, and he requires something stronger.
These are not laws i n the modern sense, and again teleology is involved,
but Plato requires more t h a n the atomists m u l t i p l e worlds can offer i n
order to explain the regular and orderly behaviour of the cosmos.

X , C o s m o l o g i c a l r e a s o n s for teleology
Plato t h e n has cosmological reasons for advocating teleology. There is a
r e c u r r i n g question w i t h i n cosmology, which continues i n the present day
(though i n a slightly different form), as to whether the order we see i n the
current cosmos can be explained i n physical terms alone. W h e n the answer
to t h i s is no, there are typically two strategies. E i t h e r one can postulate
extra explanatory principles for a unique cosmos, or postulate m u l t i p l e
cosmoi. W i t h the Presocratic debate, no one gives a one cosmos mechanical
account. Leucippus and Democritus, Anaxagoras and Empedocles a l l
either use teleological principles themselves or need to explain some
aspects of the cosmos i n terms of m u l t i p l e cosmoi, m u l t i p l e shapes and
sizes of atoms, or m u l t i p l e species. No doubt there is a sense i n w h i c h i t is
true, as Vlastos puts i t , t h a t Plato's thought is:
46 Plato's Philosophy of Science
The conscious abandonment of the line of thought which had led, in the
systems of Leucippus and Democritus, to the first rigorously mechanical
conception of the order of nature. 72

One m u s t take care though to remember the role of chance and m u l t i p l e

worlds i n t h e i r systems. Plato's general strategy for cosmology can be
defended i n several ways. I t is not at a l l clear i n the ancient w o r l d t h a t a
one w o r l d mechanical account is either viable or plausible, so we cannot
criticise Plato for rejecting this. Plato rejects the m u l t i p l i c a t i o n of entities
(cosmoi, types of particle, species) as ontologically profligate and aetiologi-
cally inadequate and has good reasons (cosmological and epistemological)
for adopting a teleological approach. Moreover, Plato breaks i m p o r t a n t
new ground i n being the first to give a thoroughgoing teleology. To r e t u r n
to m y general objection to Vlastos' view, this is no 'metaphysical fairy tale'.
Plato's teleology attempts to answer serious cosmological and epistemo­
logical problems.
F r o m the modern viewpoint, i t has yet to be shown t h a t Plato's general
strategy is incorrect. A t present there is the problem of explaining the
values of the fundamental constants, for which a bare one w o r l d account
is not enough. Plato's arguments about the explanatory value of the
m u l t i p l e worlds hypothesis still bite, and so postulating extra explanatory
principles is s t i l l a viable approach. Relative to modern cosmological
anthropic principles, Plato's teleology is excessive. While the details of
Plato's approach may require historical m i t i g a t i o n (and there is plenty to
be had), his general strategy does not.
One m i g h t though argue along the following lines. Where science has
made its most spectacular theoretical progress, i t has done so by rejecting
teleology i n favour of more or less mechanical processes of evolution. We
now have non-teleological accounts of the evolution of the cosmos from the
big bang, and of the evolution of life. Thus we should do w i t h o u t our
non-progressive teleological prop, and merely consider non-teleological
explanations. A g a i n though, there are several lines of defence. Firstly, even
i f this is so, i t is an argument d r a w n from the experience of two m i l l e n n i a
of research subsequent to Plato. Plato as the instigator of this t r a d i t i o n of
teleology is i n no position to judge its history. Moreover, Plato would have
considered himself to have made great progress, and to have l a i d down a
h i g h l y progressive programme i n astronomy and cosmology. Secondly,
proponents of the cosmological anthropic principles w o u l d vehemently
deny t h a t science has always made progress by jettisoning teleology. They
w o u l d argue t h a t anthropic principles have made i m p o r t a n t predictions i n
modern science, most notably the prediction of (then u n k n o w n ) resonant
energy levels for carbon w h i c h are critical for stellar nucleosynthesis.
Historically, one m i g h t argue t h a t the Platonic programme i n astronomy
was enormously successful. A p a r t from the overtly teleological and Neo-
platonic approaches of Copernicus, Kepler and the overtly teleological and
 1. Cosmology, Cosmogony and Teleology 47
neoaristotelian Harvey, much of the t h i n k i n g of the scientific revolution
was underpinned by the idea of a benevolent C h r i s t i a n deity creating a
good w o r l d . A further question we m i g h t ask about cosmology is this.
73

A l t h o u g h we can now give a non-teleological account of the evolution of the

cosmos, this seems only to push the problem back a level such t h a t we now
have the problem of explaining the values of the fundamental constants.
Should we succeed i n doing this, we m i g h t only push the problem back
another stage. Ultimately, we may have to leave some problem unex­
p l a i n e d or use teleology. P l a t o t h e n has serious a n d reasonable
cosmological reasons for employing teleology.
 2

Astronomy, Observation
and Experiment

Plato's comments on astronomy and the education of the guardians at

Republic 528e ff. have been hotly disputed, and have provoked m u c h
1

criticism from those who have interpreted t h e m as a rejection or denigra­

t i o n of observational astronomy. I n this chapter I argue t h a t the key to
2

i n t e r p r e t i n g these comments lies i n the relationship between the concep­

t i o n of enquiry t h a t is i m p l i c i t i n the epistemological allegories, and the
programme for the education of the guardians t h a t Plato subsequently
proposes. We have, I suggest, been too eager to stress the similarities here,
w h e n recognition of the differences may supply us w i t h the tools for a
better understanding of Republic 528e ff., one t h a t to a large extent
disarms the anti-empirical critique. M y discussion proceeds i n three
stages. F i r s t , Plato takes great care to place his comments on astronomy
i n the context of the preceding epistemological allegories. Is there any
evidence here, where Plato m i g h t be thought to discuss enquiry i n general,
t h a t he rejected or denigrated observation? I argue t h a t even i f Plato
advocated a 'two worlds' (TW) ontology, he s t i l l envisaged a dynamic
process of enquiry, i n t e r r e l a t i n g sensibles and intelligibles, the investiga­
t i o n of each being necessary b u t not sufficient to achieve the overall a i m .
Secondly, Plato appears to be deriving how we ought to go about educating
the guardians from how we ought to conduct our enquiries. W h i l e the two
are i n t i m a t e l y related there are i m p o r t a n t differences, recognition of
w h i c h t u r n s the supposed rejection of observation into an affirmation of
the need for an i n i t i a l empirical approach. I contend t h a t this makes good
sense of the i n t e r n a l structure of 528e ff., and of its relations to other parts
of the Republic. Thirdly, I discuss the Republic i n relation to the evidence
offered by later works on the question of astronomy and observation. I
raise the questions of whether Plato's views on the nature of the sensible
w o r l d necessarily lead to the denigration of careful and prolonged empiri­
cal work, and whether his views on the role of observation develop.
A second passage which is though to be indicative of Plato's anti-empiri­
cal attitude is at Timaeus 68b-d, where he is alleged to say t h a t i t is useless
to make any practical test of his theory of m i x i n g pigments to obtain
various colours, and to t r y to do so would be to reveal one's ignorance of
 2. Astronomy, Observation and Experiment 49
the differences between m a n and god. I argue t h a t this passage is actually
about the proportions of the various types of fire particle i n the v i s u a l
stream w h i c h constitute a colour, and t h a t while the passage is clearly
pessimistic about practical tests i n this case, there may be some quite
sophisticated and reasonable worries about practical tests i n play here,
and practical tests may well be applicable i n many other cases. I also argue
t h a t Plato's conception of the technai is consistent both w i t h m y analysis
of Republic V I I and Timaeus 68b-d. For each techne there is both the
possibility and the necessity of learning something by empirical means
before we move on to the higher and more theoretical p a r t of the discipline.

I. Two worlds and investigation

I n order to support the view t h a t the comments at Republic 528c ff.
disparage observational astronomy, one m i g h t begin by asking, i n w h i c h
section of the line allegory does Plato place astronomy? As we are told t h a t
the 'real objects' of astronomy are 'accessible to reason and thought alone'
(529dl-5), and stand i n relation to the visible heavens as the objects of
geometry do to any diagram which depicts t h e m (529d7-e6), we m i g h t w e l l
feel t h a t astronomy belongs i n the dianoia section, L3 ( I number the line
sections L 1 - L 4 ascending). I f so, and we are t a k i n g a standard T W view
3

of Plato, t h e n one can see how anti-empirical considerations m i g h t easily

be generated. I characterise the T W interpretation as the view t h a t for
Plato four distinctions are co-extensive, and divide w h a t there is into t w o
disjoint sets. O n the one h a n d there is t h a t w h i c h is intelligible, stable,
knowable and incorporeal, while on the other is t h a t w h i c h is sensible,
changeable, opinable and physical. The first set of entities are said to be,
while the second are i n a state of becoming. I f we are dealing solely w i t h
4

intelligible objects i n astronomy, perhaps we ought t h e n to 'proceed by

means of problems' and leave the things i n the heavens alone' i f we are to
pursue the 'real science of astronomy' (530b6-9). The final phrase here is
translated i n a manner w h i c h is sympathetic to this position, and later I
w i l l discuss whether this is the best rendering. Here one m i g h t also
emphasise the abstract definition of astronomy given at Republic 528a and
528e, as the study of solids i n motion, w i t h its close links to arithmetic,
plane geometry and solid geometry i n the c u r r i c u l u m for the guardians.
Thus one can find H e a t h commenting t h a t for Plato:

Science then is only concerned with realities independent of sense-percep­

tion; sensation, observation and experiment are entirely excluded from i t . 5

W i t h i n the framework of assigning a line section to science i n general and

astronomy i n particular, other views have been canvassed. Lee for example
has argued t h a t science belongs i n the pisteis section, L 2 , and comments
that:
50 Plato's Philosophy of Science
In the Timaeus, Plato includes the natural sciences in this subsection, as
they can never reach ultimate truth, being concerned with the changing
world.6

Cornford on the other h a n d has argued t h a t there is no 'science of N a t u r e '

for Plato as there can be no exact t r u t h i n our account of the objects of
physics, w h i c h for Cornford are those of the lower p a r t of the l i n e . Rather
7

t h a n w o r k w i t h i n this debate, I w i s h to question whether this general

approach is correct. I m p l i c i t here is the assumption t h a t science for Plato
can be demarcated by its place on the line, and most commonly, t h a t the
border of the t w o realms forms an upper or a lower boundary. M y m a i n
strategic concern w i l l be to argue t h a t Plato allows and indeed emphasises
a continuous progress i n enquiry and the cognitive ascent of the philo­
sopher, from sensibles to intelligibles. So too he believes t h a t subsequently
the intelligibles may inform us about the sensibles. I f so, t h e n there can
be no demarcation for investigation between the two worlds, or by position
on the line generally. Consider Symposium 211b6-dl, where D i o t i m a is
i n s t r u c t i n g Socrates i n the ways of love:

This is the correct way of proceeding or being led in matters of love;

beginning with particular instances of beauty one must always ascend, as if
employing rising stairs, from one beautiful body to two, and from two to all,
and from beautiful bodies to beautiful ways of living, and from this to the
beauty of learning, and from here to the ultimate higher study, which is
concerned with beauty itself and nothing else, so that finally one may know
what beauty is. 8

L e t us call t h i s investigation 'erotics'. The key point here is t h a t border of

the t w o realms does not form the demarcation between erotics and non-
erotics. The investigation begins w i t h sensibles and ascends to i n t e l l i g i ­
bles as p a r t of the same discipline. This is entirely i n accord w i t h the
Republic. Concerning investigation thought of i n terms of the line allegory,
at 510b4-9 Socrates says:

In one division the mind is compelled to investigate by using the originals of

the visible order i n their turn as images, and has to base its search on
hypotheses and proceed from them.

The force w i t h w h i c h this is expressed is interesting. The m i n d is com­

pelled, anankezetai (510b5) to investigate i n this manner, r u l i n g out any
other method of accessing the intelligibles. The Seventh Letter strongly
supports this view (though of course there are doubts about its authentic­
ity). 342b distinguishes name, account, image and knowledge, and states
t h a t the first three are the necessary means for a t t a i n i n g the f o u r t h . A t
9

343d we f i n d t h a t we must 'Work between a l l of these four, passing up and

down over each'. Thus we need to grasp the image (i.e. the physical
manifestation, cf. 343a3 ff.) i n order to access the intelligibles, and as we
 2. Astronomy, Observation and Experiment 51
m u s t pass 'up and down' between a l l four i n investigation, there can be no
demarcation at the border of the two worlds. These passages strongly
suggest t h a t investigation for Plato is the ascent of the line, including the
traversing of the border of the two realms, rather t h a n the exclusive
property of any one section. I f so, t h e n i n d i v i d u a l investigations are better
conceived of as dynamic vertical ascents rather t h a n static, horizontal
positions on the line. I t is i m p o r t a n t not to prejudge the 'horizontal/verti­
cal' debate here when t r a n s l a t i n g the names of the line sections; e.g. by
rendering episteme, used to refer to the upper parts of the line, as 'science'.
Thus science and i n d i v i d u a l disciplines are demarcated by t h e i r subject
m a t t e r (e.g. astronomy is general enquiry applied to the heavens) rather
t h a n by position on the line.
Here one m i g h t stress the links between the allegories of line and cave.
O u r s t a r t i n g point is i n the cave and the problem is how we use enquiry to
make our ascent. To ask w h i c h section of the line a discipline belongs to is
to miss this dynamism. Rather we m i g h t ask, given the current develop­
ment of a discipline: W h a t section of the line is i t now i n , and so how ought
we to pursue i t i n the future? There can be l i t t l e doubt t h a t the general
epistemology and methodology set out i n the allegories is intended to be
applied to astronomy, There are several other passages w h i c h are helpful
10

here. These can be grouped under two headings, those w h i c h show t h a t an

empirical approach is insufficient, and those w h i c h show i t to be necessary.
I t is often possible to pair these w i t h i n one work, e.g. Phaedrus 268a-269a
where we find t h a t to possess the skill (techne) of medicine one must have
more t h a n the necessary 'empirical knack' (empeiria), one must i n fact
understand the nature of w h a t is involved, here the body or the soul. 11

There is t h e n an essential empirical component to investigation for Plato.

Obviously there are restrictions on the usefulness of empirical work, for
while i t may be necessary for the i n i t i a l stages of the ascent i n the line
allegory, i t is clearly futile for the higher parts (e.g. Republic 510b where
the investigation disposes of images). Hence we find passages w h i c h decry
its use for this latter purpose, b u t not blanket bans. One m i g h t also
compare Gorgias 485c ff. and 500e ff. where identical distinctions are
d r a w n between empeiria and techne concerning rhetoric and medicine, and
one m i g h t also look to Gorgias 465a and 501a on the requirement t h a t to
be i n possession of a techne one must be able to give a r a t i o n a l account. For
necessary/insufficient pairs i n later works, one can compare, e.g., Timaeus
47a and 91d, Philebus 55e and 56a, and Laws 719e-720d.

I I . T h e d y n a m i c s of i n v e s t i g a t i o n
Also i m p o r t a n t is the subsequent descent of the line. A t Republic 520c,
Socrates speaks as i f to his fully educated guardians, t e l l i n g t h e m t h a t i t
is t h e i r duty to descend into the allegorical cave, w h e n they w i l l :
52 Plato's Philosophy of Science
see far better than those in that place, and will know each of the images,
what it is and what i t is of, because you have seen the truth concerning
beauty, justice and goodness.

T h a t intelligibles can help to inform us about sensibles shows any sup­

posed demarcation at the border of the two realms to be passable i n both
directions, 'upwards' from the sensibles and subsequently 'downwards'
from the intelligibles. Fine has argued t h a t i n the Republic, Plato is a
coherentist who offers us an i n t e r r e l a t i o n model of knowledge. Instead of
12

j u s t i f y i n g our notion of the F o r m of the Good i n terms of a n y t h i n g more

fundamental, or i n terms of knowledge by acquaintance, we do so i n terms
of its explanatory power, by showing how well i t allows us to explain the
nature and interrelationships of Forms and sensibles. T h a t the sensible
w o r l d is (as far as possible) teleogically structured, and can be explained
i n teleological terms is a point t h a t Plato is more insistent on i n the
Timaeus, b u t is s t i l l an assumption t h a t underlies the discussion of the
Republic. * I f so, as we descend the line i n showing the explanatory power
1

of the F o r m of the Good, we must draw sensibles into our explanatory

structure. So i f we t h i n k of enquiry i n terms of explanation, again we pass
the border of the two realms i n both directions. 14

The passage at Republic 520c also suggests a second strategy by w h i c h

we m i g h t undermine the supposed demarcation, w h i c h is to argue t h a t
Plato is not i n fact committed to TW. I n relation to the later dialogues, this
may w e l l be a highly profitable line of thought. Whether i t can be success­
ful for the Republic though is open to doubt. T h a t we can opine intelligibles
does not breach T W strictures, i f we take i t t h a t i t is possible to have
knowledge subsequent to opinion of intelligibles, b u t t h a t this is not the
case for sensibles. Conversely, we are able to perceive or t h i n k about
sensibles, b u t cannot perceive intelligibles. A t Republic 506c Socrates has
belief b u t not knowledge of the form of the good, while at 523a ff. we reason
about our contradictory perceptions of our fingers. Thus we ought to read
'knowable' as 'opinable and knowable' and 'opinable' as 'only opinable'. So
too w i t h 'perceptible' and 'intelligible'. This is not to suggest t h a t we can
reason about sensibles i n the strongest sense, as t h e i r deficiencies prevent
us h a v i n g knowledge about them, b u t surely we can t h i n k about them, or
b r i n g t h e m before the m i n d by some route other t h a n perception, perhaps
by memory or i m a g i n a t i o n . This weaker construal of T W meets a l l the
15

'unattractive consequences' of Fine's stronger construal, according to

w h i c h there can be no belief concerning F o r m s . These consequences lead
16

Fine to reject T W for the Republic. Where cognitive states are correlated
w i t h objects (e.g. 511de), on m y view these are m a x i m a l states to w h i c h we
may progress from lower ones. Thus we can come to know intelligibles i n
17

the manner of the Meno by attaching an explanatory account, though the

Republic has a more sophisticated view of this process. Thus the Republic
does not 'radically reject' the Meno's account of knowledge. Rather t h a n
18
 2. Astronomy, Observation and Experiment 53
reading Republic 520c as allowing unrestricted knowledge of sensibles, i t
may t e l l us no more t h a n t h a t we can know w h a t a particular sensible is
(this is an image of Beauty) and w h a t i t is of (we can know Beauty itself). 19

Knowledge of sensibles themselves m i g h t s t i l l be elusive. The analogues

for the relation of Forms and particulars t h a t Plato employs i n the Repub­
lic suggest t h a t sensibles are imperfect relative to Forms, and t h a t these
deficiencies affect statements about t h e m . I n astronomy we are told t h a t
20

i t would be foolish to seek t r u t h i n the visible stars, as they are physical

and so undergo deviations w i t h the result t h a t there is no constant relation
between t h e i r periods. I f there is no exact and consistent t r u t h , t h e n for
21

Plato there can be no knowledge. W h a t m i g h t be possible however is

knowledge of the r a n k i n g of sensibles i n order of t h e i r approximation to
the relevant k n o w n ideal, once we know w h a t t h a t ideal is. This would be
i m p o r t a n t i n politics, because the guardians w i t h t h e i r knowledge of
abstract justice could t h e n correctly r a n k actions as more or less j u s t , and
derive t h e i r r i g h t to rule from this ability. I t is also possible t h a t k n o w l ­
22

edge of the Forms would allow us improved opinions of sensibles and a

better understanding of t h e i r functioning, w i t h o u t allowing us knowledge
of t h e m .

I I I . Vlastos on investigation
Vlastos has argued t h a t for Plato, while there are necessary empirical
preliminaries t h a t must be undertaken, a discipline itself is an a p r i o r i
m a t t e r w i t h intelligible objects. He cites Phaedrus 268el-6, where Socra­
23

tes (hypothetically) says to a m a n who believes he has knowledge of

harmony because he can strike the highest and lowest notes:

My good sir, anyone intending to become skilled in harmony must know

these things, but nothing prevents someone at your stage from being without
any knowledge of harmony. You know that which of necessity must be studied
before harmony, but not harmony itself.

I f we were to accept this argument, we would need to distinguish the

discipline name, e.g. 'astronomy', referring only to the intelligible p a r t of
the investigation, from the whole investigation, i n c l u d i n g the necessary
empirical phase. The nature of the investigation would be unaffected, as
merely n a m i n g its intelligible part does not establish a demarcation. One
w o u l d t h e n have to discuss which of these we ought to refer to as Plato's
astronomy. W h a t is clear, however, is t h a t enquiry considered as a whole
does indeed have an essential empirical component.
Should we accept Vlastos' view though? First, i n general Plato refers to
the whole investigation by the discipline name. We have seen t h a t at
Symposium 211b erotics covers a broad investigation, and passages at
Republic 526de and 530d which we shall examine i n the next section w i l l
54 Plato's Philosophy of Science
help to confirm this. Secondly, I have argued t h a t an i m p o r t a n t factor i n
investigation for Plato is the i n t e r r e l a t i o n of sensibles and intelligibles,
i n v o l v i n g the subsequent descent of the line, giving us some information
about the contents of the cave and d r a w i n g the phenomena i n t o our
explanatory system. Vlastos' model of enquiry for Plato though is very
m u c h a unidirectional one, completely discarding the empirical p r e l i m i ­
naries once we begin the 'real' discipline. A p a r t from the question of
whether this is a good description of enquiry as envisaged by the episte-
mological allegories, one m i g h t also wonder about its relation to Vlastos'
view t h a t the distinction between forms and particulars is an attempt to
formulate a distinction between necessary and contingent or a p r i o r i and
a posteriori. If, as Jordan has argued, we can undermine this view of the
24

forms, t h e n we can undermine any related conception of enquiry too. 25

A different objection to m y position could be based on the association of

the methodologies of astronomy and geometry. One m i g h t propose the
following dilemma. I f geometry is to be pursued solely t h r o u g h i n t e l l i g i ­
bles, t h e n so is astronomy, and i f astronomy is at least p a r t l y empirical
t h e n so is geometry. There can be l i t t l e doubt t h a t there is a close associa­
t i o n between astronomy and geometry for Plato. However, we m i g h t ask
26

the following question: Ought we to assimilate astronomy directly to

geometry, or ought we to consider how both relate to a general theory of
investigation? I f the latter, and the general theory recognises variables
w h i c h affect i n d i v i d u a l disciplines, then there may be i m p o r t a n t differ­
ences between how we should presently pursue astronomy and geometry. 27

The problem w i t h the dilemma as posed is t h a t i t fails to recognise the

dynamic nature of investigation for Plato. While geometry is at the mo­
ment pursued t h r o u g h intelligibles, the line allegory makes i t clear at
several points t h a t we can only access this level i n by ascending from the
sensible v i a the contemplation of diagrams, and t h a t we m u s t eventually
28

transcend the current level. How we currently ought to pursue a disci­

29

pline for Plato is a question t h a t is dependent on how far t h a t discipline

has developed. A l t h o u g h not directly expressed by Plato, we m i g h t allow
h i m the following consideration. Only a few of the most elementary
perceptions are required to make us enquire into the nature of number, 30

w h i l e geometry requires more, and more specific perceptions for us to

investigate the relevant intelligibles, and so on i n a continuum v i a astro­
nomy and harmony. So while astronomy and geometry share the same
31

basic methodology, i t would be wrong to equate the two directly as astro­

nomy w i l l require a m u c h broader empirical base. As astronomy may be at
a different stage of its development, i t may require a different prescription
for its future pursuit, such t h a t we cannot w i t h o u t further consideration
impose recommendations for geometry onto astronomy. Even given t h a t
Plato does not directly assimilate astronomy to geometry, we m i g h t be
critical here feeling t h a t there is a fundamental division between disci­
plines t h a t require empirical confirmation and those t h a t do not. Plato
 2. Astronomy, Observation and Experiment 55
though approaches the m a t t e r from a different perspective. He seems to
have i n m i n d a general account of the genesis and development of disci­
plines, and the evolution of geometry from l a n d measuring and other
'practical geometry' to abstraction to diagrams and t h e n to intelligibles is
something well theorised by the line allegory. This may well be related to
32

the way i n w h i c h he treats h u m a n intellectual development i n the cave

allegory.

IV. D o i n g a s t r o n o m y a n d t e a c h i n g a s t r o n o m y
Let us now consider Republic 528e-530d. Plato's critics have made m u c h
of t h i s passage, while others have held t h a t i t is difficult to produce a
coherent i n t e r p r e t a t i o n of i t as a w h o l e . The context here is crucial, and
33

I m u s t disagree w i t h Crombie who suggests that:

I f there is anything about scientific method in the Republic, it will be i n the

curriculum for the further education of his rulers which Plato lays down in
book 7. 34

The epistemological allegories have already given us a general methodo­

logy, w h i c h as we have seen, certainly applies to astronomy. A t Republic
521c there is a m a r k e d change of direction. Socrates switches to discussing
how the guardians of the ideal state m i g h t be produced, comments t h a t
w h a t is at issue is the conversion of the m i n d 'from a k i n d of t w i l i g h t to the
t r u e day' (Republic 521c5), and t h e n says:

It is necessary therefore to ascertain what sort of studies have the power to

effect this.

The question now is how we m i g h t employ a discipline for specific educa­

t i o n a l ends rather t h a n how we m i g h t pursue an i n v e s t i g a t i o n . This35

means t h a t we must be very cautious i n d r a w i n g any conclusions concern­

i n g scientific method here. As a shorthand I shall refer to 'teaching
astronomy' and 'doing astronomy' respectively. A t Republic 526d8-e7 Soc­
rates says that:

It is necessary to consider whether the greater and more advanced part of

the subject, if taken further, has the effect of making i t easier to grasp the
form of the good ... I t is useful if it compels the contemplation of being, rather
than becoming.

The method of education is t h e n to encourage the guardians to consider

the intelligibles. Note t h a t only part of the whole subject has this effect. 36

Socrates also employs a distinction between astronomy as i t is used now

i n philosophical education, and how i t ought to be so used. A t Republic
56 Plato's Philosophy of Science
529bc he criticises the current use of astronomy, and at 529c4-6 Glaucon
asks:

But how is i t necessary to learn (manthanein) astronomy, as opposed to the

manner in which they learn (manthanousin) it now, i f it is to be useful for
the studies which we recommend?

I t is i m p o r t a n t here to translate manthanein as 'to learn' rather t h a n as

'to study', w h i c h would introduce an ambiguity w h i c h is not i n the Greek. 37

Study is ambiguous between learn and research (i.e. teaching and doing
astronomy) i n a way i n which manthanein is n o t . I t is this sentence, i n
38

conjunction w i t h 526de, which sets the terms for the discussion w h i c h is

to follow. The a i m is to determine which part of astronomy deals w i t h the
intelligible and so ought to be used i n philosophical education, i n distinc­
t i o n to the current curriculum. How we ought to teach astronomy is going
to be derived from how we ought to do astronomy, and contrasted w i t h how
i t is t a u g h t now. This accords well w i t h the broader relations of the t w o
sections we have examined. How to educate the guardians is being derived
from the earlier epistemological discussion.
First, Socrates considers the status of the celestial bodies, w h i c h t h o u g h
they may be the best and most precise of visible things, are inferior to the
t r u e entities, the orbits which the real speed i n true number and p a t t e r n
move relative to each other, w h i c h are accessible to reason and thought b u t
not to eyesight. The stars are likened to a diagram, as w e l l d r a w n as
39

m i g h t be, b u t s t i l l h a v i n g inherent shortcomings, which we ought to use

as i n geometry. We do not examine a diagram hoping to discover any
precise t r u t h s , b u t we consider the intelligibles w h i c h the diagrams ex­
press. I n line w i t h the programme of 526de and 529c, Socrates separates
40

the sensible and intelligible parts of astronomy. Note t h a t there is no ban

on observation here, where investigation is being discussed we are to use
the visible heavens as a model, as i n the line allegory. H a v i n g set up the
appropriate distinctions, at Republic 530b6-cl he t h e n draws the concomi­
t a n t conclusion:

It is by means of problems, then, that we shall proceed with astronomy as

we do geometry, and we shall leave the things in the heavens alone, i f we
propose by really taking part in astronomy to make useful instead of useless
the understanding that is by nature i n the soul.

The key question here is whether 'proceed w i t h astronomy' refers to

teaching or doing astronomy, or both. Consider the structure, wording and
placement of this sentence. I t is the c u l m i n a t i o n of this section, and as
such should be completing the programme of 526de and 529c. I f i t refers
only to teaching astronomy, t h e n i t gives an appropriate answer to the
questions raised. I f we are to use astronomy properly for education we
m u s t employ the intelligible part of the investigation, and not divert our
 2. Astronomy, Observation and Experiment 57
students w i t h observations, as philosophical educators do now. The ' i f
41

clause is specifically and exclusively educational, and the conditional

structure allows us to draw a contrast w i t h how astronomy is t a u g h t now
and how i t ought to be done. Crucially, the attempt to derive how to teach
42

astronomy from how to do astronomy i n terms of the part/whole relation­

ship i m p l i e d at 526e makes sense on this reading. For the structure of the
passage to work, doing astronomy must involve something more t h a n
teaching astronomy. I t may also be significant t h a t Plato uses ean, mean­
i n g 'to leave alone, to leave be or to heed not', rather t h a n the stronger
phrase chairein ean, meaning 'to renounce' as he does i n other contexts.
Ean fits the contrasts perfectly. We leave observation on one side for
education, b u t not for other purposes, where one m i g h t expect chairein ean
i f Plato were renouncing observation entirely. 43

The crucial point made by this carefully constructed and placed sen­
tence is t h a t astronomy as i t is taught now and as i t ought to be done do
involve examining the physical heavens - this sentence i m p l i c i t l y affirms
rather t h a n explicitly denies this - i n distinction to how we ought to use
astronomy i n philosophical education, w h e n we should leave the things i n
the heavens alone, and consider only the intellectual problems.
I f w h a t I have argued earlier is correct, then this makes good sense i n
terms of the methodology suggested by the epistemological allegories j u s t
t e n Stephanus pages earlier, and Plato has taken great pains to place his
comments i n the context of t h a t methodology. I t is thus possible to appre­
ciate the distinction between doing and teaching t h a t Plato is showing us
i n this passage. I f we ignore this, t h e n much of the structure of this
passage is otiose. A n additional advantage of this reading is t h a t i t does
not require us to a t t r i b u t e any gross exaggeration or overstatement of
position to Plato. As H e a t h has suggested:

I t may be that, when Plato is banning sense-perception from the science of

astronomy in this uncompromising manner, he is consciously exaggerating. 44

There is no ban and there is no exaggeration, b u t there is a comparison

between teaching and doing astronomy. We ought also to consider the
section on harmony at Republic 530d6-8 w h i c h immediately follows this
passage, p a r t i c u l a r l y as Socrates says:

As our eyes are framed for astronomy, so are our ears for harmony, and these
two are in some way kindred disciplines.

This sentence is either heavily ironic, or affirms t h a t sense perception has

a role i n both of these disciplines, i f our eyes and ears are framed for
t h e m . Presumably much of w h a t is said about the methodology for
45

harmony applies to astronomy and vice versa. The key to this section is
again the question t h a t is being asked. A t Republic 530e5-7 Socrates says
58 Plato's Philosophy of Science
t h a t we m u s t m a i n t a i n the principles t h a t have already been l a i d down,
and:

We must prevent those that we shall nurture from ever attempting to learn
(manthanein) in an imperfect manner, one that does not always bring them
to where i t is necessary for them to go, just as we were saying concerning
astronomy.

The same comments about the translation of manthanein apply here as at

Republic 529c. The p r i m a r y question t h e n is not how we should do
harmony, b u t w h a t p a r t of harmony is suitable for use i n the education of
the guardians, although there is n a t u r a l l y some consideration of the
former i n order to be able to derive the latter, as there was correspondingly
i n the astronomy section. The comments on harmony are t h e n amenable
to the same sort of doing versus teaching analysis as those on astronomy.

V. ' R e a l ' a s t r o n o m y
A further m a t t e r of translation concerns the phrase ontos astronomias
metalambanontes at Republic 520b8. This is sometimes translated w i t h
the adverb ontos t a k e n w i t h astronomias, such t h a t we get ' p a r t a k i n g i n
real astronomy'. T h a t is clearly favourable to positions w h i c h see Plato
46

b a n n i n g observational astronomy here, and to Vlastos' position whereby

'real astronomy' deals w i t h the problems set up by the empirical p r e l i m i ­
naries. The more n a t u r a l reading though is to take adverb ontos w i t h the
47

verb metalambanontes, giving 'really p a r t a k i n g i n astronomy', allowing

some difference between teaching and doing astronomy. This is a particu­
l a r difficulty for Mourelatos, who bases his discussion around the question
'What exactly is the "real astronomy" discussed i n 529c-530c?', t a k i n g
530b6-cl as a pronouncement on how 'real astronomy' ought to be done. 48

He divides the interpreters into two groups, the 'a p r i o r i ' and the 'save the
phenomena' v i e w s . The first group sees Plato as pursuing 'pure, ideal, or
49

strictly a p r i o r i astronomy', dealing solely w i t h intelligible objects. We

50

have already seen the theoretical difficulties w i t h this position, and the
problems w i t h generating i t from the epistemological allegories. Here we
see t h a t to generate i t from Republic 530b6-cl we need to force the
t r a n s l a t i o n and ignore the doing/teaching contrast t h a t w o u l d seem to be
the key to this passage.
Mourelatos' second group sees Republic 530b6-cl as an exhortation to
study the real and not the apparent motion of the heavenly bodies, the
reported challenge to 'save the phenomena' by regular circular motion
perhaps being the problem t h a t we are meant to ascend to. Mourelatos
cites Vlastos as the most able supporter of this position. According to 51

Vlastos, the meaning of the key phrase 'we shall leave alone the things i n
the heavens' (ta d'en toi ouranoi easomen) is t h a t (i) we should dismiss the
 2. Astronomy, Observation and Experiment 59
things i n the heavens as objects of knowledge, and (ii) we ought to use
t h e m to set up the problems t h a t 'real astronomy' t h e n tackles, solutions
here a m o u n t i n g to knowledge. Vlastos recognises t h a t 528e-530b offers
52

no direct support for the second part of his assertion, and so he seeks 53

general epistemological support elsewhere. This though is problematic, for

as we have seen enquiry for Plato seems to require a closer i n t e r r e l a t i o n
between empirical and intellectual work t h a n Vlastos' account allows.
Vlastos' difficulties here, I suggest, stem from t r y i n g to read how to do
astronomy directly from a passage which deals w i t h how to teach astro­
nomy. There t h e n appears to be a gap between Plato's general conception
of enquiry and how we should do astronomy. Vlastos' w o r k can be seen as
an attempt to bridge this gap from both ends, by downplaying the interre­
l a t i o n of e m p i r i c a l and i n t e l l e c t u a l factors i n i n v e s t i g a t i o n (hence
'empirical preliminaries' and 'real astronomy') and seeking a greater role
for empirical factors t h a n m i g h t be i n i t i a l l y suggested by Republic 530b6-
c l . However, i f we reckon there to be a fairly strong i n t e r r e l a t i o n between
observation and the subsequent processes i n enquiry, recognise the
do/teach contrast and do not reckon 530b6-cl to be a statement about 'real
astronomy', t h e n there are no gaps to bridge. Doing astronomy as opposed
to teaching astronomy is entirely consonant w i t h Plato's general concep­
t i o n of enquiry.
I f this is so, i t w i l l also defuse some of Vlastos' criticism of Plato's
programme. Vlastos argues t h a t Plato envisages a sharp division of
54

labour between those 'high grade drudges' who observe the heavens, and 55

the philosophers who t h i n k about the problems suggested by the results,

and t h a t this w i l l be detrimental to astronomy. This conflates doing and
teaching astronomy once more. The guardians, as p a r t of t h e i r philosophi­
cal e d u c a t i o n w i l l i n d e e d a v o i d o b s e r v a t i o n , a n d m e r e l y use t h e
observations of others to set up problems for t h e m to consider. Here the
division of labour may be even sharper t h a n Vlastos suggests, b u t for the
purposes of investigating the heavens this division may not exist at a l l .
Dicks' explanation of Plato's comments at Republic 530b6-cl is unconvinc­
i n g . He remarks that:
56

There had been accumulating during the fifth and fourth centuries B C a mass
of crude, but none the less practically useful observational material ... what
was now needed was for astronomers to sit down and do some hard thinking
about astronomical theory ... that is precisely what Plato realised, and this
is why he urged astronomers to concentrate on the mathematical side of their
subject ... to the extent even of calling a temporary halt to the mere
accumulation of more observations. 57

However, there is no mention by Plato or any of his ancient commentators

of any (temporary or otherwise) h a l t to observation. W h a t is more, the
difficulties t h a t Plato raises concerning observation are philosophical
worries. The question is not of how much observation we should do before
60 Plato's Philosophy of Science
considering more theoretical matters, but whether observation w i l l help
at a l l w i t h the education of the guardians. Again, difficulties are created
by failing to recognise the possible differences between doing and teaching
astronomy. A t Timaeus 39c5-dl, we are told that:

Men have not pondered the periods of the other stars ... thus men do not
name them nor investigate the mathematical relations between them.

As Plato gives considerable significance to the periods of the planets and

t h e i r interrelations i n the Timaeus, he can hardly have been calling for a
temporary h a l t to observation. Bulmer-Thomas has argued t h a t w h a t we
are meant to leave alone at Republic 530bc are ta oVen toi ouranoi, the
'things i n the heavens' and t h a t these are synonymous w i t h ta en ouranoi
poikilmata at 529c7. According to Bulmer-Thomas, the poikilmata
58
are
patterns rather t h a n i n d i v i d u a l bodies, and should be interpreted as
constellations. Thus Plato is not rejecting observational astronomy, b u t
only the fanciful grouping of stars into constellations w h i c h Bulmer-
Thomas argues was happening at the t i m e . However, Plato's philosophi­
59

cal worries w o u l d again appear to cut deeper t h a n t h i s . F i r s t , the

epistemological problem of the compresence of opposites, w h i c h not only
underlies the i m p o r t a n t discussion concerning the 'sight-lovers' of Repub­
lic V and the epistemological allegories, but is also used to set up the
discussion of education i n Republic V I I , affects a l l physical entities. F r o m
60

this perspective, i t is h a r d to see what advantage there m i g h t be i n

studying i n d i v i d u a l stars as opposed to constellations. Secondly, there is
the problem t h a t at Republic 530b2 where the heavenly bodies are said to
'deviate' such t h a t there is no constant relation between t h e i r periods. The
periods mentioned are day, night, m o n t h and year w h i c h are defined by
the relation of single heavenly bodies (earth, sun and moon) to each other, 61

so again i t is difficult to see how studying i n d i v i d u a l bodies rather t h a n

constellations is going to help here. Once we recognise the differences
between doing and teaching astronomy, the problem becomes more acute,
for i t is not at a l l clear how studying i n d i v i d u a l bodies as opposed to
constellations, fanciful or otherwise, is going to help w i t h the education of
the guardians. One m i g h t t h e n pose the following dilemma. I f we do not
recognise any difference between doing and teaching astronomy, then,
contrary to the philosophical t h r u s t of the passage, the guardians m u s t
observe the heavens. I f we do recognise some differences, and Republic
530b6-cl contrasts doing and teaching astronomy, t h e n as doing astro­
nomy does not involve leaving alone ta d'en toi ouranoi, i f the poikilmata
are fanciful constellations then these are what we must investigate. Plato
may well have objected to the fanciful grouping of stars into constellations,
b u t t h a t cannot be the whole story here.
 2. Astronomy, Observation and Experiment 61

V I . T h e Republic and other w o r k s

So far I have concentrated largely on the Republic. One reason for this is
t h a t i f Plato d i d denigrate observational astronomy, t h e n the purported
evidence is strongest here, and i t is w o r t h investigating whether this
charge can be repudiated on the basis of the Republic alone. There are two
further reasons for this approach. First, i n later chapters I w i l l argue t h a t
there is a significant change i n the nature of celestial motion between the
Republic and the Timaeus, the heavenly bodies no longer being subject to
the 'deviations' of Republic 530b2, b u t following predictable courses w i t h
invariable relations between their periods. This is a complex m a t t e r w i t h
many implications for other positions t h a t Plato m i g h t be thought to adopt
i n the Timaeus. I f there is regular celestial motion i n the Timaeus and
other later works, as I shall argue, t h e n this may significantly affect the
role t h a t observation plays i n astronomy. Secondly, there is clearly a
serious re-examination of epistemology and the nature of knowledge i n the
Theaetetus and other later works, w h i c h may revise some of the theories
of the Republic. A n y evidence from the later works w o u l d need to be
subjected to detailed scrutiny before i t could be adduced i n a discussion of
the Republic's astronomy.
Bearing these reservations i n m i n d , the later works certainly do not
recognise any denigration of observational astronomy. The clearest pas­
sage is Timaeus 47a ff.:

Sight, in my opinion, is the cause of the greatest benefit to us as not a word

of our current account of the universe could ever have been given i f neither
the stars, the sun, nor the heavens had been seen. But now the sight of day
and night, of months and the cycle of the years has procured for us number
and the concept of time, and has also led us to seek the nature of the
universe. From these we have derived philosophy, and no greater good has
or will come as a gift from the Gods to mortal men ... God devised and gave
us sight in order that we might observe the rational revolutions of the
heavens and use them against the revolutions of thought that are in us.

This w o u l d seem to make i t entirely clear t h a t observation is an essential

p a r t of the investigation of the heavens, and as the passage goes on to
discuss the merits of hearing, is well w o r t h comparing against w h a t we
have been t o l d about sight and hearing i n Republic V I I . I would agree w i t h
Vlastos t h a t i n the Timaeus:

Plato's language here is saturated with the terms of observational astro­

nomy. 62

This is something we can examine i n more detail i n chapter five. I f Plato

did issue a challenge to 'save the phenomena', t h a t too would seem to sit
strangely w i t h an anti-empirical reading of Republic V I I . W h i l e Plato was
62 Plato's Philosophy of Science
a theoretician rather t h a n a practising scientist, another question here is
his attitude to observational studies carried out by others. L e t us first
consider the case of Aristotle. Aristotle was essentially a theoretical as­
tronomer who took over a t r a d i t i o n from Eudoxus and Callippus. I n doing
63

so he w o u l d have been absorbing the large body of data t h a t provided the

basis for these theories, and then improving the relation between theory
and data. Indeed, on more t h a n one occasion Aristotle refers to results
obtained by E g y p t i a n and Babylonian astronomers. However, Aristotle
64

does not appear to have made detailed observations himself. I f we d i d not

know t h a t he had taken over a large body of data, we m i g h t reckon the
empirical support he provides for his theories to be sketchy and arbitrary.
We m i g h t take roughly the same line w i t h Plato, as he too is a theore­
t i c i a n rather t h a n a gatherer of data. As he is perhaps the progenitor of
the t r a d i t i o n the Aristotle later takes on board, we cannot deploy quite the
same argument i n his case. We need some more direct evidence t h a t he
took over a body of data, and this I suggest is provided by the Epinomis
and Timaeus. A t Epinomis 986e6-987a6 we are told:

The first to observe these bodies was a non-Greek. Tradition nurtured those
who first took these things to mind, due to the excellence of the summer
season which both Egypt and Syria adequately possess, and revealed to their
sight, so we say, they always beheld the stars together ... These observations
have since disseminated everywhere and have been shown to be true by the
test of time.

I f we are to believe the Epinomis, i t would seem t h a t Plato was w e l l aware

of detailed E g y p t i a n astronomical observations, and was fulsome i n his
praise of t h e i r accuracy. I t is w o r t h noting t h a t i n the Timaeus, while
65

Plato does not refer directly to astronomical records, he praises the Egyp­
tians for h a v i n g the most ancient and comprehensive records, and at 66

Laws 967b ff. he is full of praise for those who have accurately studied the
heavens. A t Timaeus 39c ff. Plato is critical of those who have not studied
a l l of the planetary orbits. Plato is also aware of Hesiod (Epinomis 990a)
as one who is not a t r u e astronomer, being someone who has only 'observed
settings and risings' rather t h a n studied orbits. As the Timaeus, Laws, and
Epinomis are a l l later works t h a n the Republic, they may represent a
change of m i n d on Plato's part. However, according to m y arguments, they
need not; the Republic does not condemn or denigrate observational
astronomy, b u t has a f i r m view of its position as the necessary s t a r t i n g
point for investigation, and of the l i m i t a t i o n s of a purely empirical ap­
proach.
One m i g h t argue t h a t while the Republic has a role for observation,
because its physical entities do not behave i n a perfectly regular manner,
i t discourages careful and prolonged empirical investigation. As Shorey
comments, however:
 2. Astronomy, Observation and Experiment 63
That the stars in their movements do not perfectly express the exactness of
mathematical conceptions is no more than modern astronomers say. 67

I t is i m p o r t a n t to recognise t h a t there are two explanations as to w h y this

should be so. According to the Republic, 68
physical bodies are not capable
of regular behaviour and so are likely to 'deviate' from any ideal projection.
A modern explanation though m i g h t state t h a t while bodies do perfectly
obey laws, w h e n we deal w i t h a phenomenon like planetary motion there
are problems of complexity and isolation. Once we postulate e.g. an
69

inverse square gravitational law, w i t h a force propagating to infinity, t h e n

we need to include every massive particle i n the universe i n our calcula­
tions i f we w a n t complete precision. A t some point we w i l l have to accept
'anomalies' relative to our model of planetary motion as due to extraneous
forces on a system w h i c h we cannot isolate. So both moderns and middle
period Plato agree t h a t there are anomalies. Yet we do not infer from this
t h a t modern science disparages careful observation. One m i g h t reply t h a t
the moderns have had great success i n closing the gap between anomaly
and theory, and t h a t there is a difference, perhaps of several orders of
magnitude, between w h a t the moderns and the ancients w o u l d be pre­
pared to accept as an anomaly. T h a t is undoubtedly true, b u t the ancients
h a d success here as well (especially i n astronomy), and we have l i t t l e
knowledge of w h a t they would have regarded as a serious anomaly. One
m i g h t also argue t h a t modern chaos theories and recognition of the butter­
fly effect actually p a i n t a somewhat bleaker picture of w h a t investigation
m i g h t achieve t h a n middle period Plato or the ancients i n general m i g h t
have accepted, b u t again we do not take these ideas as a discouragement
to detailed observational work. We need not agree w i t h F a r r i n g t o n t h e n
who argues t h a t Plato's view t h a t the heavenly periods bear no exact
relation to each other was thoroughly reactionary, as i t was l i k e l y to
discourage precise observation and theorising. 70

After the comments on the r e g u l a r i t y of the heavens at Republic 530b3-

4 concerning the heavenly bodies we are told t h a t ' I t would be absurd to
seek i n every manner to grasp the t r u t h of these things'. Now, i f this is
merely a point about the philosophical status of the heavens, we m i g h t
w e l l agree. However, some, including Vlastos have construed i t as a
methodological point, disparaging prolonged and detailed observational
work. We may use the heavens to set up the problems t h a t 'real astronomy'
subsequently investigates, b u t perhaps this involves no more t h a n a
cursory e x a m i n a t i o n . Given t h a t the doing/teaching distinction and the
71

dynamic i n t e r r e l a t i o n model of Platonic enquiry w i l l undermine any theo­

r e t i c a l s u p p o r t for t h i s p o s i t i o n , is t h e r e a n y w a r r a n t for t h i s
interpretation? This seems doubtful. The point of this passage, as we saw
i n section IV, is to determine the philosophical status of the objects of
astronomy i n order to determine w h a t part of astronomy the guardians
should study, rather t h a n to make methodological prescriptions for doing
64 Plato's Philosophy of Science
astronomy. As for the evidence elsewhere i n Plato, passages i n the Timaeus
and Epinomis t h a t we have seen earlier would seem to indicate t h a t Plato
h a d no objection to prolonged and detailed observational work, though as
a theoretician this is not something he would carry out himself, nor is i t
something t h a t he would recommend for the philosophical education of the
guardians.
This is not a crucial matter though, for as I commented earlier, by the
t i m e he composed the Timaeus Plato had changed his views on the
regularity of the heavens. Thus the question of whether Plato would have
ignored perturbations i n the orbits of planets as real anomalies, and so
would never have discovered Neptune and Pluto become redundant. A 72

common view of Plato's middle period attitude to the imperfection of the

sensible w o r l d does not necessarily result i n the disparagement of detailed
observational work. One m i g h t argue another line though, t h a t i f Plato
thought t h a t astronomy had now risen to line section L 3 , he may have
thought the empirical w o r k to be complete, and so d i d not encourage
further observation. While i t may well be the case t h a t Plato believed
astronomy to have progressed as far as L 3 , Timaeus 39c ff. quoted above
indicates t h a t Plato thought the empirical investigation of the heavens to
be far from complete, a point which tells against Vlastos' sharp division of
'real astronomy' and i t s preliminaries as well.

V I I . C e l e s t i a l m o t i o n i n the Republic
As I shall argue t h a t there is a significant change i n Plato's attitude
concerning the regularity of celestial motion from the Republic to the
Timaeus, let us examine exactly w h a t the Republic has to say on this
matter. The key passage i n the Republic is at 530a ff., where Socrates asks:

Won't the true astronomer be similarly persuaded when he gazes upon the
movements of the stars? He will hold that the architect of the heavens and
what is in them has constructed these things as beautifully as is possible for
such works. But concerning the proportions (summetria) of night to day, of
these to month, of month to year, and of the other stars to these and each
other, wouldn't he consider it absurd to expect these things always to behave
in the same manner and never to deviate (paralattein) in any way, as they
73

are physical and visible, and absurd too to seek by every means to grasp the
truth about them? 74

The position of the Republic seems quite clear on whether there is a

constant relationship of the periods of the stars to each other, and hence
whether there is regular celestial motion and whether the visible heavens
are amenable to precise mathematical description. There are i m p o r t a n t
75

criticisms to be made of this sort of position, though whether they apply to

the Timaeus as well is another matter. As Sorabji notes:
 2. Astronomy, Observation and Experiment 65
The Republic passage is particularly interesting, because Plato suggests
that, if the heavenly motions do not conform with mathematics, this shows
that there is something wrong with the bodies, not the mathematics. This is
contrary to the spirit of those who believe that nature obeys exceptionless
laws, and who would seek ever to improve their mathematical formulae, to
approach closer to the laws. 76

As to w h y Plato believes the heavens behave i n this manner, I have two

suggestions, one metaphysical, one cosmological. At Republic 529cd Plato
has Socrates say:

The things that are decorations in the heavens, since they have decorated
the visible, are regarded as the best and most exact things of their type, but
they fall far short of the true entities, the orbits which real speed and real
slowness, in true number and all true figures, move in relation to each other,
and carry the things in them; these are accessible to reason and thought, but
not eyesight.

One m i g h t propose two quite different analogues for how either mathe­
matics, or the 'real speed and real slowness', relate to the motions of the
visible heavens. First, there is the relation between an abstract geometri­
cal e n t i t y and any diagram of i t t h a t we m i g h t draw. Secondly, there is the
relation between a number, and any group of things t h a t has t h a t number.
The key difference between these two is t h a t while a geometrical diagram
cannot perfectly instantiate an abstract geometrical e n t i t y ( I cannot draw
a perfectly straight line, nor can I draw something t h a t has length b u t not
breadth), a group has its number precisely. Possibly i t is the first relation
t h a t is i n play here and is causing the problems. This i n t e r p r e t a t i o n has
the advantage i t can be applied quite generally, which seems to be the
i n t e n t i o n of the epistemological allegories, and pays due attention to the
fact t h a t Plato pushes the diagram analogy fairly h a r d i n the line allegory.
I f we look to the line allegory, i t may well be t h a t this analogue is used
quite generally for the relation of the intelligible to the sensible. Republic
510d ff. is i m p o r t a n t here as this analogue is used as the model for the
relation between sections L 2 and L3 of the line. One m i g h t argue though
t h a t this looks rather like the old (mis)reading of the sticks and stones
passage of Phaedo 74a ff. O n this view, Plato does have a concern about
diagrams, though not because diagrams are inherently inaccurate, b u t
because, i n any diagram, while A may be equal to B i t is not equal to C,
and while A may be double C i t is not double B. Diagrams are i n h e r e n t l y
particular, and suffer from compresence. Thus we find Socrates saying at
Republic 529de:

One must use these decorations of the heavens for the sake of that study, just
as i f one had chanced upon some thoroughly well drawn and perfected
diagram by Daedalus or some other skilled craftsman or artist. Anyone
acquainted with geometry, seeing such things, would consider them beauti-
66 Plato's Philosophy of Science
fully crafted, but would think i t laughable to examine them seriously, with
the hope of grasping in them the truth concerning equals or doubles or any
other ratio.

So this passage does not require a n y t h i n g imperfect about the physical

w o r l d , one j u s t has to recognise t h a t particulars are always subject to
qualification i n a way i n w h i c h equal and double themselves are not. The
problem w i t h this i n t e r p r e t a t i o n though is t h a t Plato t h e n goes on to t e l l
us t h a t the periods of the planets are subject to deviation, as they are
physical and visible, and i t would seem difficult to explain w h y Plato
should do t h a t i f the compresence of opposites/qualification of particulars
is the only difficulty i n play here. N o r w i l l i t be able to deal w i t h a passage
often cited i n relation to the Republic i n this context, Politicus 273b ff.
where i t is the bodily element of the cosmos w h i c h is responsible for its
gradual degeneration.
A n alternative diagnosis of the problems of Republic 528 ff. m i g h t come
from the m y t h of Er. Here there is a pivot, the spindle of necessity (616c3)
about w h i c h the cosmos rotates. While the fixed stars may be an i n t e g r a l
p a r t of the spindle, the hemispheres of the sun, moon and planets w h i c h
sit i n the bowl of the great spindle cannot be as they are said to rotate
slowly i n the opposite direction (617a6-8). The spindle anyway is said to
t u r n upon the knees of necessity' (617b4), so each m o v i n g p a r t of this
model has a pivot. The critical passage for our current purpose is at
617c5-dl:

Clotho with her right hand grasped and turned the outer revolution of the
spindle, leaving intervals of time between, while Atropos with her left hand
did similarly with the inner. Lachesis grasped each in turn with each hand.

I t w o u l d appear t h a t some active intervention is required to keep the

cosmos i n m o t i o n . Here Clotho and Atropos take hold of the r i m s of the
77

spindle and of the hemispheres and t u r n t h e m intermittently. A spindle

seems a p a r t i c u l a r l y apposite model here, as this is j u s t the sort of t h i n g
t h a t gradually slows down and has to be kept moving w h e n i n use. So
w i t h o u t the actions of the daughters of necessity the cosmos w i l l gradually
g r i n d to a halt. However, i f t h e i r activity is i n t e r m i t t e n t , t h e n as Republic
530ab has i t , the relations of the periods of the various heavenly bodies
w i l l not be constant. 78

W i t h this view, one m i g h t argue t h a t there are t w o difficulties t h a t r u n

t h r o u g h Republic 529c-530d, which are the compresence of opposites (so
Republic 529de) and a cosmological problem epitomised i n m y t h i c a l form
w i t h the m y t h of E r concerning astronomy only (so Republic 530b). The
problem here is t h a t the epistemological allegories seem quite general, and
one must w o r r y t h a t we are told t h a t the periods of the heavens deviate
because the heavens are physical and visible, not because of something
 2. Astronomy, Observation and Experiment 67
specific to the heavens. So too astronomy is being treated as a general case
for the purposes of the epistemological allegories.
A t h i r d possibility is t h a t Plato has not yet fully sorted out his model of
participation and his worries over the nature of the world, and so runs the
compresence of opposites problem i n parallel w i t h an imperfection of the
physical w o r l d view i n the Republic. One m i g h t argue t h a t he does
something similar i n the Phaedo, i f we compare the sticks and stones
passage w i t h the 'fish's eye' view which is p u t forward at Phaedo 109b ff.,
w h i c h compares the relation of the submarine to the t e r r e s t r i a l and the
terrestrial to whatever Plato envisages here as beyond the terrestrial. A
complicating factor here is t h a t i n some works (Phaedo, Republic, Poll-
ticus) Plato may countenance a degenerating cosmos as I shall discuss i n
greater detail later.
Whatever the diagnosis of w h y there is not regular celestial motion i n
the Republic I shall argue t h a t i t is clear t h a t the Timaeus, and indeed the
Philebus, Laws and Epinomis adopt a substantially different position.
Perhaps Plato moves to the second analogue for the relation of mathemat­
ics to these motions, one t h a t would certainly better u n d e r p i n such a
cosmology A d m i t t e d l y direct evidence of such a change is sparse, b u t the
contrast t h a t the Philebus at least is interested i n is between the arithme­
tic of the philosophers and t h a t of the masses, the former w o r k i n g w i t h
pure numbers, the l a t t e r ' w i t h two armies and two bulls and t w o extremely
small or extremely large things'. The Philebus also suggests t h a t this
79

analogue is the appropriate one for many other fields of study. I f this 80

second analogue is i n place for the relation of mathematics to the motions

of the heavens i n the Timaeus, t h e n t h a t would explain the move to regular
motion, and i f i t is i n place quite generally, i t would explain the Timaeus'
constant insistence t h a t events occur kafarithmon. Certainly i t is the case
t h a t there are cosmological changes from the m y t h of E r i n place i n the
Timaeus such t h a t the cosmos no longer requires a pivot to t u r n on, any
external bonds to hold i t together or any external interference to keep i t i n
motion.

V I I I . Timaeus 68b-d
There is a passage i n the Timaeus which some commentators have t a k e n
as showing t h a t Plato believed the idea of experiment to be problematic.
Timaeus 68d tells us that:

I f some investigator were to make a practical test of these, he would show

his ignorance of the difference between human nature and god's, in that god
has sufficient knowledge and power to blend the one into many and dissolve
the one back into many, but there is not now, nor will there ever be a man
capable of these things.
68 Plato's Philosophy of Science
I t is i m p o r t a n t here to take due care i n t r a n s l a t i n g basanon at 68d3.
A l t h o u g h experiment is a t e m p t i n g translation, i t is critical to recognise
t h a t basanon has broader resonances and of course ought not to be
assimilated to any modern notion of experiment. L S J give 'a test to t r y
whether something was genuine or n o t . . . enquiry, esp. by torture'. O n the
i n t e r p r e t a t i o n of this passage, Sedley comments:

Plato knows well enough that his own formulae for colour mixture would not
be confirmed by experimental test. But that, we are told, is because the very
idea of experimental science is flawed. To try to reproduce natural processes
in the laboratory is to play god, in ignorance of the fact that in this respect
we can never duplicate god's work. 81

Sedley takes 68b-d to be about the m i x i n g of pigments, b u t there is an

alternative view held by Cornford, and by Lloyd who comments that:

The blending in question appears to be not a matter of mixing pigments, but

one of combining fundamental atomic particles. 82

Whether 68d is about m i x i n g colours or pigments is i m p o r t a n t for assess­

i n g Plato's views on learning by m a k i n g practical tests. Let us p u t this
passage i n context first. I take i t t h a t this section on sight and colours runs
from 67c to 68d, and is the fourth part of the piece r u n n i n g from 65c ff.,
w h i c h wants to analyse taste, smell, hearing and sight i n terms of geomet­
rical atomism. A t 67c ff. we are told t h a t there are several types of fire
particle, small, m e d i u m and large, w h i c h contract, leave unaffected or
dilate the visual stream. I t is the m i x t u r e of these which create the
sensation of 'colour' ( w i t h due caveats about Greek conceptions of col­
our/hue/intensity, etc.). So we m i g h t expect the m a i n conclusions here to
83

be about the way t h a t geometrical atomism relates to perception.

A t Timaeus 68b5-9 we are told t h a t i t would be foolish to say w h a t
proportions they are mixed i n , as there is 'no necessary or likely account'
(mete Una anangke mete ton eikota logon). Does this refer to m i x i n g
pigments (the immediate context) or m i x i n g fire particles (the more gen­
eral context)? I t would be extraordinary i f we could not have some sort of
account of how many drops of black and w h i t e make a shade of grey. W h a t
proportion of large to small fire particles are involved though m i g h t w e l l
appear to an ancient to be something t h a t is beyond us, and is likely to stay
beyond us. Fire particles are far below our perceptual thresholds. So there
w i l l be problems for humans both i n gaining information about t h e m , i n
terms of identifying or quantifying types, and m a n i p u l a t i n g them. Those
m i g h t w e l l appear, quite reasonably, to be intractable problems to an
ancient. These are not problems which god has, however. W h e n he orders
the cosmos, earth, water, air and fire only have traces of t h e i r proper
nature, as is emphasised j u s t after our section at Timaeus 69bc. Clearly he
is capable of operating at t h a t level i n a way t h a t humans are not.
 2. Astronomy, Observation and Experiment 69
There is also a question about the status of Timaeus' account here. A t
68d2, the results of m i x i n g pigments are those w h i c h w i l l 'save the likely
tale' (diasozoi ton eikota muthon). However, i f 68b5-9 is about pigment
m i x i n g , there is no necessary or probable reason, and i f 68d2-7 is about
pigment m i x i n g too, any practical test w i l l be useless. Better t h e n t h a t we
take 68b5-9 to be about m i x i n g fire particles as t h e n Timaeus can have his
likely account of pigment m i x i n g , while being able to say n o t h i n g about the
m i x i n g of fire particles. I f 68d2-7 is about fire particles too, t h e n he can
have learned the probable account t h r o u g h experience. One m i g h t t r y to
avoid this by arguing t h a t while 68b5-9 is about proportions of pigments,
the rest of this section is about simple qualitative results of m i x i n g
pigments. T h a t though would leave us i n the distinctly odd position of
being able to have some account of w h a t colour occurs w h e n we m i x
pigments, b u t being unable to have any account of the relative quantities
w h i c h w o u l d give different shades, hues or intensities. I f 68b5-9 is about
m i x i n g fire particles, and one translates accordingly, this alters the bal­
ance of the passage such t h a t i t is less odd t h a t the final section refers to
m i x i n g fire atoms rather t h a n colours.
How does Plato treat s i m i l a r activities to pigment m i x i n g i n other
contexts? There is a group of practical activities i n w h i c h Plato w o u l d say
ability was acquired by empeiria kai tribe, experience or empirical knack,
and g r i n d i n g away. Typical examples here are cooking, learning a musi­
84

cal i n s t r u m e n t , base rhetoric and crude medicine. Plato is of course

85 86

highly critical of the l i m i t a t i o n s of learning i n such a manner. As we have

seen, such empirical preliminaries may be essential for learning or inves­
t i g a t i n g a subject thoroughly, but they are very far from being the whole
story. Unless we analyse the nature of w h a t we are dealing w i t h , or
introduce some mathematics we w i l l never be able to progress to techne.
Nowhere though does Plato suggest t h a t we cannot learn something,
however basic and l i m i t e d , by empeiria kai tribe. I take tribe i n particular
to be quite pejorative, w i t h overtones of'grinding out', and learning by t r i a l
and error, b u t one does not show one's ignorance of the difference between
m a n and god by a t t e m p t i n g these activities. So i t would be odd i f the
results of m i x i n g pigments, which Plato m i g h t be expected to say can be
learned by a painter by empeiria kai tribe, were singled out for extremely
strong criticism at Timaeus 68b-d.
W h a t is i t t h a t god has sufficient knowledge and power to do (hikanos
epistamenos Kama kai dunatos, 68d6) t h a t we do not? Surely not m i x i n g
pigments, as i t is evident we at least have the power to do t h a t (many i n t o
one at least), and Timaeus has j u s t been t a l k i n g about the results (and i t
87

is those w h i c h on the pigment m i x i n g interpretation are problematic, not

the ability to do mixes). We have seen t h a t here is a considerable difference
between w h a t god and humans m i g h t do w i t h fire particles. Another
possibility here is suggested by Sophist 265c-266d. Sophist 265c gives us
the standard idea t h a t we and other l i v i n g things, along w i t h earth, water,
70 Plato's Philosophy of Science
air and fire are a l l products of divine workmanship. 266bc though tells us
t h a t dream images, and those which occur i n the day, and are said to be
n a t u r a l , such as shadows and the interaction of the two l i g h t streams, are
i n fact divine works too. So t r y to investigate how fire particles m i x to
create images m i g h t well be to show ignorance of the difference between
m a n and god, b u t t h e n this is a special case restricted to the way t h a t fire
particles m i x . There is no indication of this view i n the Timaeus though,
either i n the discussion of m i r r o r s or of w h a t god does.
Does Plato anticipate failure for his account of colour mixes i f tested?
A l t h o u g h some of the terms used appear odd to us, do we know enough
about Greek colour terms to be sure of this? According to Aristotle, b r i g h t
(lampron) seen t h r o u g h a dark m e d i u m or reflected i n a dark surface is
seen as r e d , and he has some very odd things to say about the colours of
88

the rainbow. I f Plato does anticipate failure, w h y give this account at all?
89

A l l he needs to say for the structure of this section is t h a t variations i n

colour are due to variations i n fire particles i n the visual stream.
O n an i n i t i a l reading, the references of Timaeus 68b5-9 and touton at
68d2 w o u l d appear to be to m i x i n g pigments. There are several difficulties
w i t h this view though, w h i c h also gives Plato a pretty poor position, and
one t h a t is not only contrary to his consistent conception of techne i n many
other works would allow us to learn about pigment m i x i n g by empeiria kai
tribe, b u t also is unparalleled i n the Timaeus and elsewhere. Here we see
the importance of the previous analysis of Republic V I I for i n t e r p r e t i n g
Plato. One m i g h t t r y to cite passages there as support for a radical
anti-experiment reading of Timaeus 68b-d. So I would rather take the (on
the face of i t ) less likely (but perfectly possible) references of 68b5-9 and
68d2-7 to be to fire particles. Doubtless Plato could have been clearer as to
exactly w h a t he meant at Timaeus 68d, as Lloyd comments, b u t I don't
90

see t h a t we need to conclude he was hostile i n general to experimental

science from t h i s passage.

IX. Plato and experiment

I f Timaeus 68b-d is about the m i x i n g of fire particles, and we cannot by
practical test find out the proportions of those fire particles w h i c h make
up a colour, to w h a t extent are the worries here generalisable to other
situations where we m i g h t hope to make some progress by the use of
practical tests? One m i g h t worry t h a t this w i l l affect a l l cases w h i c h
involve either blending the many into one or dissolving the one into many,
as only god is sufficiently wise and powerful to do this. However, i t is
evident t h a t we can m i x or dissolve some ones and manies, both physically
( m i x i n g pigments) and philosophically. So Timaeus 68d m u s t only be
91

about ones and manies t h a t only god can blend or dissolve, and not a l l ones
and manies. I have t w o suggestions on this matter.
I f we take our cue from Sophist 266a ff., t h e n w h a t god can do b u t we
 2. Astronomy, Observation and Experiment 71
cannot is produce l i v i n g creatures (animals and ourselves), organise the
elements (earth, water, air, fire) and produce images, either those t h a t
occur i n dreams, or those w h i c h occur by day and are said to be n a t u r a l
(autophue). 92
I f so, t h e n the extent of Timaeus 68b-d is severely l i m i t e d .
There is something special about images w h i c h means we cannot find out
about t h e m by practical test. Perhaps too there is something special about
the production of l i v i n g creatures t h a t we could not find out about by
practical test either. However, t h a t would not affect the study of the
constitution and function of animals and ourselves, and indeed Timaeus
w i l l have a great deal to say about both of these later on. I f this is the
problem, i t is i m p o r t a n t to note t h a t i t would be s t i l l quite possible to l e a r n
(in a l i m i t e d manner) many things by empeiria kai tribe as Plato's account
of the technai has i t , something I take to be i m p o r t a n t for both the
suggestions I p u t forward i n this section. There is a down side to this view
w h i c h is t h a t i t would have god intervening i n the w o r l d to produce images
v i r t u a l l y constantly, w h i c h would r u n contrary to the general tenor of the
Timaeus. I t also seems slightly odd to p u t images on a par w i t h the
production of l i v i n g creatures. As noted i n the previous section, the Sophist
talks specifically of dreams and m i r r o r images i n relation to divine works,
and w h i l e the Timaeus has a discussion of both dreams and m i r r o r images
at 45d f f . , and talks of how god designed the eye and eyelids, etc., i t makes
93

no m e n t i o n of dreams or images or m i r r o r images as being due to divine

work.
M y second suggestion on w h a t god can do b u t we cannot is t h a t god does
have the knowledge and the ability to manipulate the various i n d i v i d u a l
fire atoms b u t we do not. As god has the ability to do the i n t r i n s i c
organisation of atoms, he clearly can manipulate the atoms themselves,
w h i l e i t is clear t h a t ( i n the context of fourth-century Greece) we cannot.
This immediately looks to be a more widespread problem, though one m u s t
note t h a t there is no difficulty here i f we have the knowledge and ability
to manipulate, so practical tests on larger scale agglomerations of atoms
(perhaps at the perceptible level) would be perfectly possible, as would
l e a r n i n g by empeiria kai tribe. While the Timaeus is pessimistic about
finding out the proportions of types of fire atom t h a t produce colours, i t is
happy to give an account w h i c h is i n accord w i t h empirical considerations
of other phenomena at the atomic, and even the sub-atomic level. W i t h the
discussion of digestion at Timaeus 80c ff., for instance, we are given an
account of the incoming atoms of food being cut up and reconstituted i n
order to replenish the body, and we can even have an account of the way
i n w h i c h the triangles go to make up the elements. W h y can we have this
account? I t is i m p o r t a n t to note t h a t i n relation to the fire particles, there
is no necessary or probable reason (Timaeus 68b7) for the proportions of
the sizes constituting a colour. This is not something w h i c h is said of the
other geometrical atomism analyses of perception which precede t h a t of
sight, nor is i t said anywhere else about geometrical atomism analyses of
72 Plato's Philosophy of Science
the body or a n y t h i n g else. So i f fire particles are too small for us to have
the knowledge and power to manipulate properly, and there are no plau­
sible hypotheses to be had, t h e n perhaps there is indeed no point i n
practical tests i n this particular case. I n other cases though there may w e l l
be 'necessary or probable' hypotheses which we can test the consequences
of against experience or against practical test, even where we lack the
knowledge and ability to manipulate the relevant entities.
This may indicate t h a t Plato is w o r r i e d about a more sophisticated
difficulty w i t h practical tests. I n modern parlance this is k n o w n as the
underdetermination of theory by data. I n the next chapter I shall look at
the relation between Plato's views and the modern underdetermination
problem i n detail, i n particular the way t h a t Plato's worries are generated
by Meno's (the learner's) paradox. This is a problem w h i c h worries h i m i n
many places i n the Timaeus, and is arguably one of the d r i v i n g forces
behind the h i g h l y teleological account. The aspect of underdetermination
t h a t may be i n play here is t h a t no amount of data w i l l determine w h i c h
theory we ought to adopt, but extra (non-empirical) criteria are required.
As we shall see, Plato does have a means of resolving underdetermination
and the non-empirical criteria he adopts, involving ideas of the good, have
interesting relations to modern realist resolutions. The problem w i t h fire
particles though appears to be t h a t i n this case these criteria are inade­
quate as there are no necessary or probable reasons to prefer one
hypothesis over any other, and as we m i g h t devise many hypotheses to
cover the results of any practical test, we cannot learn of the proportions
of sizes of fire particles constituting a colour by means of a practical test.
This gives us a general problem w i t h experiment w h i c h Plato believes to
be soluble, and a highly specific one to do w i t h fire particles w h i c h is not,
and the possibility of learning by empeiria kai tribe.
Clearly Timaeus 68b-d contains some pessimism about w h a t practical
test m i g h t be able to achieve. How general t h a t may be is another matter.
Sophist 266a ff. w o u l d suggest the problem is highly specific, b u t i t is not
clear t h a t Plato has these considerations i n m i n d i n the Timaeus. I f the
problem is t h a t we do not have the knowledge or the power to manipulate
microscopic entities (and never w i l l have), t h e n although this problem is
not general (we can do practical tests where we can manipulate), t h e n one
must feel t h a t Plato was unduly pessimistic. Criticism on this m a t t e r
t h o u g h ought to be tempered by the context of fourth-century Greek
science, where w i t h o u t any technology of optical aids or delicate and
intricate experiment, or progress i n these areas, access to the w o r l d of the
atom may well have seemed impossible. I f underdetermination is the
worry, t h e n there is a particular and unresolvable problem w i t h fire
particles and a general b u t resolvable problem w i t h practical tests. Even
i f colours were the issue at Timaeus 68b-d, these considerations i f duly
reformulated w o u l d also l i m i t the extent to w h i c h there is a doubt about
practical tests i n general.
 2. Astronomy, Observation and Experiment 73
None of this is to deny t h a t there are weaknesses i n Platonic conception
of science i n relation to experiment, j u s t as there are w i t h v i r t u a l l y a l l
Presocratic conceptions of science. A l l are too optimistic about w h a t
94

theorising can achieve and a l l fail to constrain or test t h e i r theories

properly by experiment, even i f they have commonplace observations to
support t h e i r theories. W h a t I do w i s h to deny here is t h a t we can single
out Plato as an exception i n h a v i n g a highly anti-empirical attitude. There
is no ban on or denigration of observation i n Republic V I I , nor is there a
general b a n on experiment at Timaeus 68b-d. This is i m p o r t a n t , for as we
shall see, these passages are m u c h quoted i n diverse circumstance as being
indicative of Plato's anti-empirical attitude. N o r can we cannot support an
anti-empirical reading of either by reference to the other, or anti-empirical
readings of other passages by reference to these critical passages. Plato
has a coherent conception of technai which is consistent w i t h both of these
positions. There is b o t h the possibility and the necessity of l e a r n i n g
something by empirical means before we move on to the higher and more
theoretical p a r t of the discipline. Thus Plato does not, as F a r r i n g t o n puts
i t , 'substitute mathematics for physics', nor does he exclude observation
95

and experiment from science. The anti-empirical charges against Plato's

96

conception of science, as far as they single h i m out as an exception against

the background of the Presocratics, are i l l founded.
 3

Meno's Paradox and

Underdetermination

The m a i n a i m of this chapter is to argue t h a t Plato had epistemological

reasons for employing teleology There are considerable similarities be­
tween some of the problems posed by Meno's paradox and a problem i n
m o d e r n philosophy of science k n o w n as u n d e r d e t e r m i n a t i o n . W h i l e
Meno's paradox is broader and more general, i t can be seen to have the
following problem is common w i t h underdetermination. W h i c h hypothe­
ses, from an indefinite field of hypotheses, should we adopt for further
consideration, and w h i c h one should we finally accept? The modern prob­
l e m gives us the difficulty of w h i c h criteria we ought to employ w h e n
experiment underdetermines theory. W h i c h solution ought to be adopted,
and indeed whether there is a solution here are hotly debated issues. 1

There are those who hold t h a t while scientists do employ non-empirical

criteria i n theory selection, the choice of these criteria cannot be justified
i n any absolute manner. I t is a purely pragmatic, i n s t r u m e n t a l or conven­
t i o n a l choice, to provide simpler theories and easier calculations. These
theories may be more convenient for humans to w o r k w i t h , b u t there is no
question of whether they are true or false. Some hold t h a t theory choice is
determined by sociological rather t h a n intellectual factors and so argue
t h a t we should pursue a sociology of science, or more radically a sociology
of knowledge. Such positions are i n general associated w i t h some form of
i n s t r u m e n t a l i s m , anti-realism or r e l a t i v i s m . Those on the other h a n d
2

w h o believe t h a t some form of non-empirical c r i t e r i a for theory choice

can be j u s t i f i e d t e n d towards some form of r e a l i s m . Plato believes t h a t
he has a n answer to Meno's paradox. T h i s is i m p o r t a n t , as i t affects the
w a y i n w h i c h Plato t h i n k s about certain scientific problems, especially
i n cosmology. I n the Timaeus Plato several times faces the p r o b l e m of
choosing from an i n d e f i n i t e field of possibilities ( w h i c h cosmos/ w h a t is
the n a t u r e of order/ the u l t i m a t e particles/ celestial m o t i o n ) and i n each
case proposes a n answer w h i c h is shaped by his s o l u t i o n to Meno's
paradox.
 3. Meno's Paradox and Under determination 75

I. T h e modern underdetermination problem

The modern problem of underdetermination i n the philosophy of science,
i n its simplest form, can be stated quite baldly. However m u c h empirical
data one has, t h a t on its own is insufficient to determine w h i c h theory, l a w
or explanation to adopt. Thus this problem is often referred to as the
underdetermination of theory by data. The modern form is perhaps best
3

i l l u s t r a t e d by the graph d r a w i n g problem. Here we have a set of data from

4

an experiment, and we would like to determine the mathematical relation­

ship between x and y. We can draw a straight line t h r o u g h the data, b u t
5

the difficulty is t h a t there are many relationships w h i c h we can postulate

w h i c h w i l l a l l account for the data. I t should be evident t h a t not only can
one f i t many curves to the data, but t h a t one can generate indefinitely
many such curves. The equations for those curves w i l l be more complex
t h a n for a straight line, but w i t h no further criteria t h a n simply accounting
for the data, they cannot be r u l e d out. While one can draw a straight line
t h r o u g h the data, there are indefinitely many other possible ways of
accounting for the data (see Figs 5 and 6). How can we justify a preference

Fig. 5. Data plot. Fig. 6. Data plot with curves.

for any one way? This is a critical problem, for we m i g h t hope to generate
laws of nature i n this manner. Figures 5 and 6 could represent F = m a or
V = I R or s i m i l a r relationships thought to be fundamental to science. More
data w i l l not be any help here, for while i t m i g h t rule out some sets of
curves, there w i l l s t i l l be indefinitely many left. There is a deliberate
ambiguity i n these diagrams, w i t h lines and curves meeting at each
d a t u m . One could account for the data not only by straight line or regular
curve, b u t by any combination of these, i.e. (between successive points)
straight line, t h e n small amplitude curve, t h e n large amplitude curve.
W i t h no criteria other t h a n accounting for the data, n o t h i n g says we have
to use continuous functions (Fig. 7). Simply by v a r y i n g the amplitude of
any sine-type curve w h i c h passes t h r o u g h the points (and w i t h sufficient
ad hoc m a n i p u l a t i o n , w h i c h again we have no grounds to rule out, there
76 Plato's Philosophy of Science

Fig. 7. More data, still underdeter­

mination. More data does not help
w i t h underdetermination, as there
will still be indefinitely many ways of
accounting for the data, even using
continuous functions.

always is such a curve) one can generate an indefinite number of functions

r e l a t i n g x to y. I n order to determine which is the correct relationship then,
we need some more criteria. However, i f science is conceived as being a
purely empirical exercise dependent only upon the data, t h e n there are no
further criteria.

Fig. 8 Data plot, not saving the

data. Anyone producing experimental
results like these would be very happy
to draw a straight line here, even
though the 'best' straight line does not
go through any of the data points.

Practically, scientists are not often bothered by this philosophical diffi­

culty. I n fact, they are quite happy to do something worse (Fig. 8). Given
this rather fine set of data, no practising scientist would have a problem
w i t h d r a w i n g a straight line t h r o u g h the data. However, the straight line
6

does not go t h r o u g h any of the points and so does not 'save the phenomena'
at a l l . The use of error bars, while they may soften some of the practical
difficulties, w i l l not solve the philosophical problems. E r r o r bars indicate
the supposed amount of experimental error i n a measurement and give a
spread of values around the measurement instead of a point (Figs 9 and
10). Even i f error bars soften the problem about d r a w i n g straight lines
7

w h i c h do not go t h r o u g h a l l the points, they produce a further underdeter­

m i n a t i o n problem allowing many straight lines to be d r a w n t h r o u g h the
error bars.
 3. Meno's Paradox and Underdetermination 77

Fig. 9. E r r o r bars plot. Fig. 10. Underdetermination with

error bars.

One s t i l l has the problem, however, t h a t one can draw indefinitely many
curves t h r o u g h the data. While i n this situation one can draw a straight
line ( i n fact indefinitely many straight lines!) t h r o u g h the error bars, one
can easily conceive of a situation where t h a t w o u l d not be possible. 8

Another theoretical, i f not practical difficulty is t h a t of the 'rogue' result.

Anyone who has ever t r i e d to get an experiment to w o r k w i l l recognise data
like t h a t shown i n Fig. I I . Such results are always likely to occur,
9

especially where small external factors (e.g. dust) have a critical effect.
Given underdetermination though, w h a t grounds do we have to rule out
'rogue' results, as is commonly done? I t is easy to recognise w i t h the rogue
result t h a t something has gone wrong w i t h the experiment. I t is less easy
to formulate and justify criteria for excluding such results. Again, i f
10

science is entirely data driven t h e n we have no criteria for excluding such

results.

Fig. 11. Poor data plot.

So far we have looked at simple mathematical relationships between

two variables, b u t i t should be evident t h a t the basic problem here w i l l
apply to more complex theories, and indeed purely verbal explanations as
well. For any theory or explanation, by m a k i n g sufficient ad hoc modifica­
tions, or even by simply adding material, one can generate an indefinite
number of theories or explanations w h i c h w i l l cover the facts. 11
78 Plato's Philosophy of Science
How t r o u b l i n g is this form of underdetermination? This depends on
w h a t resources one believes are available. I f for instance one takes the
position (at least w i t h i n science) t h a t our only proper guides i n theory
choice are experimental data and logic, t h e n theories w i l l r e m a i n under-
d e t e r m i n e d . I n practice, there are of course f u r t h e r c r i t e r i a w h i c h
scientists use i n theory selection. Here one m i g h t cite simplicity, unity,
elegance, beauty, and relation to other theories, p a r t i c u l a r l y at a higher
l e v e l . Unfortunately this is not a straightforward m a t t e r as i t is not easy
12

to capture exactly the sort of simplicity or u n i t y t h a t ought to be employed,

nor do straightforward accounts of simplicity or u n i t y give a good descrip­
t i o n of scientific progress. The nature and status of these criteria is t h e n
a m a t t e r of considerable debate.

I I , T h e Duhem-Quine thesis
There is another form of underdetermination, k n o w n as the Duhem-Quine
thesis. I t is sometimes held t h a t while H u m e has shown t h a t t r y i n g to
prove a theory w i t h positive instances is f u t i l e , the Duhem-Quine thesis
13

shows t h a t t r y i n g to disprove a theory w i t h negative instances is s i m i l a r l y

doomed. The central tenet of the D u h e m thesis is:

An experiment in physics can never condemn an isolated hypothesis but only

a whole theoretical group. 14

I f the group is falsified, we do not know w h i c h theory i n the group to reject

and w h i c h to retain, and so theory choice w i l l again be underdetermined.
Quine too is a holist i n this sense:

The failure of a prediction falsifies only a block of theory as a whole, a

conjunction of many statements. The failure shows that one or more of those
statements is false, but it does not show which. 15

As I argue t h a t Plato believes he has a solution to the problem of positive

instances, I shall not be greatly concerned w i t h negative instances i n w h a t
follows. The problem of negative instances is specifically a t w e n t i e t h -
century one, largely i n response to Popper's falsificationism, a t t e m p t i n g to
show t h a t no theory can be conclusively falsified. However, Gillies has
argued t h a t w h a t is often taken as the Duhem-Quine thesis are actually
two separate theses, and Quine's views are distinctly more r a d i c a l . 16

D u h e m is concerned only w i t h physics, is concerned w i t h groups of theo­

ries, and offers an account of scientific good sense. Quine is concerned
17

w i t h the whole of science, believes the ' u n i t of empirical significance to be

the whole of science' (and not i n d i v i d u a l theories or Duhem's groups) and
18

offers no systematic account of theory choice i n the face of underdetermi­

nation. I n support of underdetermination, Quine notoriously held that:
 3. Meno's Paradox and Underdetermination 79
Any statement can be held true come what may, i f we make drastic enough
adjustments elsewhere in the system. 19

Of course anyone can hold any statement, b u t as L a u d a n has argued, the

key m a t t e r is whether i t is rational to do so. I n support of his view Quine
20

suggests t h a t we can a r b i t r a r i l y redefine terms i n the theory, abandon 21

the laws of logic or plead h a l l u c i n a t i o n . Plato has ample resources to deal

22

w i t h the question of the r a t i o n a l i t y of any of these procedures, especially

i n a dialectical context. He also has i m p o r t a n t resources to deal w i t h the
resulting position. There are certain parallels between Quine's position
and t h a t of Protagoras. Elements of the strategy Plato employs against
Protagoras i n the Theaetetus, most notably the generalised application of
theses, including reflexive application, and the consideration of how the
resulting positions can deal w i t h the assertion of the negation of the m a i n
thesis or themselves can also be deployed against Quine.
A t first reading, one m i g h t get the impression t h a t Quine is referring
only to scientific statements when he holds t h a t any statement can be held
true. I n fact though there is no restriction on the application of Quine's
underdetermination principles and so they can be applied quite generally
to w h a t others m i g h t distinguish as philosophical and scientific problems.
As Quine rejects the analytic/synthetic distinction, gives us the image of 23

'the t o t a l i t y of our so-called knowledge or beliefs ... is a m a n made fabric', 24

and indeed explicitly states t h a t 'ontological questions are on a par w i t h

questions of n a t u r a l science', t h a t would have to be so. Indeed, i n places
25

Quine is quite explicit about t h i s . So any statement, not j u s t scientific

26

ones, w i l l be underdetermined. This has some unfortunate consequences.

Quine may make comments about simplicity ('total science ... has as its
objective the simplicity of l a w s ) , conservatism (as to w h y we 'adjust one
27

strand of the fabric of science' rather t h a n another), and pragmatism 28

(theory choice where rational is pragmatic), w h i c h would appear to give

29

h i m a reasonable position. He appears to have some criteria for choosing

between systems despite empirical underdetermination. B u t these criteria
too are underdetermined, i f the underdetermination principles apply gen­
erally, and so there is no choice between criteria and so no choice between
r i v a l systems. Secondly, the underdetermination principles themselves
(failure of negation to determine theory, redefining terms, a l t e r i n g logic,
h a l l u c i n a t i n g etc.) are underdetermined (by each other or reflexively).
Finally, i f x = 'any statement can be held true come w h a t may, i f we make
drastic enough adjustments elsewhere i n the system', t h e n the negation of
x can be held to be true, resulting i n paradox. The resulting position is
30

indefensible (there are no longer any criteria to distinguish i t from any

other position). I n such a situation, as Plato is fond of pointing out, there
can be no experts and no learning, and enquiry is h a m s t r u n g . As Plato is 31

well aware from Meno's paradox, and later discussions as w e l l , one has 32

to be very careful not to undermine the possibility of enquiry and l e a r n i n g

80 Plato's Philosophy of Science
altogether. Whether Quine can mount a defence against this sort of
33

attack is another matter. I merely indicate Plato's strategy against such

positions, to show t h a t i t is s t i l l viable i n modern debates.
This approach is also i m p o r t a n t because of other modern trends, not­
ably w i t h i n the sociology of science and the sociology of knowledge. There
are those who would argue, either on grounds of outright r e l a t i v i s m or on
grounds of underdetermination, t h a t as scientists employ non-logical,
34

non-empirical criteria i n theory choice, t h e n theory choice should be

explained by 'social rather t h a n logical factors'. Thus there ought to be a
35

'sociologising of epistemology'. Plato would reject this r e l a t i v i s m (what is

the point of discussion i f a l l theories are equally true?), and so too under­
determination. Epistemology and methodology r e m a i n as philosophical
rather t h a n sociological matters for Plato. Indeed the reflexive argument
pioneered by Plato i n the Theaetetus remains one of the m a i n weapons i n
the battle against factual relativism.

I I I . Meno's p a r a d o x
I f one wishes to be a realist about science, t h e n one must be able to solve
the problems created by underdetermination. Effectively this means being
able to justify the use of some non-empirical criteria. Plato, I contend,
recognises a similar, although more generally expressed problem, w h i c h
he i n i t i a l l y presents as Meno's paradox. This paradox is set out as follows,
at Meno 80d:

Meno: How w i l l you search for this t h i n g , Socrates, not k n o w i n g at a l l

w h a t sort of t h i n g i t is? For w h a t sort of t h i n g t h a t you do not know w i l l
be proposed i n your search? Or even supposing t h a t you should meet w i t h
this t h i n g , how w i l l you know t h a t i t is this t h i n g which you do not know?
Socrates: I understand the sort of point you make, Meno. B u t don't you
see w h a t a contentious argument you introduce here, t h a t i t is not possible
for a m a n to search for either t h a t which he knows or t h a t w h i c h he does
not know? I t is not possible for h i m to search for w h a t he knows, as he
knows i t , and there is no need to search for such a t h i n g . N e i t h e r can he
search for w h a t he does not know, as he does not know w h a t sort of t h i n g
he seeks.

I n context, the m a t t e r under debate is v i r t u e b u t Socrates' reply to Meno

is quite general, and can be taken as a problem for a l l kinds of enquiry.
Meno gives us two problems. H o w does one begin an enquiry (what do we
propose i n our search?), and how does one close an enquiry (how do we
recognise the answer w h e n we meet it?). One of the problems i n each case
is t h a t of theory choice. There are a range of theories one m i g h t adopt,
either as candidate theories at the outset of an enquiry or as final theories
at the close of an enquiry. W h a t criteria do we have for theory choice? I f we
 3. Meno's Paradox and Underdetermination 81
accept Socrates' distinction between k n o w i n g (fully) and not k n o w i n g
(anything), t h e n either we have no need for enquiry or we have (as yet) no
criteria for distinguishing between the merits of theories, and so theory
choice is radically underdetermined. There are of course differences here
w i t h the modern underdetermination problem. Meno's paradox is more
general, dealing w i t h a l l types of enquiry, is not specifically set up as a
problem about experimental evidence, and there is no m e n t i o n of choice
from an indefinitely large number of theories. Meno's paradox can also be
seen as being regressive. I f we do not have the criteria f o r j u d g i n g theories,
t h e n we must search for them. B u t how do we begin and end t h a t search?
I t does though give us the same problem about how to finish an enquiry -
w h a t criteria do we use to select our final theory? So too Meno's paradox
sets a problem for s t a r t i n g an enquiry, i n terms of how we select candidate
hypotheses for testing, whether the testing be philosophical or experimen­
t a l . This is a significant b u t perhaps underrated problem for modern
philosophy of science, p a r t i c u l a r l y where theoretical assumptions need to
made i n order to facilitate experimental design.
One m i g h t argue t h a t modern science is spared some of the more radical
consequences of Meno's paradox, as we now know which sort of theory or
explanation we are looking for, even i f there is an underdetermination
problem w i t h the final form. While t h a t may be so for the majority of
science, an exception to this may be cosmogony, and i n particular the b i g
bang. A good deal of our mathematical physics becomes inapplicable, as
the zeros associated w i t h the i n i t i a l singularity generate i n f i n i t i e s , and
36

as neither t i m e nor space are considered to exist prior to the b i g bang,

causal explanations seem inappropriate as well. I f there was l i t e r a l l y
n o t h i n g prior to the b i g bang, i t is h a r d to see w h a t determined i t (espe­
cially i f the laws of physics d i d not exist). Thus we do not know w h a t sort
of explanation to propose or to accept for the b i g bang. The typical reaction
- Ts t h a t really an explanation? - to some modern cosmogony w o u l d seem
to be a healthy one, and an indication t h a t Meno's paradox s t i l l has some
bite beyond the more standard problem of underdetermination.
So how does Plato answer Meno's paradox for the investigation of the
n a t u r a l world? I n the Timaeus, we have seen choices from indefinite fields.
So we have the choice of two sorts of fundamental particle, one cosmos, and
the slightly less evident choice about the origin of humans. The procedure
here is clear enough. We must, i n the first instance, hypothesise w h a t we
believe to be best, as is made clear at Timaeus 54ab i n relation to the
fundamental particles. A t this point the Timaeus is quite tentative, and
says i f anyone can hypothesise better particles, he w i l l be treated as a
friend rather t h a n an enemy. The next task is to see how good a theory
these particles can generate, whether this theory coheres w i t h others and
whether we can have plausible explanations of the phenomena. This I take
to be one of the i m p o r t a n t points of giving an extended discussion of w h a t
we can do w i t h the theory of geometrical atomism i n the l a t t e r p a r t of the
82 Plato's Philosophy of Science
Timaeus. The nature of the good t h a t Plato invokes w i l l be an i m p o r t a n t
concern here, b u t first I w a n t to look at how Plato's application of the good
develops. I f we go to the Meno, Plato's first answer to Meno's paradox is
the theory of recollection. For the purposes of this book, I take this to be
37

a relatively short-lived and uninteresting answer. I f we go to the Republic,

we find a fully fledged epistemology w i t h hypotheses leading to the
cognition of the form of the good. The Republic perhaps reflects one of the
key differences between Plato and Quine. Quine believes that:

We have no reason to suppose that man's surface irritations even unto

eternity admit of any systematisation that is scientifically better or simpler
than all possible others. I t seems likelier, if only on account of symmetries or
dualities, that countless alternative theories would be tied for first place. 38

Plato supposes j u s t the opposite, t h a t there is one and only one system
w h i c h is best. Whether he is justified i n doing so is another matter, t h o u g h
i t is w o r t h n o t i n g t h a t at least w i t h i n modern science this is a m a t t e r of
contention, as many would hold t h a t there is a unique unified u l t i m a t e
physical theory, a theory of everything (TOE) w h i c h physics is heading
towards, and indeed may be relatively close to achieving. Meno's paradox 39

bites here too, as there is considerable debate about w h a t the general form
of such a theory w o u l d be. Instead of the Republic though, I propose to
begin w i t h the Phaedo. The Phaedo has some interesting points to make
balance of explanation and ontology, and highlights differences w i t h the
physiologoi.

IV. S o c r a t e s ' a u t o b i o g r a p h y
A t Phaedo 96a ff., Plato has Socrates recount his experiences of e n q u i r i n g
into w h y things should come to be and pass away. Socrates first tells of his
youthful i n f a t u a t i o n and subsequent disenchantment w i t h peri phuseos
historian, the 'enquiry concerning nature'. He t h e n hears of the works of
40

Anaxagoras, and hoping to find teleological explanations, rushes to read

his books, b u t is disappointed. Unable to find or be t a u g h t the sort of
explanations he requires, Socrates sets out on his deuteros pious, a second
voyage or perhaps second best way. The protos pious, the 'first voyage' has
been identified w i t h the views of the physiologoi w h i c h interested Socrates
i n his youth. This i n t e r p r e t a t i o n runs into a major difficulty though, as the
deuteros pious involves both the method of hypothesis and the theory of
forms. I t w o u l d seem unlikely t h a t Plato would have Socrates describe
these as a second best to the physiologoi. There are t w o standard strate­
gies for dissolving this tension. The first is to suggest t h a t the deuteros
pious description is ironic and so is not to be t a k e n seriously, the second to
postulate t h a t the protos pious is some unattainable ideal, relative to
w h i c h the deuteros pious is a second best. W h a t I w a n t to suggest instead
41
 3. Meno's Paradox and Underdetermination 83
is t h a t Socrates' autobiography is p r i m a r i l y a discussion of ontology and
explanation. The question I believe is being posed is: W h a t is the m i n i ­
42

m u m number of entities we must postulate to have a theory t h a t w i l l do

a l l the explaining we require of it? The central issue is t h e n a balance
between ontological parsimony and explanatory richness. The deuteros
pious may t h e n be a second attempt at this balance, explanatorily richer,
b u t r e q u i r i n g a greater ontological commitment.
One way of proceeding w i t h this programme is to begin w i t h a m i n i m u m
number of entities, to examine how they are explanatorily deficient, and
t h e n to b u i l d a richer ontology specifically designed to remedy these
inadequacies. This is indeed w h a t we find i n Socrates' autobiography. The
puzzles w h i c h Socrates cannot solve at Phaedo 96d ff. may seem a rather
odd collection. Undoubtably, part of the function of these puzzles is to help
us formulate w h a t is an adequate explanation. Some criteria of adequacy
here are t h a t the possession of a property ought not to have two opposite
explanations, as when both division and addition are cited as reasons for
something being two, nor should a single explanation explain the posses­
sion of opposite properties, as when 'a head' is cited as the reason for one
person being small and another being large, nor should the explanation
cite the opposite of the property. Something cannot be large by v i r t u e of
something s m a l l . Why, though, does Socrates mention some clearly
43

inadequate explanations i n each case? I f we are restricted to a parsimoni­

ous ontology these may be the only sorts of explanation available. I t is also
interesting to note the nature of the account of m i n d t h a t is on offer here.
A t Phaedo 96b, we find t h a t among the questions t h a t Socrates asked
himself i n his y o u t h are:

Is i t blood with which we think, or air, or fire, or is i t none of these? Is i t the

brain (ho engkephalos) which grants the sensations of hearing, sight and
smell, are memory and opinion produced from these, and is i t from memory
and opinion acquiring stability that knowledge is produced?

Note the use of ho engkephalos, b r a i n , rather t h a n Plato's usual he psuche,

mind/soul here, and the external and causal explanation of the acquisition
of knowledge. Knowledge here is a state t h a t occurs due to the causal effect
of perceptions eventually creating stable opinions i n the b r a i n , rather t h a n
being due to any activity of the m i n d . The discussion t h e n takes a detour
via Anaxagoras. Socrates is excited by the prospect of teleological explana­
tions on hearing t h a t Anaxagoras believes t h a t intelligence orders every­
t h i n g . He is disappointed though, and at Phaedo 98b says:
44

My wonderful hopes came to ruin, since proceeding with my reading I

perceived a man neither making use of (his) intelligence, nor assigning to i t
any reason for the order of things, but giving the reason as air, aether, water
and many other absurd things. 45
84 Plato's Philosophy of Science
I take i t t h a t Plato has t w o objections here. First, Anaxagoras gives
explanations couched i n the same inadequate, entirely physical terms as
those of the physiologoi. Secondly, he postulates an e n t i t y w h i c h does no
explaining. Anaxagoras is reputed to have remarked t h a t though a l l m e n
have intellect not a l l of t h e m use i t . This m i g h t be seen as a proto-
4 6

Ockhamite objection, t h a t Anaxagoras is m u l t i p l y i n g entities beyond ne­

cessity. This has an i m p o r t a n t function as a contrast to the lessons of the
physiologoi's ontology. The reader has now been armed w i t h the criteria
w i t h w h i c h to judge any future proposals. A r e they too parsimonious and
explanatorily deficient, or are they too profligate, w i t h entities t h a t are
inefficient or even non-explanatory? Closing the first part of his autobio­
graphy at Phaedo 99c, and h a v i n g bemoaned the fact t h a t no one has given
h i m proper teleological explanations, Socrates says:

They do not truly believe that i t is the good (agathos) and proper which binds
and holds everything together. I would most gladly become anyone's student
concerning such a reason and how i t prevails; but since I was deprived of
this, neither able to find i t myself nor to learn i t from any other, would you
like, Cebes, for me to demonstrate how I worked out and created for myself
a second voyage in search of explanation?

W h a t one w o u l d expect i f Plato is following the programme I have sug­

gested is t h a t some more entities w i l l now be postulated, designed to
remedy the deficiencies of the views already examined, and this t h a t we
find r i g h t at the beginning of the deuteros pious. A t Phaedo 100b Socrates
says:

I hypothesise there to be something beautiful itself by itself and similarly a

good (agathos) and a large and all the others.

T h a t these new entities are designed to help solve some puzzles is uncon-
troversial. A t 100c ff., Socrates goes t h r o u g h the difficulties introduced at
96d ff. and shows how we can know have some adequate explanations
w h i c h cite forms as explanatory entities.
Now consider the relation of the deuteros pious to the proposed protos
pious. The deuteros pious is more profligate of entities t h a n the protos
pious, b u t is also more explanatorily sufficient. There is a sense i n w h i c h
i t is a second best relative to the protos pious, i n r e q u i r i n g greater
ontological commitment, b u t also a sense, i n terms of explanatory ade­
quacy, i n w h i c h i t is better. The deuteros pious is also secondary i n terms
of exposition i n terms of the programme t h a t I suggest Plato is engaged i n
here, and secondary i n terms of the chronology of the metaphor of Socrates'
autobiography, w i t h o u t necessarily being a second best.
We can also consider the deuteros pious as an epistemological metaphor.
One meaning of the phrase is to take to the oars when there is no w i n d to
move the s h i p . Thus when the easy methods w h i c h we hope w i l l b r i n g us
47
 3. Meno's Paradox and Underdetermination 85
to our destination w i t h o u t doing any work ourselves fail, we must make
headway by our own active efforts, 'row our own boat' i f you like. One sense
m i g h t be t h a t w h e n we cannot be taught, we must actively seek knowledge
ourselves. This has deep resonances throughout Plato's work, from the
relation of the Socratic elenchus to the experts, to the Apology, to Meno's
paradox, t h r o u g h to the Theaetetus and the criticism of Protagoras. L i k e
using the sails, being taught is easier, b u t should this fail to provide
understanding, t h e n we must take to the oars and play an active part
ourselves. Note the relative ontological commitments here. Being t a u g h t
requires only a passive account of the b r a i n and a causal account of the
acquisition of knowledge, like t h a t of the physiologoi at Phaedo 96b, while
the deuteros pious requires a more active account of m i n d .
One can also consider this metaphor i n the following manner. I f sense
perception is inadequate for w h a t we require, t h e n we must have recourse
to reason, again r e q u i r i n g a more active account of the m i n d . Doubtless
the deuteros pious description is ironic, b u t Platonic irony is not a binary,
on/off phenomenon and the dichotomy of the serious Socrates giving us
Platonic doctrine against the ironic Socrates who is not to be believed is
too simple. Rather, Plato's irony can focus our attentions on ambiguities,
48

forcing us to do some reasoning for ourselves i n order to arrive at a

consistent account. The deuteros pious is presented as simultaneously a
best and a second best, and the irony focused on this description forces us
to sort out i n exactly w h i c h separate respects the deuteros pious is best and
second best. We are led to consider the ontological cost of explanatory
adequacy.

V. Teleology a n d the Phaedo

Does the deuteros pious give us any teleology? A t Phaedo 97d Socrates
makes i t clear i n discussing Anaxagoras t h a t he is interested i n teleology:

Considering these things, I was delighted to think I had found in Anaxagoras

a teacher of the reason for things in accordance with my own way of thinking.
I thought that firstly he would tell me, whether the earth is flat or round,
and having told me this, he would explain the reason and the necessity,
stating the nature of the better and why i t is better for the earth to be as i t is.

Earlier, I suggested t h a t Plato hypothesises the entity required for such

explanations, the agathos, as the deuteros pious gets under way. One m i g h t
argue t h a t i f Plato makes no use of this agathos, t h e n he is prey to the same
objection t h a t he deploys against Anaxagoras, t h a t of postulating an e n t i t y
w h i c h does no explanatory work. However, i t may be t h a t j u s t as the
puzzles of 96b ff. were given solutions i n terms of the newly hypothesised
entities, so the puzzle of 97d is resolved at Phaedo 108e ff. Socrates says
that:
86 Plato's Philosophy of Science
I am now persuaded that, first, if the earth is round and in the centre of the
heavens, then it requires neither air to prevent i t falling nor any other
necessitation of this sort, but the uniformity of the heaven itself in every way
and the equipoise of the earth itself is sufficient to restrain it. For something
which is equipoised and is placed in the middle of something homogeneous
cannot yield to being moved aside in any way, but will remain steadfast.

T h a t Socrates discusses the shape, position and stability of the earth, the
three things mentioned for explanation at 97d ff. can h a r d l y be a coinci­
dence, especially as Socrates also comments t h a t his theory has no need of
air to prevent the earth from falling, perhaps a reference back to Phaedo
99b where he attacks the vortex and 'kneading-trough' theories:

One man makes the earth stay in position by means of the heavens, employ­
ing a vortex, and another has i t flat like the lid of a kneading trough, 49

resting on the air for support. But as for the power by which things are now
situated in the best way the could be, they neither search for i t nor attribute
any special force to it. They believe they will find some Atlas stronger, more
immortal and better able to hold everything together, and they do not truly
believe that it is the good and proper which binds and holds.

Specific targets here may be Empedocles and Anaxagoras, though the

criticism is quite general and Plato may have others i n m i n d too. Con­ 50

cerning the Phaedo 108e ff., Sedley has argued t h a t Plato gives a teleologi­
cal explanation:

The earth is stable because its stability makes i t an integral part of the best
possible overall arrangement of the cosmos. 51

According to Sedley, the teleological reading he derives here is not evident,

b u t Plato has left enough clues for w o r k i n g out the true explanation of the
earth's stability. Perhaps the m a i n objection to the idea t h a t the Phaedo
countenances teleological explanation is t h a t he has Socrates say t h a t he
was 'deprived' (esterethen) of i t , 'neither able to find i t myself nor to learn
i t from any other'. I would agree w i t h Wiggins against Vlastos on the
52

t r a n s l a t i o n of Phaedo 99c. The aorist esterethen T was deprived' of

53

teleology does not prejudice any future venture. The reading Vlastos
argues for, t h a t Socrates is for a l l t i m e deprived of teleology requires a
perfect tense, i.e. esteremai. So while the protos pious was devoid of
teleology, t h a t is not necessarily so for the deuteros pious, especially i f the
deuteros pious is being more generous i n terms of ontology so allowing
fuller explanation. W h a t is different about the deuteros pious? I t appears
to be more active. Socrates has failed to find out about teleology by being
someone's pupil, now he w i l l t r y to find out for himself. B u t does he not t r y
to find out for himself i n the protos pious? I n a sense yes, b u t i t is i m p o r t a n t
to recognise how he tries to do this. W h e n Socrates refers back to his
youthful experiences at Phaedo 99e he says:
 3. Meno's Paradox and Underdetermination 87
I was afraid that I might totally blind my soul (psuchen), looking at things
with my eyes and endeavouring to grasp them with each of my senses.

He t h e n goes on to say t h a t i n the new enquiry he 'took refuge i n theories',

w h i c h marks the beginning of the method of hypothesis. So instead of a
merely empirical enquiry or being taught (both of w h i c h would require
only a passive account of the b r a i n , note t h a t psuche, soul, returns i n place
of engkephalos, b r a i n , here) we are to have a theoretical enquiry where
Socrates w i l l have to do some active theorising, one i n w h i c h he certainly
theorises the existence of the agathos, and may i n the m y t h give an
indication of the application of the agathos. I t w o u l d h a r d l y be any great
surprise t h e n i f we need to do some of this w o r k ourselves, to take a more
independent and active approach to the text i n order to uncover some
teleology, or as Sedley puts i t :

[Plato] expects his reader to work to uncover the deeper-seated teleology of

the myth. 54

Such readings are of course impossible to prove, and as Fine has pointed
out, i t may be the case t h a t Plato is merely offering us an account he sees
as preferable to those of the physiologoi, w i t h o u t yet being teleological. 55

Whatever the position of the Phaedo here, clearly there are significant
developments by the Republic and further developments i n the Timaeus.

V I . Healthy hypotheses
Whether or not the Phaedo gives us any teleology, i t does have some
i m p o r t a n t points to make about method. I f i t is the case t h a t Plato
advocates a coherence theory of t r u t h i n the deuteros pious, t h e n this may
shed some l i g h t on one of the perennial difficulties associated w i t h the
method of hypothesis, the meaning of sumphonein at Phaedo 100a5 ff. 56

Socrates says:

Therefore this is the way I have begun, hypothesising in each case whichever
account (logos) I consider to be the healthiest (erromenestaton), and what­
ever seems to me to agree (sumphonein) with this, I posit as being true, and
whatever does not, as not true, whether concerning the reason for or any
other entity.

The problem here is t h a t the logical relation w h i c h sumphonein denotes

w i l l w o r k w e l l for one a r m of this procedure, b u t not the other, whether we
take this relation to be entailment or consistency. W i t h entailment, a l l t h a t
is entailed by a true hypothesis is true, b u t many things not so entailed
are also true. Further, as Gentzler points out, the set of propositions not
entailed by our hypothesis w i l l contain pairs of negations, and i t cannot be
reasonable to posit both members of these pairs as false. W i t h consis-
57
88 Plato's Philosophy of Science
tency, a l l t h a t is inconsistent w i t h a true hypothesis is false, b u t many
things t h a t are consistent w i t h i t are also false. I t may also be the case t h a t
w h i l e B and C are consistent w i t h our hypothesis A, they are not consistent
w i t h each other. Socrates also says t h a t he can demonstrate the i m m o r t a l ­
i t y of the soul using this method. Consistency though w i l l merely leave us
w i t h a set of assertions, not a proof. Such considerations have led some
58

to conclude t h a t we cannot read sumphonein univocally t h r o u g h Socrates'

account. Robinson comments:
59

Both our interpretations of the metaphor of accord and discord have run into
grave paradox ... There is no third interpretation. We have to choose be­
tween consistency and deducibility as the meaning of'accord'. The better is
consistency, as the paradox to which i t leads is much less than to which
deducibility leads.60

However, i f Plato has i n m i n d a coherence theory of t r u t h t h e n the problem

dissolves. Sumphonein simply means 'coheres', and i t is definitionally t r u e
i n a coherence t r u t h system t h a t whatever coheres is true and whatever
does not is not true. Where Plato wants to express logical consequence he
usually employs sumbainein. 61
Gentzler gives the following definition of
coherence:

A proposition P coheres with Q if and only if P is consistent with Q and stands

in a suitable inductive or deductive relation to Q. 62

The broader the conception of t h a t which has to do the cohering, perhaps

the easier i t is to see how such coherence could work. W h i l e i t is possible
to argue t h a t propositions may cohere, i t is perhaps easier to see how
accounts m i g h t generate coherence. The question t h e n is how we should
translate logos at 100a4. The narrow translation, given by Robinson, is
'proposition', where I have favoured the broader 'account'. Some points i n
favour of the 'account' view. I f one t h i n k s i n terms of the Meno, t h e n
k n o w i n g something means being able to logon didonai, generally rendered
'to give an account', w h i c h suggests more t h a n a proposition. Similarly, at
101d6 i n the Phaedo, w h e n one has to give an account (logon didonai) of a
hypothesis, one would expect more t h a n a single further proposition. I t
may well be t h a t hypotheses are complex too, involving more t h a n one
proposition. I f so, t h e n Phaedo 1 0 I d , where we are meant to examine the
consequences of an hypothesis, whether they agree w i t h one another or not
makes more sense. While one proposition cannot have contradictory con­
sequences, complex hypotheses can. 63

Gentzler raises the w o r r y t h a t there w i l l be an infinite set of proposi­

tions t h a t are consistent w i t h P but w h i c h do not stand i n a suitable
inductive or deductive relation to P, and so must (unreasonably) be reck­
oned false, and that:
64
 3. Meno's Paradox and Underdetermination 89
Perhaps Plato did not assume a clearly defined decision procedure for
restricting them to a manageable number; rather he took i t for granted that
his audience would have an intuitive, if rough idea of the difference between
the logical consequences of our hypotheses that were worth considering and
those that were not. 65

W h i l e this is a different problem from f o r m u l a t i n g hypotheses i n the first

instance, again one can see how reference to the good m i g h t be i m p o r t a n t
for Plato here. The clearest passage on the later decision m a k i n g process
comes at Epinomis 991d ff. We are told t h a t there is an i m p o r t a n t matter,
w h i c h i f conceived correctly is a great boon to men, and i f not we can only
call upon the gods. To those who learn i n the proper manner, every
diagram, every system of number and harmony, and the ordered motions
of the stars w i l l reveal themselves to be part of a u n i t a r y and coherent
(homologian) whole, as there is a single bond between these things, and
we m u s t keep t h i s i n m i n d while investigating. The conclusion is that:

If one goes about this in some other way, i t is necessary to call upon luck, as
we say.

One m i g h t compare this w i t h other passages, most notably Gorgias 503d

ff. where those who are properly skilled w o r k w i t h the best i n m i n d , and
do not do things at random or produce random effects. Those who do not
aim at the best, and are not i n possession of the proper skills can only l e a r n
by empeiria kai tribe, an empirical knack and g r i n d i n g away, t h a t is
effectively by luck and t r i a l and error. We m i g h t also look to the Republic
66

where knowledge is structured around the good (or perhaps the structure
of knowledge/the w o r l d is the good), and i t is notable t h a t we have the first
appearance of the demiurge ordering the heavens i n best way possible way
at Republic 530a. I t is also i m p o r t a n t t h a t the demiurge, the skilled
w o r k m a n god, creates taxis and cosmos i n the universe j u s t as skilled
craftsmen (e.g. doctors) create the same attributes i n t h e i r subject matter.
This of course is the key l i n k for Plato. I f the w o r l d has been structured for
the best, t h e n we can have an appropriate epistemology w h i c h uses
reference to the good i n order to overcome such problems as underdeter­
m i n a t i o n . One interesting aspect of this is the contrast of chance and the
67

good i n both cosmology and epistemology. W i t h o u t the good any structure

i n the universe is a matter of (highly improbable) chance, and i f we do not
look to the good t h e n learning too w i l l be a matter of chance. I f one believes
t h a t l e a r n i n g is not a matter of chance, and t h a t there is such a t h i n g as
expertise, t h e n we m i g h t t u r n this the other way round. I f we w a n t an
epistemology w h i c h w i l l be able to give us realist answers, t h e n we m u s t
suppose the cosmos to have good order. Thus Plato has epistemological
reasons, rooted i n Meno's paradox/underdetermination for teleology. One
m i g h t also trace this back to the original Greek idea of the cosmos as an
ordered whole t h a t is understandable by h u m a n beings. For Plato, the
90 Plato's Philosophy of Science
cosmos is an ordered whole because i t is teleologically ordered, the ac­
counts of the atomists and physiologoi being inadequate. The cosmos also
needs to be teleologically ordered so t h a t h u m a n beings can come to
understand i t . As we saw i n Chapter 1, Plato takes the order/chaos
analysis to a fundamental level, of the intrinsic as well as the extrinsic
order of the constituents of the cosmos. Plato does not betray the original
idea of the cosmos, b u t he does see some of the difficulties associated w i t h
i t and attempt some solutions. One m i g h t see here considerable affinities
w i t h modern cosmological anthropic principles.
How m u c h of this does the Phaedo have? Here an i m p o r t a n t question is
how to translate and interpret erromenestaton at 100a4. Robinson here
translates 'strongest', which w o r k i n g w i t h i n a correspondence theory of
t r u t h is fair enough. However, erromenestaton also has connotations of ' i n
good health, vigorous, stout', and I have translated 'healthiest'. W i t h i n a
coherence theory of t r u t h based on relationship to the agathos, t h i s has
i m p o r t a n t connotations, of selecting the 'healthiest' account most l i k e l y to
cohere w i t h 'for the best' explanations. Perhaps a good example of this is
Phaedo 108e and the question of the earth's shape and position. Plato does
not have Socrates propose any random shapes or positions here, b u t ones
w h i c h are l i k e l y to produce coherence, 'healthy' hypotheses framed w i t h
'for the best' explanation i n m i n d . I f Sedley is correct about teleology i n the
Phaedo, t h e n we see the beginnings of a matching of a teleological struc­
t u r e for t h e w o r l d w i t h a n o r i e n t a t i o n o f e p i s t e m o l o g y t o w a r d s
explanations citing the good. One critical difference between the Phaedo
and the Republic of course is t h a t i n the Phaedo the agathos is j u s t one
form among many, and not as i n Republic the key form, though one m u s t
note t h a t at 99b ff. Socrates is critical of the physiologoi who do not
a t t r i b u t e any 'special force' to the good, nor do they recognise t h a t i t is the
good w h i c h binds and holds everything together.

V I I . Teleology a n d epistemology
Plato t h e n has epistemological reasons for employing teleology. I n order to
solve Meno's paradox, and i n particular the difficulty of w h i c h hypotheses
from an indefinite field to adopt for further consideration (how to start an
enquiry) and w h i c h one to accept u l t i m a t e l y (how to end an enquiry), he
gives us an account where 'healthy'hypotheses are selected for t h e i r likely
coherence and relation to the good. Vlastos is critical of Plato's use of
teleology i n the following manner:

Suppose we could have asked the physiologoi - any of them from Anaximan-
der down - the following question: 'In your inquiries into nature, when you
have to decide whether or not something is thus and so, would you think i t
right to settle the issue by arguing, "It would be better, more beautiful, i f
things were thus and so; ergo, they are thus and so?" ' There can be little
doubt that we would have got an emphatic 'No' from each. 68
 3. Meno's Paradox and Underdetermination 91
This is not quite Plato's position on healthy hypotheses, b u t the criticism
could be easily reformulated to take account of that. I f I may suggest
Plato's reply, i t m i g h t be T i n e . B u t now tell me how you are going to resolve
such matters. How are you going to justify a choice from an indefinite
number of possibilities? I f you don't have an answer, your position is
indeterminate.' This w i l l r e m a i n a standard i m p l i c i t criticism of the
69

atomists and the physiologoi on several issues for Plato. Just how do they
make a choice about conceptions of order/chaos, the nature of matter, the
behaviour of matter, etc., and can this be justified? To r e t u r n to some issues
from Chapter 1. H o w can the supposition of a vortex, rather t h a n any other
form of motion, be supported? O n w h a t grounds does Anaxagoras' nous
choose the ordering of the cosmos? There we saw t h a t Plato has teleological
answers to epistemological problems i n cosmology and cosmology. One
point of this chapter has been to explore the grounds t h a t Plato has for his
method i n approaching certain problems i n cosmology and cosmogony.
How does this account relate to modern realist positions i n the philo­
sophy of science? Here we need to look more to the Republic t h a n the
Phaedo. One modern view is t h a t the u l t i m a t e goal of science is the final
theory or theory of everything (TOE). This would sit atop a p y r a m i d a l
structure of scientific knowledge, w i t h other theories of physics, t h e n
chemistry and biology supporting i t and being i n t u r n supported by i t as
p a r t of the structure. Some would argue t h a t we are beginning to have
quite a good idea of w h a t such a theory would look l i k e . Certainly i t w i l l
70

involve a unification of q u a n t u m mechanics and general relativity, and i t

is hoped t h a t such a theory would predict the observed values of the
fundamental physical constants. The ramifications of such a theory m i g h t
w e l l be mathematically highly complex, but i t is thought t h a t the founda­
tions w i l l be conceptually simple and elegant. One m i g h t w e l l argue t h a t
this is the case for general relativity and modern theories i n particle
physics. As E i n s t e i n famously put i t , 'God is subtle, b u t not malicious.' 71

For this sort of position, there are several criteria for theory selection i n
order to be able to meet underdetermination. One looks for conceptual
simplicity, coherence w i t h other h i g h level theories, and the ability to unify
existing theories at a higher level. Proponents of this position would argue
t h a t i t gives a good description of scientific progress (the standard problem
being how science can 'progress' i f one incorrect theory follows another) as
more general and unified theories supersede theories t h a t are less so
approaching the T O E .
We m i g h t well compare aspects of the Republic w i t h this conception of
science. The Republic too has a p y r a m i d a l structure, w i t h theories inter­
dependent on one another. The Republic is of course more general, dealing
w i t h a l l of knowledge rather t h a n j u s t science, b u t there is a s i m i l a r
emphasis on unification and coherence w i t h higher level theories. Relative
to the modern realist position I have sketched, Plato's view, as presented
i n the Republic appears overly reliant on the good and ethical and aestheti-
92 Plato's Philosophy of Science
cal judgements. I n later works though the conception of the good appears
to be somewhat more technical, involving unity, mathematics and struc­
ture. I n this context Aristoxenus' comments on Plato's lecture On the Good
may be of some considerable interest:

Each came expecting to learn something about the things which are gener­
ally considered good for men, such as wealth, good health, physical strength,
and altogether a kind of wonderful happiness. But when the mathematical
demonstrations came, including numbers, geometrical figures and astro­
nomy, and finally the statement, good is one it all seemed to them, I imagine,
utterly unexpected and strange. 72

Sayre has argued t h a t the comments a t t r i b u t e d to Plato i n his Lecture on

the Good can be understood w i t h reference to the views he puts forward i n
the Philebus. 73
I t is evident early on i n the discussion of the good and the
good life t h a t neither can be equated w i t h any one t h i n g exclusively, such
as intelligence or pleasure, b u t must consist of a m i x t u r e . As the discussion
draws to a close and the good is brought into view, and at Philebus 65a
Socrates comments:

If it is not within the power of one idea to hunt down the good, let us grasp
it with three, beauty, proportion and truth, and let us say that all these as a
unity among the elements of the mixture are most correctly held responsible,
and it is on account of the goodness of these that the mixture itself has
become good.

A t 66a ff. Socrates and Protarchus proceed to r a n k intelligence and

pleasure against the three of this unity. F i r s t come 'measure, the mean and
w h a t is appropriate' (66a7), second are 'proportion, beauty, completeness,
sufficiency' (66b 1), t h i r d come 'wisdom, intelligence and understanding'
(66b6), f o u r t h are 'knowledge, s k i l l and r i g h t opinion' (66b8) and finally
come the most pure forms of pleasure. Now we know from earlier i n the
Philebus t h a t measure is achieved by the imposition of l i m i t , w h i c h is
equivalent to participating i n unity. So i f measure is the p r i m a r y constitu­
ent of the good, then as measure is participation i n unity, so i n one sense
u n i t y m i g h t be equated w i t h the good. Gaiser summarises the m a i n
features of Plato's principles as preserved by the doxographers as follows:

The goodness (arete) of a thing is shown by its permanence, beauty and form.
These qualities depend on order (taxis, kosmos); that is, on a well propor­
tioned arrangement of parts within the whole. The basis of order is then
unity, and thence unity or oneness is the cause of all good, or the good in
itself.
74

This conception of the good, found i n the Philebus and the Lecture on the
Good differs from the position of the middle period, and presents a more
technical and mathematical view of the good. W h e n we then t a l k about the
 3. Meno's Paradox and Underdetermination 93
good structure and arrangement of the universe, we are more likely to
refer to mathematical notions such as proportion and due measure rather
t h a n ethical or aesthetical criteria. The latter may still be relevant t h o u g h
(aesthetical considerations may be i m p o r t a n t for modern scientists too;
there is certainly a disposition to believe t h a t the most beautiful, elegant
theory ought to be correct, including theories about how the cosmos is
structured). One m i g h t here consider t h a t when Timaeus describes the
construction of the w o r l d soul (see Timaeus 35e ff.), he does so p r i m a r i l y
i n mathematical terms of proportion and due measure w i t h other consid­
eration subservient. One would like to be considerably more specific about
the increased role for mathematics i n investigation and explanation, b u t
unfortunately Plato is very vague here, j u s t giving us theoretical outlines
of a greater mathematical orientation. How much of a greater orientation
is another question. The c u r r i c u l u m of the guardians is structured to lead
to cognition of the good, and contains mathematics, astronomy and har­
mony, and indeed w h a t is good for the state is t h a t which binds i t into one
75

(Republic 462b). Perhaps t h e n there is only a change of emphasis, b u t

76

t h a t i n itself may be quite significant.

I f this general view of Plato's later conception of the good is correct, t h e n
i t considerably narrows the gap between Plato's teleology and a realist
conception of modern science, which would w a n t to cite unity, simplicity,
elegance etc. i n order to be able to answer underdetermination. I t does not
close i t , of course, and one would not expect that. I t does show though t h a t
Plato's teleology is no mere 'metaphysical fairy tale'. Vlastos is also con­
cerned that:

The insistence on circle and sphere being the most beautiful and most perfect
shapes gave undue importance to aesthetics. 77

This criticism though seems misplaced, i f the basis of Plato's theory of the
good lies deeper i n ideas of unity, comprehensibility and order.

VIII. Duhem and instrumentalism

D u h e m has argued t h a t Plato began an i n s t r u m e n t a l i s t t r a d i t i o n of 'sav­
i n g the phenomena' i n astronomy, one which was highly i n f l u e n t i a l i n
a n t i q u i t y . Duhem's views have been h i g h l y i n f l u e n t i a l themselves,
78

though Lloyd and Musgrave have now seriously undermined many of his
claims about later astronomers. Here I shall consider Duhem's views on
79

Plato i n detail. D u h e m draws on Simplicius' commentary on Aristotle's Be

Caelo, saying that:

We find the Platonic tradition formulated in the following terms: Plato lays
down the principle that the heavenly bodies' motion is circular, uniform, and
constantly regular. Thereupon, he sets the mathematicians the following
problem: What circular motions, uniform and perfectly regular, are to be
94 Plato's Philosophy of Science
admitted as hypotheses so that it might be possible to save the appearances
presented by the planets? ... The object of astronomy is here defined with
utmost clarity: astronomy is the science that so combines circular and
uniform motions as to yield a resultant motion like that of the stars. When
the geometric constructions have assigned each planet a path which con­
forms to its visible path, astronomy has attained its goal, because its
hypotheses have then saved the appearances. 80

Saving the appearances is t h e n the end point for astronomy done i n this
t r a d i t i o n . D u h e m contrasts this instrumentalist t r a d i t i o n w i t h another
and sums up the differences between the two approaches: 81

The hypotheses of astronomy can be viewed as mathematical fictions which

the geometer combines for the sole purpose of making the celestial motions
accessible to his calculations; or they can be viewed as a description of
concrete bodies and of movements that are actually realised. I n the first case,
only one condition is imposed on hypotheses, namely, that they save the
appearances: in the second, the intellectual freedom of the astronomer turns
out to be much more limited, for if he is an advocate of a philosophy which
claims to know something about the celestial essence, he will have to
reconcile his hypotheses with the teachings of that philosophy 8 2

D u h e m places Plato f i r m l y i n the first of these categories, and considers

the position of the physicist to be like the prisoners i n the cave allegory:

A slave to positive method, the physicist is like the prisoner in the cave: the
knowledge at his disposal allows him to see nothing except a series of
shadows in profile on the wall facing him; but he surmises that this theory
of silhouettes whose outlines are shadowy is only the image of a series of
solid figures, and he asserts the existence of these invisible figures beyond
the wall he cannot scale. 83

For D u h e m , however, such an assertion is invalid. I t exceeds the legiti­

mate methods of the physicist, who must r e m a i n a prisoner i n the cave. 84

Plato does make some comments on the prisoners w h i c h would suggest the
best they could do would be something like i n s t r u m e n t a l i s m , i n particular
Republic 516cd, where there is competition among the prisoners to predict
w h i c h shadows w i l l appear next. B u t for Plato we are not stuck i n the cave,
and we can, and indeed ought to escape. As commented earlier, i f the
85

guardians d i d not derive some benefit from escaping from the cave, such
t h a t they can do better t h a n the permanent inhabitants on t h e i r r e t u r n ,
they w o u l d have no intellectual mandate to govern. As Republic 520c says
t h a t they w i l l 'see far better t h a n those i n t h a t place, and w i l l k n o w each
of the images, w h a t i t is and w h a t i t is of' w h e n they descend, i t seems t h a t
they w i l l be far better off t h e n the prisoners, to an extent incompatible
w i t h i n s t r u m e n t a l i s m . D u h e m may be an i n s t r u m e n t a l i s t on the basis of
this allegory, b u t Plato most certainly is not. To make this more specific to
astronomy, there is no suggestion i n the discussion of astronomy as part
 3. Meno's Paradox and Underdetermination 95
of the education of the guardians t h a t seeking the 'true realities' is not part
of the task of the true astronomer. 86

The dynamism of the enquiry i n the Republic is i m p o r t a n t as i t relates

to Duhem's treatment of hupothesis and its cognates i n the passage from
Simplicius is also highly suspect. The Platonic challenge asks for hypothe­
ses, b u t Duhem's use of'hypotheses' i n the subsequent discussion, and i n
other works makes i t clear t h a t he does not recognise any significant
distinction between Plato's use of hupothesis, and his own use of hypothe­
ses, and so he can make a case t h a t Plato was an instrumentalist, at least
as far as astronomy is concerned. However, Plato's use of hupothesis, both
87

i n the line allegory and i n the Phaedo is surely significantly different from
Duhem's. For Plato, hypotheses are something we seek to go beyond as
88

p a r t of a dynamic enquiry, something we perhaps even r e t u r n to destroy

as hypotheses. They are certainly not the end point of an enquiry as they
are for Duhem, b u t somewhere near the beginning. So too standard Greek
usage, while not as loaded as Plato's use, would also have hypotheses are
something assumed for further discussion. So w h e n we come to the Pla­
tonic challenge, we have no reason to suppose t h a t we should rest satisfied
w i t h our hypotheses of regular, circular and ordered motion, and feel t h a t
we ought to go no further. One m i g h t add t h a t i n the Phaedo the a i m is to
reach ti hiknanon, something adequate, and i n the Republic i t is to
eventually to come to the arche anupothetos, the unhypothesised princi­
ple. 89

According to Duhem, the only constraint on hypotheses should be t h a t

they should save the phenomena. The intellectual freedom of the astro­
nomer should not be restricted by having to make his hypotheses conform
to some conception of the 'celestial essence'. Is this the case w i t h Plato?
Again, surely not. I n the passage from Simplicius Plato demands t h a t
hypotheses should posit regular, circular, and ordered motion. I n view of
w h a t Plato says at length i n the Timaeus, Laws and Epinomis, i t should
come as no great surprise t h a t Plato l i m i t s hypotheses, and l i m i t s t h e m i n
this manner. These are some quite considerable constraints on the sort of
hypotheses Plato w i l l consider. Let us be clear w h a t those are to start w i t h .
D u h e m translates the passage from Simplicius as:

Quels sont les mouvements circulaires et parfaitment reguliers qu'il con-

vient de prendre pour hypotheses, afin que Ton puisse sauver les apparances
presentees par les astres errants? 90

T h a t I suggest significantly underplays the constraints. While Simplicius

in De Caelo 488.23 has Plato asking for motions which are homalon kai
tetagmenon, 492.3 has h i m asking for motions w h i c h are homalon kai
enkuklion kai tetagmenon. Presumably tetagmenon asks for something
more t h a n homalon, rather t h a n merely reiterates i t , and the idea of taxis
is an i m p o r t a n t one for Plato, especially i n the Timaeus and Laws. Here
96 Plato's Philosophy of Science
Laws 898ab is important. F i r s t i t speaks of regular and uniform motion,
and t h e n of motion which is according to hena logon kai taxin mian ampho,
one reason and order. Laws 898b contrasts motion which is not regular or
u n i f o r m and also motion w h i c h is med'en kosmoi med en taxei mede en tini
logoi, 'not arranged or ordered or according to some reason'. O n the
evidence of the Laws then, i t would seem then t h a t Plato asks for regular,
circular and ordered motion. A t 991e, the Epinomis is even more emphatic
on this matter. Every aspect of the analysis of the heavens (diagrams,
numbers, harmonies) must be brought into one system.
The critical question here is, Are the restrictions which Plato places on
hypotheses merely means to make calculations of the motions of the
heavens simpler? The answer to this has to be no. W h y do the heavenly
bodies move i n this manner for Plato? Because i n addition to t h e i r bodies
w h i c h are ' m a i n l y fire', they have souls and are intelligent, and so choose
91

the appropriate best available m o t i o n . As Laws 893b ff. makes abun­

92

dantly clear, h i g h on the list is regular and circular motion. Plato restricts
hypotheses to combinations of this sort of motion because of his view on
the nature of the heavens. So too Plato wants ordered motions, w h i c h I
take t h a t to mean t h a t we must consider i n d i v i d u a l motions i n relation to
the ordered whole of the heavens. As I w i l l argue i n greater detail i n a
Chapter 5, the model of the heavens which Plato gives us i n the Timaeus
is too simple to account for the k n o w n phenomena, because of the insis­
tence on regular, circular and ordered motion. I n the following passage
D u h e m is highly critical of Aristotle, but everything he says here is equally
applicable to the conditions for celestial motion w h i c h Plato sets:

There were so many restrictive conditions that he imposed upon the hy­
potheses of the astronomers, and he would not have hesitated to reject a
combination of motions that presumed to dispense with any of them. 93

I t is also i m p o r t a n t t h a t Plato's idea of techne goes beyond i n s t r u m e n t a l i s t

limitations. Gorgias 500e tells us t h a t the techne of medicine investigates
the nature of those i t treats and the reason for w h a t i t does, and Phaedrus
270b tells is t h a t i t is necessary to analyse the nature of the body (for
medicine) and the soul (for rhetoric), i f you i n t e n d a t t a i n the techne and
not merely have practice and experience (me tribe monon kai empeiria).
T h a t the Timaeus presents a m y t h and a likely story is a problem for
the i n t e r p r e t a t i o n of t h a t work. Quite how we ought treat such an account
I shall r e t u r n to later. Even i f we take the likely account at face value
though, t h a t w i l l not help the instrumentalist view, on which a l l accounts
w h i c h save the phenomena have equal m e r i t , and we adopt the simplest.
The likely account of the Timaeus is the most likely account, more likely
t h a n any other. I t is also i m p o r t a n t to recognise t h a t much of w h a t the
Timaeus has to say on astronomy and cosmology is repeated elsewhere,
 3. Meno's Paradox and Underdetermination 97
most notably i n the Laws, and indeed the Epinomis, w i t h o u t the l i k e l y
account qualifiers.
The most fundamental matter here, from w h i c h m u c h of the above
flows, is quite simple. This is t h a t Plato believes he has a solution to the
problem of underdetermination for the formation and selection of hypothe­
ses, and is not i n the least concerned about underdetermination i n relation
to negative instances. N o r do I see t h a t Plato would be worried by other
anti-realist arguments. Van Fraassen's argument t h a t the merits of expla­
nations are context r e l a t i v e , such t h a t explanatory m e r i t cannot decide
94

between theories. I t is open to Plato to argue t h a t relation to the (form of

the) good and the structure generated around i t as i n the Republic is the
critical context. The historical argument t h a t so many past theories have
been superseded t h a t we should not t r u s t current ones fails to trouble
m a n y modern realists and was unavailable anyway i n the context of
fourth-century Greek science. A r t h u r Fine has argued t h a t any proce­
95

dure for choosing between hypotheses cannot itself be justified by t h a t

procedure, on p a i n of vicious circularity, and must be justified by some­
t h i n g more stringent. Plato has at least two options here. The first would
96

be to argue t h a t his decision-making procedure is derived from something

more fundamental, possibly considerations about how to conduct any
investigation at a l l , and here Plato would be less worried about j u s t i f y i n g
scientific procedure by extra-scientific means t h a n Fine m i g h t be. Sec­ 97

ondly, given some of the epistemological considerations examined earlier,

Plato m i g h t well argue t h a t any circularity here is virtuous rather t h a n
vicious w h e n we consider the broader p i c t u r e . 98

Thus i t is h a r d to see t h a t Plato has any motivation towards i n s t r u m e n -

talism. I see no reason to suppose t h a t Plato was an instrumentalist, either
i n general or i n astronomy, or t h a t he began a t r a d i t i o n of i n s t r u m e n t a l i s m
i n astronomy.

I X . S a v i n g the p h e n o m e n a
Some doubts have been raised about the nature and authenticity of the
challenge t h a t Simplicius ascribes to Plato. Let us look at the full passage.
Simplicius tells us that:

First of the Greeks [to hypothesise regular, ordered, circular motion] was
Eudoxus of Knidos, (1) as Eudemus states in the second book of his History
of astronomy, and so too Sosigenes having taken this from Eudemus, (2) and
as Sosigenes says, Plato had posed the following problem for those engaged
in such studies: Which hypotheses of regular and ordered motion are able to
save the phenomena of the planets?' 99

Some commentators w o r r y t h a t apparently we have the testimony of

Eudemus and Sosigenes for (1), we only have the testimony of Sosigenes
for ( 2 ) . I t seems extremely doubtful whether we can set great store by
100
98 Plato's Philosophy of Science
Simplicius' 'characteristic pedantry' here. A t 488.23 the motion asked for
101

is homalon kai tetagmenon, b u t w h e n he gives a recapitulation of the

passage above at 492.31 ff., at 493.3 the motion asked for is homalon kai
engkuklion kai tetagmenon. 102
As Vlastos comments, i t may simply be the
case t h a t Sosigenes failed to acknowledge Eudemus, or t h a t for stylistic
reasons (not w i s h i n g to repeat himself) Simplicius failed to acknowledge
such an acknowledgement. Even i f (2) was not i n Eudemus, Simplicius
103

repeats (2) at 492.31 ff. w i t h o u t mentioning either Eudemus or Sosigenes,

and raises no doubts about a t t r i b u t i n g the challenge given i n (2) to Plato.
W h e t h e r Simplicius h a d access to the works of b o t h Eudemus and
Sosigenes is a question here, though almost certainly i t is undecidable on
the current evidence. T h a t he quotes from Eudemus' History of Geometry
(at l e n g t h ) and his Physics (very often) does not show he h a d access to
104 105

the H i s t o r y of Astronomy, and t h a t he quotes extensively from the first t w o

b u t very l i t t l e from the t h i r d may be good evidence t h a t he d i d not. H i s
praise of Eudemus' style may also be second h a n d . N o r do I see t h a t any
106

great mileage is to be had from w o r r y i n g how Eudoxus was t r u l y the first

w i t h these hypotheses i f Plato h a d set such a problem. T h a t Plato set a
problem and Eudoxus was the first to attempt a solution is a perfectly
sensible reading here. There is some evidence t h a t Plato was not the
107

first to hypothesise regular circular motion, as there is a passage from

Geminus w h i c h tells us that:

For the Pythagoreans, who were the first to apply themselves to investiga­
tions of this kind, assumed the movements of the Sun, the Moon and the five
planets to be circular and uniform. They would not admit, with reference to
things divine and eternal, any disorder such as would make them move at
one time more swiftly, at one time more slowly, and at another time stand
still ... with the imperishable stars i t is not possible to adduce any cause of
swiftness or slowness. 108

I f we accept this and Simplicius, I would agree w i t h Van der Waerden t h a t

the easiest resolution here is to suppose t h a t Eudoxus was first to address
the problem as stated by P l a t o . Depending on w h a t sort of relationship
109

the Pythagoreans envisaged between this model and the actual heavens,
Plato may have been the first to believe t h a t the phenomena may be saved
i n t h i s manner. 110

This is not to say t h a t there are not legitimate worries about whether
Plato actually issued the challenge reported i n Simplicius. As Z h m u d has
argued, there was a tendency to portray Plato as the architect of science
w h i c h can be traced back to the early Academy. I t is odd t h a t i f the
111

challenge was i n Eudemus t h a t we do not hear about from other sources,

and the challenge is not mentioned by either Theon or Geminus. I t is
possible t h e n t h a t we have only have Sosigenes' testimony, and t h a t he
may be embellishing other similar accounts. As Z h m u d has argued, there
 3. Meno's Paradox and Underdetermination 99
are similarities between Philodemus' History of the Academy and the
passage i n S i m p l i c i u s :
112

At this time the mathematical sciences (ta mathemata) were also greatly
advanced, with Plato being the architect of this development; he set prob­
lems for the mathematicians, who in turn eagerly studied them.

Whether t h a t shows a great deal is another matter. There is n o t h i n g i n the

challenge which could not have been said by Plato, as problems and
hypotheses can be found i n the Republic, regular, circular and ordered
motion i n the Timaeus and Laws. Sosigenes could quite possibly have
lifted a l l the required language from Plato himself, independent of any
other t r a d i t i o n . I t is unlikely t h a t we can come to a definitive answer on
whether Plato actually issued the challenge or not. Unless we can u n e a r t h
some new authoritative text there w i l l always be some doubts and even i f
Z h m u d is correct, t h a t does not preclude the possibility t h a t Plato issued
the challenge and t h a t Sosigenes' report is i n fact true.
Some more tractable questions are whether Plato could have issued
such a challenge, and whether, given the tenor of his later view on
astronomy and cosmology, he was likely to issue such a challenge. So too
we can consider the relation of Plato and Eudoxus independently of the
challenge. There are some standard objections to Plato issuing the chal­
lenge and subsequently influencing Eudoxus. One we have already dealt
w i t h , t h a t on the evidence of Republic V I I Plato rejected or denigrated
empirical astronomy and so could not have influenced someone such as
Eudoxus. Neugebauer for instance comments t h a t Plato could not have
positively influenced Eudoxus:

His advice to the astronomers to replace observations by speculation would

have destroyed one of the most important contributions of the Greeks to the
exact sciences. 113

Two other objections need detailed r e b u t t a l and w i l l be dealt w i t h later i n

this book. I f Plato d i d not himself advocate regular, circular and ordered
motion how could he issue the challenge and so influence Eudoxus? I t is
clear t h a t i n the Republic he did not advocate such celestial motion. We
can look here either to 530b, where we are told t h a t we cannot expect the
periods of the planets not to deviate (paralattein) i n any way, or to the m y t h
of E r and the i n t e r m i t t e n t actions of the daughters of necessity. Arguably,
t h a t is the case i n the Phaedo and the Politicus as well. However, I shall
argue at length i n later chapters t h a t this is not the case i n the Timaeus,
and t h a t the Philebus, Laws and Epinomis agree w i t h the Timaeus. The
later Plato may well have issued this challenge and i n doing so influenced
E u d o x u s . 1 have argued t h a t Plato as not an instrumentalist, and i n the
114

next few chapters we w i l l see how he struggles w i t h the foundations of

scientific realism. T h a t m i g h t be seen as a barrier to any serious influence
100 Plato's Philosophy of Science
on Eudoxus, i f Eudoxus was, as some have asserted, an i n s t r u m e n t a l i s t ,
interested only i n a geometric construction of the planets, paths and
disinterested i n cosmology. A g a i n though, I shall argue t h a t a t t r i b u t i n g
such a position to Eudoxus is quite u n w a r r a n t e d on the evidence we have,
which, i f anything, points i n the other direction. One cannot t h e n argue
against the authenticity of Simplicius' report on the grounds t h a t i t w o u l d
have been impossible or highly u n l i k e l y for Plato to have issued such a
challenge or for Eudoxus to have accepted i t . N o r is there a n y t h i n g i n
Simplicius' report itself t h a t m i g h t lead us to have serious doubts over i t .
 4

Celestial Motion i n the Timaeus

The m a i n a i m of this chapter is to argue t h a t i n the Timaeus the motion of

the heavenly bodies is regular, stable and amenable to precise m a t h e m a t i ­
cal prediction, contrary to Plato's view as expressed i n the Republic. This
is i m p o r t a n t b o t h as a m a t t e r of Platonic scholarship, where i t has been
controversial, and as an episode i n the history of science. Plato's belief i n
1

the regular, circular and ordered motion of the heavenly bodies arguably
sets the stage for astronomy for the next two millennia, so the develop­
ment of his views is of some interest. I t is also significant t h a t we can find
this conception of celestial motion i n the Laws and Epinomis, and argu­
ably the Philebus as well. I f so, t h e n later Plato is perhaps i m m u n e to some
of the criticisms t h a t m i g h t be aimed at the Republic. One of the key
implications here is t h a t there is a considerable strengthening of the
connections between events i f Plato gives up the idea of the inherent
i r r e g u l a r behaviour of physical entities. The criticism t h a t Plato cannot
have exceptionless laws because of this inherent i r r e g u l a r behaviour is
t h e n inapplicable for the later works. I also propose an i n t e r p r e t a t i o n of
the relation between reason and necessity w h i c h accords w i t h regular
celestial motion and the new ideas about the behaviour of bodies. The
change to regular and stable celestial motion perhaps also gives us a new
criterion for the relative dating of Plato's works. I shall argue t h a t the last
expression of the Republic view comes i n the Politicus, and t h a t those
works w h i c h share the Timaeus' view, the Philebus, Laws and Epinomis
m i g h t a l l be dated later ( w i t h due caveats on dating). I also argue t h a t i t
is possible to discern a change i n Plato's attitude to the i n e v i t a b i l i t y of
political, social and m o r a l degeneration, and t h a t this may be l i n k e d to the
cosmological changes. I argue t h a t m y group of later dialogues disown the
inevitable degeneration of the earlier works.

I . D i s o r d e r l y a n d d e g e n e r a t i n g cosmologies
We have seen t h a t according to the Republic, the periods of the heavenly
bodies deviate and do not conform to regular periods. The Phaedo provides
a framework w h i c h would support s i m i l a r conclusions. There the m y t h
draws a parallel, w h i c h we m i g h t t e r m the 'fish's eye' view of the w o r l d .2

Just as a fish m i g h t look up and see the sun and stars t h r o u g h the water,
102 Plato's Philosophy of Science
and believe the sea to be the true heaven, only to be disabused i f i t were to
put its head above water, so we, l i v i n g i n the hollows of the earth, live i n a
m u r k i e r w o r l d t h a n the t r u e heaven. Exactly w h a t lesson we are to draw
about the nature of the heavens and of astronomy is unclear, t h o u g h i t may
w e l l be t h a t we see an analogue of the b l u r r e d and unstable image of the
heavens as a fish would see of the w o r l d i f i t were to look up t h r o u g h
t u r b u l e n t waters. I f so, t h e n the visible heavens m i g h t well defy precise
3

mathematical prediction and the periods of the planets would deviate.

The m y t h of the Politicus is also interesting i n this respect. We are
presented w i t h a cosmos which, on the face of i t , is subject to two cycles.
F i r s t , there is the reign of Cronus where God guides the r o t a t i o n of the
heavens. When, at the appointed time, he relinquishes his control, the
motion of the heavens reverses and the reign of Zeus ensues. I t is made
clear at Politicus 272b t h a t we are currently l i v i n g under the reign of Zeus:

You have heard of the life under Cronus, Socrates, but the tale of life under
Zeus, that of the current state, you are alive to experience yourself. 4

The key passage is Politicus 273b ff., where the Stranger describes the fate
of the w o r l d i n the reign of Zeus:

In the beginning i t remembered the teaching of its creator and father most
accurately, though this eventually dulled. The reason for this was the physi­
cal element in its constitution, which had been in i t from the earliest stage,
and partook of great disorder before the universe came to be in its present
ordered condition ... When the world nurtures within itself living things
under the guidance of the helmsman, it produces little evil and much good.
However, when i t becomes separated from him, i t fares best during the time
immediately after the release, but as time proceeds and it grows forgetful,
the old condition of disorder gains sway more and more, and towards the
conclusion of time little good and much of its opposite flourishes, and there
is danger of the destruction of the world and those in it.

T h a t t h i s change affects the heavenly bodies is attested at Politicus 269a.

We m i g h t also infer this from the Plato's general conviction t h a t the stars
have physical bodies and the myth's statement t h a t a l l physical bodies
degenerate. I f the cosmos is continually degenerating towards chaos, and
as Politicus 273b makes clear, this w i l l be a radical and fundamental chaos
(unless god intervenes, the cosmos w i l l be 'storm driven by confusion and
broken up into an endless sea of unlikeness', Politicus 273de), t h e n clearly
the periods of the planets w i l l be subject to deviation and the visible
heavens w i l l again defy precise mathematical prediction.
One m i g h t also argue t h a t i n the Phaedo, Republic and Politicus Plato
is committed to a degenerating cosmology. The clearest statement of this
occurs i n the m y t h of the Politicus at 273b ff., b u t so too the passage at
Republic 529d ff. can be read as saying t h a t not only do the planets not
follow mathematically precise orbits, b u t t h a t t h e i r orbits are also continu-
 4. Celestial Motion in the Timaeus 103
ally decaying. I have translated paralattein at 530b2 as 'to deviate' b u t i t
is w o r t h noting its resonances as given by L S J o f to change or alter a l i t t l e ,
esp. for the worse ... to pass aside, t u r n from the path, to deviate, vary'.
Interestingly the key w o r d paralattein does not occur i n Timaeus' account
of the heavens, nor i n other works (Philebus, Laws, Epinomis) which
arguably agree w i t h the Timaeus? One m i g h t see such a degeneration as
an expression of the general principle enunciated at Republic 546a t h a t
' A l l things t h a t come to be must also decay', where i t is used to u n d e r p i n
a description of inevitable social and political degeneration. I t may not be
an accident t h a t Plato mentions Hesiod's theory of metals at Republic 546e
ff. w h e n discussing his own theory of decay. One can find similar language
6

of corruption and decay i n the m y t h of the Phaedo, where at 110a Socrates

tells us:

This earth and the stones and the whole region have been corrupted and
eaten away (diephtharmena ...kai katabebromena), as things in the sea are
by salt water; nor does anything worth mentioning grow in the sea, and
practically nothing is perfect, but there are eroded rocks and sand and
unimaginable mud and mire, and things are in no way worthy to be com­
pared with the beauties in our world.

One m i g h t argue here t h a t as Plato uses the perfect tense one is only
justified i n assuming t h a t decay has occurred rather t h a n is continuously
occurring, though one m i g h t counter by saying t h a t is a l l he requires for
his epistemological purposes and the comparison w i t h sea water m i g h t
suggest a continuous process.

I I . T h e m y t h of the Politicus
How common is the idea of a degenerating cosmos i n the ancient world?
W i t h l i t e r a t u r e one could certainly look to Homer and Hesiod. W i t h
Presocratic philosophers Empedocles is explicit t h a t we are i n a cycle of
increasing strife and decreasing love, and certainly at least some and
possibly a l l the cosmoi of the atomists were subject to destruction. W i t h
later philosophy, the Stoics held a theory of gradual w a r m i n g and loss of
e q u i l i b r i u m moving towards the cosmic conflagration, and the theme of
decay is more evident i n Epicurus and Lucretius. Lucretius tells of a
farmer who praises the achievements of his forefathers, but:

He does not realise that everything is gradually decaying and on course for
shipwreck, worn out by the long years of old age. 7

The idea of a degenerating cosmos, or of a degenerating cycle w i t h a

cyclical conception of t i m e was t h e n reasonably common i n the ancient
world. We cannot dismiss the idea t h a t Plato at some stage held a degen­
erating cosmology on the grounds t h a t such a view would have been
104 Plato's Philosophy of Science
u n d u l y pessimistic i n the context of other ancient cosmologies, nor i f I am
correct i n m y suggestions about the Phaedo and the Republic, is the
Politicus alone i n Plato's works p u t t i n g forward a degenerating cosmology.
I have assumed a standard two-phase cycle for the Politicus myth,
where Brisson and Rowe have argued for three phases. 1 do not believe
8

t h a t such a move is justified on broader philosophical considerations.

Contra Rowe, I don't see t h a t the Politicus cosmology is hopelessly pessi­
mistic, compared w i t h other works by Plato, other philosophers or Greek
literature. Does the cosmos behave like a rational creature? I n m a n y ways,
yes i t does. I t struggles h a r d to be as much like god as possible, b u t is
eventually overcome by old age and deteriorates towards death. Rational
creatures are m o r t a l , even i f gods are not, and i t is interesting to note t h a t
the cosmos of the Timaeus is a god as well as a rational creature and does
not degenerate. The Politicus specifically denies t h a t the cosmos is a god.
This consideration is also i m p o r t a n t i n relation to one of the m a i n motiva­
tions considering the details of the m y t h behind the Brisson/Rowe view. A t
272e ff., god relinquishes control of the cosmos, and so too the demi-gods
let go t h e i r appointed spheres of influence. A t this point there is a shudder,
and great destruction of l i v i n g things. There is t h e n a period w h e n the
cosmos sets itself i n order and masters itself and the things i n i t . I t is only
somewhat later t h a t the deterioration sets i n . A problem w i t h a straight­
forward reading of the m y t h , of god's order followed by degeneration, is the
chaos left after god leaves and the regeneration before the degeneration.
Degeneration ought to follow order, and not chaos, and the regeneration is
unexplained. However, i f the cosmos is a rational b u t m o r t a l creature left
to its own devices, this is indeed w h a t we ought to expect and there is no
need to postulate extra phases of the cycle. While the cosmos is young, i t
w i l l be capable of such ordering. This is not to say t h a t the m o r t a l cosmos
can order itself from any sort of chaos, and certainly not from the 'endless
sea of unlikeness' t h a t god has to rescue i t from the possibility of, nor even
from the state the cosmos is i n w h e n god performed this rescue act. The
chaos after god relinquishes control is far milder t h a n either of these.
There is a shudder, and the death of many l i v i n g things, b u t the cosmos is
s t i l l holding together fairly well. Like any l i v i n g creature, i t cannot order
itself from o u t r i g h t chaos, b u t can to some extent perform repairs on itself
and heal its o w n wounds. Also like any l i v i n g t h i n g though, i t loses this
ability as i t grows older.
So the second phase begins i n some sort of chaos w h i c h the young
cosmos can p u t into order. As such ordering is clearly for the best, there
w i l l be no problem i n h a v i n g an account of i t . As the cosmos gets older
though, i t begins to lose its faculties, and so is no longer able to sustain the
order w h i c h i t has created, and the cosmos begins to degenerate. As the
cosmos begins to age there is no need for a reversal of the cosmos, w h i c h
m i g h t w e l l be thought to b r i n g great destruction w i t h i t , as w h e n god
relinquished control. I t is interesting to compare the krates language of
 4. Celestial Motion in the Timaeus 105
Politicus 2 7 3 b l w i t h the Timaeus. I n the Timaeus, i f we master our
emotions we can become j u s t , i f we master our perceptions we can become
knowledgeable and indeed like god (as far as possible!). I f we look to the
account of ageing, i n y o u t h our triangles can cut up the incoming ones, i n
old age the reverse u n t i l we eventually die. I n the second phase t h e n the
cosmos does not actively oppose the rule of god, b u t imitates i t as best i t
can. I n the early stages i t does t h a t well, b u t being mortal, the cosmos
cannot do this w e l l for ever. I f we can keep the interpretation of the m y t h
simple w h i l e coherent, we ought to do so given the simplicity criteria t h a t
are mentioned at the beginning of the m y t h . 9

I agree w i t h McCabe t h a t i t is i m p o r t a n t t h a t our account of the current

phase is such t h a t there can be a debate about the merits of life i n each
cycle, and t h a t i t is plausible t h a t life i n the current cycle is better. So too
our account of the current cycle must generate some grounds for reconsid­
ering the nature of the statesman. O n the accounts of Brisson and Rowe,
10

life i n the first cycle would seem preferable, and there seem to be no
grounds for the statesman not being a shepherd, as i n the first cycle. The
theme of independence (autokrates) is an i m p o r t a n t l i n k between the m y t h
and the m a i n body of the work. The cosmos, like the statesman does best
when independent and capable of ordering the disparate parts of the
cosmos/state into a unified whole. The statesman, however, like the cos­
mos, is m o r t a l and so cannot do this indefinitely. Here I would agree w i t h
Lane t h a t the Politicus, 11
i n distinction to the Republic, lays emphasis on
the ability of the statesman to act at the kairos, the r i g h t moment. I f t h a t
ability is lost, as both cosmos and any particular statesman w i l l indeed
lose i t , t h e n we see the unfortunate results. The two-phase cycle t h e n
seems to make best sense of the Politicus m y t h , as long as we recognise
the cosmos is a m o r t a l rational a n i m a l which can regenerate itself and be
independent i n its y o u t h b u t degenerates i n its old age.

I I I . C e l e s t i a l m o t i o n i n the Timaeus
One i n t e r p r e t a t i o n of the Timaeus has i t i n agreement w i t h the Republic
on the question of regular celestial motion. B u r k e r t comments t h a t i n the
Republic, Plato:

was still of the opinion that real exactitude was impossible in the physical
world, so that the true astronomer should not depend on sense perception
but busy himself with purely ideal magnitudes and movements. I n the
Timaeus, too, Plato speaks without hesitation of the 'wandering' of the
planets. 12

Certainly there are things to be said i n favour of this sort of position. The
planets are often referred to by the use of planes, 'a wanderer', and its
cognates. We m i g h t w i s h to equate this usage of planes w i t h t h a t at
13
106 Plato's Philosophy of Science
Timaeus 48a9, where we are t o l d of tes planomenos eidos aitias, the
'wandering cause' w h i c h prevents the complete expression of reason i n the
universe. So while reason m i g h t w i s h the heavenly bodies to move i n an
orderly manner, the dictates of the compromise w i t h necessity to w h i c h a l l
physical entities are subject ensures t h a t they move irregularly. Thus
planes attributes at least some disorderly motion to the heavens. However,
there are considerable difficulties w i t h this view, and I shall argue t h a t
celestial motion i n the Timaeus is entirely regular, and t h a t the planes
description is used i n two separate ways. I n the context of astronomy i t
merely distinguishes the fixed stars from those bodies w h i c h move i n
complex orbits and so 'wander' across the sky, while i n other contexts i t
carries the a t t r i b u t i o n of disorderliness. F i r s t consider the use of planes i n
the context of astronomy. A t 34a Timaeus tells us t h a t the demiurge gave
to the universe:

a motion proper to its body, that of the seven motions which is best suited to
reason and intelligence. Therefore he made i t move in a circle, revolving of
itself uniformly and in the same place, and he took from i t all trace of the
other six motions and kept it free from their wanderings (kai aplanes
apeirgaseto ekeinon). 14

There is t h e n at least one motion w h i c h is aplanes. Further, at 40b

Timaeus distinguishes between 'the stars which do not wander' (hos aplane
ton astron) (40b4), and the planets 'which t u r n and as such wander' (ta de
trepomena kai planen toiauten ischonta) (40b6). This is rather a w k w a r d
for the position I have sketched, according to w h i c h there should be no
aplanes celestial motion. One m i g h t t r y to argue t h a t the crucial difference
between the fixed stars and the planets is t h a t the former are l i v i n g
creatures and the latter are not, but Timaeus 38e makes i t clear t h a t the
planets are alive too. The fixed stars of course have physical bodies,
'created for the most part from fire' (Timaeus 40a3). W h a t though of the
bodies w h i c h do have planes motion? Concerning the orbits of the planets,
at 39c Timaeus tells us that:

Men do not name them nor investigate the mathematical relations (sum-
metrountai skopountes arithmois) between them, so that they do not know
that the wanderings of these stars constitute time (chronon onta tas touton
planas), as they possess immense complexity and form astonishing patterns.

Now, i f the wanderings of the planets constitute t i m e , and these wander­

ings are irregular, t h e n t i m e w i l l be irregular. Thus planes cannot a t t r i b u t e
i r r e g u l a r i t y of motion w h e n applied to the heavenly bodies. More likely, i t
merely distinguishes the mobile hands of the celestial clock (the planets)
from the static numerals (the fixed stars). I t is i l l u m i n a t i n g to compare
this passage w i t h Republic 530a. Both t a l k of how the motions of the
heavenly bodies relate to time, but here the emphasis on the deviations of
 4. Celestial Motion in the Timaeus 107
physical bodies and the consequent i r r e g u l a r i t y of celestial motion has
been lost. Instead, i t seems t h a t most men, including perhaps the younger
Plato, have been led astray by the immense complexity of the heavens and
have not properly investigated the mathematical relationships between
the stars. Note for instance the difference between the summetria, 'propor­
tions' at Republic 530a8 which deviate, and the precision of the sum-
metrountai arithmois ' c o m m e n s u r a t e s i n n u m b e r ' t h a t we are to
investigate at Timaeus 39c8. A t 529cd the Republic emphasises t h a t i t is
the t r u e realities, those accessible to reason and thought w h i c h have 'true
number', while the visible heavens are like a diagram. I n the Timaeus, the
emphasis is on the fact t h a t we can apply mathematics precisely to the
visible heavens.
I f Plato d i d not at any stage countenance an i r r e g u l a r flow of time, t h e n
his conception of w h a t constitutes t i m e may also have changed. I n order
to judge t h a t celestial t i m e is irregular we need some independent access
to absolute time. How this is done is left unspecified i n the Republic, b u t
i t is clear t h a t absolute t i m e there cannot be constituted out of the motions
of the planets. I t may well be t h a t Plato had not considered these difficul­
ties at the t i m e of w r i t i n g the Republic and had no clear ideas concerning
the nature of absolute t i m e or how we m i g h t gain access to i t . One m i g h t
compare the clarity of the Timaeus on these points, as at 47a we are told
t h a t the sight of day and n i g h t etc. has procured for us number and the
concept of t i m e .15

One m i g h t t r y to grasp the nettle here though, and argue t h a t perhaps

for Plato, w i t h his view of the imperfection of the physical world, t i m e does
not flow i n a perfectly regular manner. There is no explicit statement on
this m a t t e r i n the m a i n corpus of his works, though we are told at
Epinomis 991c that:

We must grasp the exactness of time, how precisely it marks all the events
throughout the heavens.

There are however some points which weigh heavily against the irregular­
i t y of t i m e i n the Timaeus. A t 37d Timaeus tells us t h a t the demiurge:

Intended to create a movable image of eternity, and at the same time as

ordering the heavens, he made from the eternity that resides in unity an
eternal image moving according to number (kat' arithmon iousin), that
which we have called time.

So at Timaeus 38c we find that:

From this reasoning and thought by God concerning the generation of time,
in order that time should come to be, the sun and the moon and the five other
stars, which have the name 'wanderers' (planeta), came into being to deter-
108 Plato's Philosophy of Science
mine and guard the numbers of time (eis diorismon kai phulaken arithmon
chronou gegone).

Now, although t i m e may be less perfect t h a n t h a t w h i c h i t is an image of,

this does not entail t h a t its flow is irregular. The contrast here between
stability i n u n i t y and motion according to number is sufficient. A l l t h a t is
required is t h a t t i m e flows while eternity stands still, and the phrase
'moving according to number' would suggest t h a t t i m e flows i n an orderly
manner. This is reiterated at Timaeus 38a9 where we are told t h a t t i m e
'circles around according to number'. I t is also w o r t h n o t i n g t h a t Aristotle,
while he is critical of Plato's views on time, does not comment on this
matter, w h i c h one m i g h t expect i f Plato had advocated non-uniform t i m e . 16

A further consideration is the stability of the solar system and the nature
of the great year. A t 39d Timaeus tells us:

Nevertheless, it is perfectly possible to grasp the amount of time elapsed at

the completion of the perfect year, when the relative speeds of all the eight
circuits have brought them to a head simultaneously, as measured by the
motion of circle of the same and similar. I n this manner and for these reasons
all of those stars which have turnings on their passage through the heavens
were created, in order that this world might be as like as possible to the
perfect and intelligent living being in respect of i t imitating its eternal
nature.

I f there is a specific amount of t i m e between the grand conjunctions, t h e n

celestial motion must be regular, or we are left w i t h the h i g h l y improbable
alternative t h a t the irregular motions somehow cancel each other out. I n
t h a t case, the great year would lose its significance as a sign of the r a t i o n a l
ordering of the universe. I f the great year recurs, and there is no sugges­
t i o n i n the Timaeus t h a t i t does not, t h e n celestial motion must be regular
and the solar system stable and free from any degeneration. The predict­
able recurrence of the great year is a sign of the cosmological stability of
the Timaeus. Certainly one cannot, as A d a m does, equate the great year of
the Timaeus w i t h the periods of God's help and God's abstinence. The 17

great year could be achieved i n the reign of Cronus, (there is, however, no
mention of the grand conjunction, and i t is not the criterion by which t h a t
age comes to an end, cf. Politicus 269c and 272d), b u t could only be
achieved by chance, i f at a l l , i n the reign of Zeus w h e n the planets move
increasingly chaotically, and contrary to Timaeus 39c ff., the t i m e between
conjunctions would be unpredictable.
One m i g h t also note the way i n w h i c h i n the Timaeus we can make
precise calculations about the visible heavenly bodies i n distinction to the
position of the Republic 530a ff. The general idea t h a t the visible heavens
are amenable to calculation proliferates throughout the Timaeus, 18
as does
the idea t h a t events occur i n the physical w o r l d kat' arithmon, 'according
 4. Celestial Motion in the Timaeus 109
to number'. 19
Finally, we need to consider Timaeus 47a ff. where we are
told:

God devised and gave to us vision in order that we might observe the rational
revolutions of the heavens and use them against the revolutions of thought
that are in us, which are like them, though those are clear and ours confused,
and by learning thoroughly and partaking in calculations correct according
to nature (logismon kata phusin orthotetos), by imitation of the entirely
unwandering (pantos aplaneis) revolutions of God we might stabilise the
wandering (peplanemenas) revolutions in ourselves.

We can compare this w i t h Timaeus 40b, where planes and aplanes are used
of the planets and the fixed stars respectively. E i t h e r Plato has made an
oversight, or planes is being used i n two different senses. Outside of its
specific use i n astronomy of distinguishing planets from fixed stars, planes
again takes on its sense of disorderliness, but here we are t o l d t h a t the
motions of the heavens are not disorderly. I t is i m p o r t a n t to recognise
20

t h a t 47a ff. is not an isolated comment. As I shall discuss i n more detail i n

Chapter 9, the analogy between the w o r l d soul, the visible manifestation
of w h i c h is the heavens and the h u m a n soul is a central theme of the
Timaeus and one t h a t is worked out w i t h some care. Timaeus 37a ff. insists
t h a t a l l the judgements of the world soul are true and t h a t the circles of
the same and the different move regularly.

IV. C e l e s t i a l m o t i o n i n l a t e r P l a t o
One objection m i g h t ask how is i t t h a t as the physical w o r l d is supposed
to be an imperfect copy, there can be perfectly regular celestial motion.
Here we need to consider the difference for Plato between ideal and
uniform motion. I f we were to describe the usual N e w t o n i a n scenario to
Plato, of a body moving i n a straight line at a constant velocity unless acted
upon by external forces, would he consider such motion to be perfect?
Surely not. Laws 893b ff. develops a set of criteria for the best type of
motion. Some of these criteria are indeed to do w i t h regularity, b u t others
assert the primacy of self-generating circular motion i n one place. I t is
clear from this is t h a t we can remove some of the qualifications for
perfectly r a t i o n a l motion w i t h o u t losing regularity or uniformity, as they
are necessary but not sufficient conditions.
This appears to be the course t h a t the Timaeus takes. The motion of 21

the whole universe is one of uniform rotation entirely free from the other
six motions. Consider how the motion of the fixed stars relate to this. A t
22

40a Timaeus tells us t h a t while they have translational motion due to

obeying the motion of the same:

With regard to the other five motions, they are motionless and still, in order
that each might attain the greatest possible perfection.
110 Plato's Philosophy of Science
One can show the regularity of this translational motion i n two ways.
F i r s t , i t is equated w i t h the motion of the same (the motion of the same is
equivalent of the motion of the universe as a whole, and from Timaeus 34a
t h a t must be a regular motion), and secondly, we are told t h a t i t consists
of only one of the six motions to the exclusion of the other five. Such 23

motion is less perfect t h a n t h a t of the universe as a whole i n the hierarchy

of motions, b u t this is achieved by complexity rather t h a n i r r e g u l a r i t y . So
24

too w i t h the planets. T h e i r motion may be less perfect t h a n t h a t of the fixed

stars, b u t this is because t h e i r paths are more complex, being subject to
the motion of the different as well as t h a t of the same. One m i g h t compare
here t i m e , w h i c h is a moving image of eternity, yet moves i n a regular
manner.
The m a t t e r of the highly controversial 'contrary power' (enantian ...
dunamin) at Timaeus 38d4, w h i c h (arguably) applies to M e r c u r y and
Venus and (arguably) deals w i t h w h y these planets are always seen near
the sun w i l l be discussed at length i n the next chapter. There are some
complex matters of interpretive strategy which come out of this issue, the
result of w h i c h i n m y view is t h a t regular circular motion emerges u n ­
scathed.
The Timaeus does not, u n l i k e the Laws, specifically condemn the de­
scription of the planets as planeta, because of the ambiguous use of this
t e r m . One point of interest concerning the condemnation i n the Laws is
t h a t i t reads like a strongly worded clarification of a matter on w h i c h Plato
may have felt t h a t he was misunderstood. A t Laws 821a the A t h e n i a n
Stranger says to Cleinias:

My good fellow, we speak falsely nowadays when all the Greeks tell of those
great deities, the sun and the moon ... We assert that they never follow the
same path, along with some other stars, those we have named 'wanderers'.

A t Laws 822a the A t h e n i a n states categorically: 25

The usual opinion concerning the sun, moon, and other planets, that they at
some time wander (planatai), is not the case; precisely the opposite is true.
For each of these bodies always travel on one path, and not many, although
this may not seem so.

This is entirely i n accord w i t h Laws 898ab, where the motion t h a t the

A t h e n i a n Stranger w i l l argue is given to the cosmos by soul is:

regular and uniform and in the same place and around the same things and
in relation to the same things and according to one reason and one order.

One m i g h t compare the attitude of the Laws on accuracy and calculation

w i t h the s i m i l a r comments i n the Timaeus and the different attitude t a k e n
by the Republic. A t Laws 967a the A t h e n i a n Stranger tells us: 26
 4. Celestial Motion in the Timaeus 111
Even at that time a sense of wonderment emerged concerning the heavenly
bodies, and those who studied them accurately suspected what is now indeed
believed, that i f they were soulless and lacking in intelligence, they would
not be amenable to wonderful calculations (logismois) of such accuracy.

So too the Philebus assumes the regularity of celestial motion. T h a t the

universe has a soul, and t h a t the years, seasons and months are arranged
by intelligence are emphasised at Philebus 30a and 30c respectively. One
m i g h t argue t h a t Philebus 30c (and indeed Epinomis 983c) are deliber­
ately evocative of Republic 530ab and set out to correct the earlier position.
Philebus 30d tells us t h a t a k i n g l y soul is behind this intelligence, and t h a t
there are others w h i c h carry out similar functions. T h a t is very m u c h i n
line w i t h the Timaeus position, where the demiurge creates the order of
the world, and t h e n has lesser demons, whose embodiment are the heav­
enly bodies, m a r k out t i m e . This is i m p o r t a n t because we are t o l d t h a t
27

p a i n and pleasure are to be equated w i t h deterioration and restoration, 28

and, i n line w i t h the general tenor of the Laws, t h a t i t w o u l d be quite

u n f i t t i n g for the Gods to experience pleasure or pain. A t Philebus 3 I d i t is
emphasised t h a t p a i n occurs when the harmony of the l i v i n g being is
disrupted. One m i g h t t h i n k here of the way i n w h i c h both the Philebus and
the Timaeus describe the universe as a l i v i n g being, and the way t h a t the
Timaeus takes great care to emphasis the harmony t h a t constitutes its
body, especially at Timaeus 35a-37d. I f so t h e n the heavenly bodies cannot
suffer any deterioration and hence, like the Timaeus b u t u n l i k e the
Republic, t h e i r motions ought to be entirely regular. One m i g h t also
29

compare the general tenor of the remarks at Philebus 28d-33a w i t h those

of the Laws 967a ff. There is the same emphasis on the order of the
heavens and the supposition t h a t the stars have souls, and the same
denigration of previous thinkers who thought otherwise. There is also no
question t h a t the Epinomis advocates regular celestial m o t i o n . 30

V. T h e P o l i t i c u s m y t h a n d the Timaeus
One m i g h t also consider some more specific points of difference between
the Timaeus and the Politicus. The Timaeus is adamant t h a t the cosmos,
once created, can function i n a stable manner w i t h o u t any attention from
its creator and can only be dissolved by an act of his w i l l . I n the Politicus
3 1

t h o u g h either God must perpetually guide the cosmos or i t degenerates of

its o w n inherent nature, and is only saved from s i n k i n g into 'an endless
sea of unlikeness' by the active intervention of God. While the Politicus
32

tells us t h a t the cosmos is mounted on a pivot, the Timaeus is adamant

t h a t i t is n o t . WThile i n the Timaeus the cosmos is now ageron kai anosos,
33

'ageless and unailing', i n the Politicus God has to make the w o r l d athana-
ton kai ageron, 'undying and ageless' i n recreating the reign of Cronus
after the current cycle. As we have seen, the cosmos of the Politicus is a
34
112 Plato's Philosophy of Science
rational a n i m a l , but i t is not a god and so not i m m o r t a l , where the cosmos
of the Timaeus is a god. A further consideration here is the nature of time.
I f the Politicus is operating w i t h the conception of t i m e found i n the
Timaeus t h e n t i m e is going to become increasingly chaotic along w i t h the
motions of the planets. As we are l i v i n g i n the degenerating cycle i t is more
l i k e l y t h a t Plato is s t i l l operating w i t h the Republic's conception of t i m e .
I t w o u l d seem t h e n t h a t on cosmological grounds we can separate two
groups of works, the Timaeus, Philebus, Laws and Epinomis as advocating
regular and stable celestial motion, the Phaedo, Republic and Politicus as
not. I f we assume t h a t Plato made a single move on this question, we m i g h t
attempt to employ this as one criterion for the order of his works. W h i l e
this w o u l d give us t w o nice groups, I would reiterate here m y reservations
about producing absolute orders for Plato's works. There are a couple of
objections r e l a t i n g to the Politicus m y t h t h a t we m i g h t reject though. One
m i g h t be concerned t h a t certainly i n the Phaedo and the Politicus, and
w i t h the m y t h of E r we are dealing w i t h m a t e r i a l t h a t is specifically
introduced as m y t h . So too one m i g h t argue the m a i n t h r u s t of the
Politicus m y t h is political and not cosmological, and we do not, for
35

instance, take every detail of the geography of the Phaedo m y t h or the

cosmology of the m y t h of E r entirely seriously or ascribe these views
directly to Plato. However, I do not require t h a t we accept a l l the details of
the Politicus m y t h (or those of the Phaedo or Republic), merely t h a t we
take some of the u n d e r l y i n g principles w h i c h are used to construct i t as
being indicative of Plato's current views, as we would do w i t h Plato's other
m y t h s . The Politicus m y t h certainly shares some assumptions concern­
36

i n g cosmology and t i m e w i t h the Republic and the Phaedo. I t is highly

significant t h a t when Plato re-uses the m y t h of Cronus to make some
political points at Laws 713b ff., i t is stripped of its cosmology, and i n its
new form fits w i t h the regular and stable celestial motion position of the
Laws. T h a t we m i g h t take i n conjunction w i t h Laws 822a where Plato
m i g h t be thought to clarify fully his position on the motions of the heavens,
along w i t h other trenchant passages i n the Laws, as an indication t h a t
Plato is m a k i n g i t u t t e r l y clear w h a t his new position on cosmology i s . 37

The Laws draws the sort of political morals one m i g h t expect from the
m y t h , w h i c h are i n line w i t h the Eleatic Stranger's comments at Politicus
275b. I f the cosmology of the Politicus m y t h is not essential to its political
interpretation, t h e n w h y do we get this cosmology, especially i f i t is
something t h a t Plato does not hold with? I t is also i m p o r t a n t to note the
manner i n w h i c h the m y t h is introduced. A t Politicus 268de the Stranger
says t h a t the m y t h he is about to introduce 'may offer some amusement
(paidian)' and asks young Socrates to:

Give your full attention to this story, just as children (paides) do; for you have
not entirely outgrown such games (paidias).
 4. Celestial Motion in the Timaeus 113
We are perhaps not meant to take the m y t h literally, b u t there are
i m p o r t a n t points to be gleaned from looking at the structure of w h a t is
said, and that, even i f the m a i n t h r u s t of the m y t h is political, m u s t include
the considerable amount of cosmology we are given i n the m y t h as w e l l . 38

As McCabe comments, i t is not clear w h y such a detailed cosmology is

39

required to make the political points w h i c h some out of the m y t h , so we

m i g h t suppose the m y t h to have come further purpose concerning cosmo­
logy or divine teleology.
One final concern m i g h t be whether any astronomer could accept the
degenerating cosmology of the Politicus m y t h . T h a t m i g h t depend on how
m u c h deviation is already supposed to have occurred and how quickly the
rot is supposed to set i n . Ancient Greek astronomy s t i l l h a d to resolve
m a n y anomalies concerning the motions of the heavens, w h i c h could be
equated w i t h the deterioration t h a t had already occurred. The deteriora­
t i o n m i g h t also be very slow. Here i t is i m p o r t a n t to emphasise t h a t w h e n
at Politicus 270a8 we are given a comment on the length of the degenerat­
ing period (pollas periodon muriados), this is not 'thousands of revolutions'
but something rather stronger, l i t e r a l l y 'many ten thousands of revolu­
tions', and its most likely meaning is 'countlessly many'. I f we make the
40

reasonable assumption t h a t one rotation of the universe is one day (that

is the l i k e l y implication of Politicus 269a coupled w i t h 269bc f f . ) , t h e n one
41

thousand revolutions w i l l be approximately three years, and t e n thousand

approximately t h i r t y (or around 'one generation of men'). Politicus 269b,
272de, and 274de each make i t clear t h a t we are dealing w i t h a timescale
t h a t encompasses many generations. Whatever the outcome of any dating
considerations here, the Phaedo, Republic and Politicus share the position
t h a t the motions of the visible heavens w i l l be irregular, unstable and not
amenable to exact mathematical analysis, and t h a t is contrary to the view
of the Timaeus, Philebus, Laws and Epinomis.

V I . Reason a n d necessity
One objection here to the notion t h a t the heavens move i n a regular
manner i n the Timaeus is the question of the relation of reason and
necessity. A t Timaeus 48a we are told:

The generation of this universe was the bringing together of a combination

of necessity and intelligence. Intelligence ruled necessity by persuading her
to lead the greatest part of the things that were generated towards the best.
It was in this way, through necessity submitting to sensible persuasion, 42

that this universe was in the beginning brought together. I f this was the
manner of its generation, then really speaking one must bring in the way in
which the wandering aitia acts.

W h a t does Plato mean by necessity though, and how is i t t h a t necessity

can be 'persuaded' by anything? Plato speaks of necessity accomplishing
114 Plato's Philosophy of Science
things of chance and disorder. As reason only scores a p a r t i a l victory over
43

necessity, there is t h e n some residual chance and disorder. N o w i t seems 44

strange t h a t necessity should be associated w i t h these things. However,

one m i g h t take both chance and disorder i n two separate senses, depend­
i n g on w h a t they are contrasted w i t h . A n event m i g h t be said to occur by
chance because there is no causal chain t h a t leads to its occurrence,
contrasting chance w i t h causal d e t e r m i n i s m . However, one m i g h t also
45

contrast chance w i t h h u m a n design. I f we were to b l i n d l y t h r o w p a i n t at a

canvas, i t w o u l d be mere chance i f a n y t h i n g good were the result, t h o u g h
no causal chain need be broken i f such an event were to occur. S i m i l a r l y 46

perhaps, i f the necessity t h a t Timaeus speaks of produces a n y t h i n g of

w o r t h , this is chance i n the absence of any teleological design rather t h a n
i n the sense of any breach of causal connections. 47

We can take the same sort of approach w i t h disorder. The order t h a t i t

could be contrasted w i t h m i g h t be t h a t of physical law, or t h a t of a
teleological arrangement of the phenomena. The ordinary emission of
l i g h t , for instance, m i g h t be lawlike b u t disorderly (no order to wavelength
or direction), relative to the stimulated emission of l i g h t and its ordering
into a laser beam (ordered i n wavelength and direction). A n ancient 48

analogue here m i g h t be t h a t the principles of m i l i t a r y strategy apply to a l l

groups of men, b u t some groups are well ordered formations while others
are disordered rabbles. This is i m p o r t a n t as Plato often uses taxis and its
cognates, w i t h its connotations of d r a w i n g up into a good m i l i t a r y order i n
these contexts, a good example being Timaeus 30a: 49

For god desired that all things should be good, and nothing paltry (phlauron)
as far as was possible, he took over all that was visible, which was not at rest
but in discordant and disorderly motion (alia kinoumenon plemmelos kai
ataktos), and led i t into order out of disorder, judging the former to be entirely
better than the latter.

I f we t h e n take i t t h a t the p a r t i a l victory of reason over necessity is the

p a r t i a l i n s t a n t i a t i o n of teleology, t h e n there is not a problem w i t h perfectly
regular behaviour by physical bodies. I t may be t h a t the attempt to
instantiate completely the good produces a set of conflicting demands
w h i c h cannot be j o i n t l y met. A good example of this m i g h t be the question
of the h u m a n skull, discussed at Timaeus 75bc. I n order for us to have
acute perception, the s k u l l ought to be as t h i n as possible. I n order for i t
to protect our brains and ensure a long life, i t should be as t h i c k as
possible. S i m i l a r l y at 75e there must be some flesh around the s k u l l for
the purpose of temperature control, b u t as l i t t l e as possible so as not to
obstruct perception. I f we t h i n k of necessity i n this manner, t h e n reason
can only persuade i t as far as logical possibility w i l l allow.
Cornford questions how regular behaviour can be persuaded i n any
manner. The i m p o r t a n t t h i n g to recognise here is t h a t the demiurge
 4. Celestial Motion in the Timaeus 115
performs a once and for a l l action w h e n he orders the w o r l d from its
p r i m o r d i a l chaos. There is t h e n no question of a ghostly h a n d i n t e r f e r i n g
w i t h causal sequences after the i n i t i a l ordering. The persuasion of neces­
50

sity lies i n this i n i t i a l ordering. I f the demiurge does not bother to order
the m a t t e r of the universe, t h e n a n y t h i n g good t h a t occurs i n the sub­
sequent development of the cosmos w i l l be purely due to chance. As we 51

have seen, the best t h a t could be hoped for i n the pre-cosmic state is t h a t
the like-to-like principle w i l l have brought together clumps of like entities,
most probably the four elements, and t h a t w i l l not constitute a w e l l
ordered cosmos. However, i f matter is arranged such t h a t good w i l l result
from the law-like development of the i n i t i a l conditions, necessity is per­
haps 'persuaded' by reason to produce something good. So the earth, w i t h
its m i x t u r e of earth, water, air and fire is created where necessity would
have produced merely a clump of earth entirely surrounded by water.
Laws 889b is the critical cosmogonical passage here, as we saw i n Chapter
1§I. We cannot expect a like-to-like principle to b r i n g together u n l i k e
things except by the merest chance. So too the elements exist due to techne.
I f the demiurge not only arranges the positions of the m a t t e r i n the
universe b u t also determines the shape of the fundamental particles, t h e n
further constraints enter the picture. I f the shape of these particles m u s t
be the best possible, as the Timaeus insists, and the shape of the particles
contributes to the properties t h a t they possess, t h e n the demiurge is
saddled w i t h some more constraints on w h a t a reasoned arrangement of
the particles may be able to achieve. As Cornford points out, i t is
52

i m p o r t a n t t h a t we do not make the demiurge omnipotent. O n m y ac­ 53

count, the demiurge is l i m i t e d by causal necessity, w h i c h he can persuade

only by a l t e r i n g the i n i t i a l conditions of the system, and logical possibility
w h i c h prevents h i m from i n s t a n t i a t i n g conflicting demands of the good.

V I I . I m p l i c a t i o n s for c a u s a t i o n
A further consequence of this analysis of the Timaeus is a strengthening
of the connections between events. Sorabji has argued t h a t for Plato
physical bodies are incapable of perfectly regular behaviour, and so there
can be no question of causal necessities. I n certain works, such as the
Republic and the Politicus, this would indeed appear to be Plato's posi­
t i o n . Sorabji further argues t h a t Plato distinguished physical from logi­
54

cal necessity more sharply t h a n a modern philosopher w o u l d as he

associated necessity w i t h the 'wandering cause' i n the Timaeus, and
Sorabji associates this wandering w i t h passages from the Republic and the
Politicus w h i c h i m p l y t h a t bodies behave irregularly:

When Plato speaks of necessity here, he cannot be associating i t with perfect

invariability. For he states emphatically in the Republic (530b) and the
116 Plato's Philosophy of Science
Politicus (269de) that what is bodily cannot always behave in the same way
without alteration. 55

Now, i f the arguments of the previous sections have been correct, we can
say t w o i m p o r t a n t things about regular behaviour i n the Timaeus. F i r s t ,
contra the Republic position, the heavens do indeed move i n an entirely
regular manner. I f we ask w h y t h a t is so, t h e n we must look to the fact t h a t
the heavens are no longer mounted on any sort of pivot and the heavenly
bodies are now i m m o r t a l , intelligent self-movers. Being intelligent they
w i l l act for the best and being i m m o r t a l they w i l l continue to do so, thus
giving a solid basis for the ongoing order of the heavens. The gods, as the
Laws and Epinomis emphasise, do not change t h e i r minds. I shall r e t u r n
to consider merits of such an approach i n Chapter 6. A l l I w i s h to establish
here is t h a t the heavens move i n an entirely regular manner i n the
Timaeus.
The second critical difference w i t h the Republic and Politicus can be p u t
like this. I f we associate necessity i n the Timaeus w i t h an invariable
like-to-like principle, t h e n t h a t principle may, i n a regular, necessary and
law-like manner, produce chance and disorder, where we contrast chance
and disorder w i t h a teleological ordering rather t h a n regular, law-like and
invariable behaviour. Thus necessity wanders from teleology b u t not from
regularity. So u n l i k e the Republic and Politicus, m a t t e r is capable of
regular behaviour, and indeed persistent regular behaviour even i f i t is
incapable of intelligent behaviour.
Clearly there are some i m p o r t a n t consequences for Plato's philosophy
of science. I n Chapter 2 i n relation to the Republic on the 'deviations' of the
heavens, we saw Sorabji r e m a r k i n g t h a t 'This is contrary to the spirit of
those who believe t h a t nature obeys exceptionless laws'. T h a t is incorrect
56

for the Timaeus both for the heavens, where the fact t h a t the stars have
souls means t h a t they w i l l behave intelligently, and for matter i n general.
I n Chapter 7 I w i l l further support this claim about matter by arguing t h a t
the basic entities supposed by the Timaeus, the two half-equilateral and
the half-square triangles, can be seen as h a v i n g stable properties.
Plato has been accused of reducing physical necessity to logical neces­
sity i n the Phaedo. Whatever the case i n the Phaedo, this is not the case
51

for the Timaeus. I n relation to one of the key cases for the Phaedo, the
relation of fire and heat, i t is interesting t h a t at 61d the Timaeus says:

Let us consider why we say that fire is hot. We may investigate this by
noticing the dividing and cutting effect that i t has on our bodies. That it has
the property of sharpness is something we nearly all perceive. As for the
thinness of its sides and the sharpness of its angles and the smallness of its
particles and the speed of its motion, by virtue of all of which it is energetic
and divisive, always cutting keenly whatever it encounters, these must be
accounted for by recalling the origin of its shape, and i t is rather this and no
 4. Celestial Motion in the Timaeus 117
other property that divides up our bodies into small pieces and which
naturally gives i t the quality we call heat and supplies its name.

So is the r e l a t i o n between fire and heat here one of logical necessity? The
Timaeus postulates different sizes of fire particles. Consider the possibility
t h a t our bodies are made of smaller fire particles t h a n the ones we meet.
The fire particles we meet w i l l no longer have this effect on our bodies, and
so cannot be considered to be hot. I f t h a t seems a somewhat artificial
example (though the stars have bodies of fire, and the effect of fire on t h e m
m i g h t be interesting), consider the following. The Timaeus gives an analy­
sis of colour. Colour is due to interaction of the outgoing fire particles of
58

our visual stream and incoming fire particles. W h e n the incoming particles
are larger t h e n those of our visual stream, the stream is contracted and we
see black. W h e n they are smaller, the stream is dilated and we see w h i t e .
So is there any relation of logical necessity here? Surely not, as there is no
logical necessity t h a t our visual stream is comprised of particles of a
certain size and no logical necessity t h a t contraction of the visual stream
makes us see black.
One m i g h t argue t h a t these examples are too strong, as i t appears t h a t
causal as well as logical necessities are breached i f fire is not always hot.
That, i t would seem, is achieved by m a k i n g qualities relative to the
perceiver. As w i l l be i m p o r t a n t later i n this book, there are two senses to
this. There is the sense of the section of the Theaetetus w h i c h deals w i t h
the idea t h a t knowledge is perception, whereby a l l our perceptual judge­
ments are t a k e n to be correct. There is also though the Timaeus sense,
whereby qualities may be relative to us, b u t there is a physical basis both
for the qualities and our perception of them, and we may be i n error about
our perceptions. So ( i n n o r m a l circumstances, and perceiving correctly) we
feel hot w h e n our bodies are being dissolved by fire particles. I f our m e n t a l
revolutions are disturbed though we may judge ourselves not to feel hot
even w h e n our bodies are being dissolved by fire. A less radical example
may be t h a t I may judge (erroneously) t h a t this d r i n k is sweet w h e n i n fact
i t is bitter, i n the sense t h a t i t is carrying out the physical processes on m y
tongue associated w i t h bitterness. I take this to be a different case t h a n
59

to the following. I f I am i l l and m y tongue is affected (say i t is swollen) and

a d r i n k has a different effect on m y tongue to usual (what w o u l d have the
effect of contraction on a n o r m a l tongue and so w o u l d taste bitter, returns
my swollen tongue to n o r m a l and so tastes sweet to me). I n this case the
effect is different b u t I judge correctly, i n the other case the effect is the
same b u t I judge incorrectly.
We need to go a stage deeper for the Timaeus' causal necessities, w h i c h
hold between the entities of geometrical atomism. There is no breach of
causal necessities here. The fire particles always behave i n the same
manner even i f they would not be hot to one of the celestial deities. Indeed
the Timaeus gives us several statements on how the elements interact, and
118 Plato's Philosophy of Science
on how large and small amounts of elements i n t e r a c t . These w o u l d seem
60

quite straightforwardly to be causal b u t not logical necessities. So too we

m i g h t consider the like-to-like principle to be a causal rather t h a n a logical
necessity. One has to be slightly cautious here, as the like-to-like principle
may describe an effect of the way i n w h i c h particles i n motion interact,
rather t h a n be a causal principle i n itself. The interactions of the p a r t i ­
61

cles would s t i l l constitute causal necessities, however.

I t w o u l d appear t h a t not only is Plato interested i n exploring the
possibilities of a cosmology w h i c h postulates r e g u l a r behaviour, b u t is
also concerned w i t h explaining w h y there should be such behaviour. A t
Timaeus 61d above and i n the Timaeus generally, Plato relies on the
geometrical properties of matter to explain regular behaviour, and more
strongly, the (causally) necessary connection between events. I t is impor­
t a n t t h a t for the most basic entities those properties do not change, as w i l l
be argued i n Chapter 7, and while water may decompose and t u r n into fire,
the properties of the icosahedra of water and the tetrahedra of fire do not
change either.
I w i l l postpone discussion of the question of the nature and basis of the
intelligent behaviour of the heavens u n t i l we have seen a l i t t l e more of
Plato's astronomy and cosmology. Here too Plato seeks for an explanation
of regularities, b u t this t i m e has too explain t h e i r apparently i n t e l l i g e n t
behaviour as well, and so attributes souls to the heavenly bodies. I f those
souls always act i n the best manner they w i l l act i n an entirely regular
manner, and here Plato invokes a different basis for understanding n a t u ­
ral laws.

V I I I . Political, social a n d moral decline

I n addition to these changes concerning cosmology and t i m e , i t may also
be possible to discern a change i n Plato's attitude to the i n e v i t a b i l i t y of
political, social and m o r a l (PSM) degeneration. A t 273c i n the Politicus
m y t h , the Eleatic Stranger tells us:

From the one who brought i t together, the universe has received only good,
but from its previous state i t retains in itself and engenders in the living
creatures all the harshness and injustice that has its origin in the heavens.
When the world nurtures within itself living things under the guidance of
the helmsman, i t produces little evil and much good. However, when it
becomes separated from him, it fares best during the time immediately after
the release, but as time proceeds and it grows forgetful, the old condition of
disorder gains sway more and more, and towards the conclusion of time little
good and much of its opposite flourishes.

T h a t seems to be as clear a statement as possible of an inevitable P S M

decline. We m i g h t compare this w i t h Republic 546a ff. where Plato has
Socrates describe the inevitable descent from the ideal state t h r o u g h
 4. Celestial Motion in the Timaeus 119
timocracy, oligarchy and democracy to tyranny. One problem w i t h this
description is t h a t i t is difficult to judge Plato's attitude here as he deploys
the paidia I spoude (play/seriousness) contrast. At Republic 545d Plato has
Socrates say:

Shall we, like Homer, pray to the Muses to tell us how a faction first began
the trouble, and believe that in their tragic style they speak as if in earnest
(hos de spoude) employing high flown language, though they tease and play
with us as i f we were children (hos pros paidas hemas paizousin)?

One reading of this piece of'serious play' m i g h t be t h a t we should not take

the details of each society and the passage from one to another too literally,
but should be concerned w i t h the general character types of each society
and the tendency for society to degenerate. 62

We should perhaps distinguish between a deterministic decline, where

each stage is of necessity worse t h a n the previous, and one where w h i l e
there are factors w h i c h unchecked w i l l cause a society to decline, conscious
h u m a n action may, at least temporarily, counteract them. Those factors,
as i n the Politicus, may become stronger w i t h time. Taylor has quite
reasonably likened this passage to a medical textbook description of a
t e r m i n a l illness. T h a t describes the typical stages an unchecked ailment
63

w i l l progress through, while i n an actual case there may be many compli­

cating factors. Here conscious action by a doctor may slow the decline,
temporarily check i t or even temporarily improve the patient's condition,
w i t h o u t removing the u n d e r l y i n g cause of his or her u l t i m a t e decline and
demise. Plato I suggest is committed to this l a t t e r possibility, such t h a t
64

w h e n we look at other works we w i l l have to look for an absence of

corrupting factors rather t h a n mere evidence t h a t society can make some
temporary progress. Let us now consider the position of the Laws. A t 769d
ff., the Stranger tells us t h a t legislation ought to be w r i t t e n down and duly
amended i f found w a n t i n g i n practice, Tf the ordering of the polis is always
to improve and to i n no way decline.' The implication here would seem to
be t h a t not only can there be progress, b u t t h a t the fate of the city lies i n
its own hands, rather t h a n being subject to any degenerating forces. One
of the key political questions asked by the Laws is outlined by the Stranger
at 676b ff., where he says:

Sometimes larger states have come to be from smaller ones, and sometimes
smaller from larger, and worse states have come to be from better and better
from worse ... We must attempt to understand, i f i t lies within our power,
the reason for such change.

The rest of Laws Book I I I is t h e n devoted to answering this question.

Various factors are analysed here; at 688c ignorance and vice are dis­
cussed, at 691a, financial acquisitiveness,
65
at 697c a lack of communal
spirit, at 704d geographical factors. Yet 66
at no stage i n Book I I I , or
120 Plato's Philosophy of Science
anywhere else i n the Laws, is any explanation i n terms of an inevitable
degeneration cited or alluded to. Indeed, the point of this investigation
seems to be to argue t h a t a l l of the causes of political degeneration are
either avoidable or treatable.
After this discussion the Stranger introduces a cyclical theory of history
where after periods of increasing social sophistication m a n k i n d is reduced
to a p r i m i t i v e and r u r a l existence by severe flooding. There is no sense of
an inevitable decline here though, or of a decline a l t e r n a t i n g w i t h a more
positive phase. I t is highly significant t h a t i n this context t h a t w h e n the
m y t h of Cronus makes a reappearance at 713a f f , i t is shorn of any
commitment to inevitable degeneration. I n the Laws then, we do not find
any inevitable P S M degeneration of the k i n d we find i n the m y t h of the
Politicus. H o w m u c h of a change t h a t is from the Republic depends on how
seriously we take the passage at 545d ff. Even i f the Republic is not
committed to an inevitable decline, there is an i m p o r t a n t change of tone
i n the Laws w h i c h cannot be fully accounted for by reference to the
different subject matter. 67

As for the Timaeus, a l l the passages w h i c h concern politics are to be

found i n the opening m y t h , and as I have argued, these have various
l i t e r a r y and philosophical functions to serve which mean t h a t we should
not take t h e m to be indicative of Plato's views. The Timaeus does have
some interesting points to make about moral decline, however. O u r souls,
analogously to the w o r l d soul, are made up of two revolutions. O u r 68

intellectual welfare is related to how uniformly our intellectual revolutions

move, w h i c h is dependent on whether we take the trouble to master our
incoming perceptions, or are mastered by them. Similarly, the key m o r a l
injunction for humans concerns the relation of the emotions and the
intellect. A t 42b Timaeus tells us concerning m e n and t h e i r emotions that:

I f they have mastery over these, then they will live justly, but i f they are
mastered by them, unjustly. 69

W h i l e humans who do not take the trouble to master w h a t impinges on

t h e m may become corrupt, there is no sense of any i n e v i t a b i l i t y here, or of
any increasing disorder i n the w o r l d t h a t leads to the corruption of
mortals. I n contrast to the Politicus, where the w o r l d inevitably becomes
a more corrupting place, and where humans can do l i t t l e about this, the
Timaeus appears to provide a level playing field for mortals to determine
t h e i r own fate. One m i g h t usefully compare Timaeus 86d ad fin. w i t h the
Politicus i n the way t h a t i t analyses moral degeneration as something t h a t
is the sole responsibility of the i n d i v i d u a l soul, and i t is notable t h a t at
Timaeus 87b ff. t h a t treatable moral shortcomings are related to political
and social difficulties. Especially i m p o r t a n t here is Timaeus 90d ff., where
we are told t h a t the goal i n life is to rectify the disordered revolutions i n
our heads, because by doing so we w i l l a t t a i n the best sort of life. I can find
 4. Celestial Motion in the Timaeus 121
no indication t h a t the Philebus entertains the idea of any P S M degenera­
tion. I t is interesting t h a t the Politicus ties together the two concepts of
cosmological and P S M stability or degeneration, and t h a t i n the later
dialogues where we find cosmological stability we find P S M stability too.
I f the d r i v i n g force behind such degeneration is cosmological or metaphysi­
cal, changes here m i g h t explain the move to P S M stability. I t is interesting
to note t h a t according to Empedocles we are l i v i n g i n the increasing
strife/decreasing love p a r t of the cosmic cycle, and t h a t the Purifications
talks of the m o r a l degeneration w h i c h is brought about by this increase i n
strife.70

I X . T h e i m p l i c a t i o n s of a stable cosmology
W h a t I hope to have established i n this chapter is t h a t the Timaeus, like
the Philebus, Laws and Epinomis but u n l i k e the Phaedo, Republic and
Politicus advocates a stable cosmology and motion for the visible heavens
w h i c h is regular, circular and amenable to exact mathematical analysis.
A l o n g w i t h t h a t go some i m p o r t a n t cosmological changes. The cosmos is
no longer thought of as t u r n i n g on a pivot, contra the Republic and the
Politicus, and the cosmos now has an i m m o r t a l soul, against being soulless
i n the m y t h of E r and h a v i n g a m o r t a l soul i n the second phase of the
Politicus m y t h . There is also some evidence t h a t the later dialogues have
a different attitude to the inevitability of P S M decay, w h i c h may be l i n k e d
to the shift to a more stable cosmology. I do not t h i n k we have the means
to determine w h i c h of these changes is d r i v i n g the others, though I would
point out t h a t these appear to be co-ordinated changes.
Depending one's view of w h a t generated the difficulties i n the Republic,
there may also be an i m p o r t a n t metaphysical change, of the analogue for
how mathematics relates to the world. I n the Timaeus, the visible heavens
are amenable to precise mathematical analysis, where the Republic allows
this only for t h a t w h i c h is accessible to reason and thought and is not
visible to the eye. So too the basic entities of geometrical atomism, the
scalene and isosceles triangles, are precise i n t h e i r structure ( 1 , V3, ^2, and
1, 1, V2), and i t is critical for the way t h a t they fit together t h a t they are.
As we shall see w h e n we look at geometrical atomism i n more detail,
there is never any suggestion t h a t these are not perfect t r i a n g l e s . So71

instead of the geometrical objects and t h e i r instantiations analogue for the

relation of mathematics and geometry to the world, we have a new
analogue w h i c h allows some real precision. O n these bases we have a new
argument for the relative dating of Plato's later dialogues, w i t h the usual
caveats about such dating.
I f the motions of the heavenly bodies are regular and stable, t h e n there
are some i m p o r t a n t consequences both for Plato's relation to the develop­
ment of Greek science and for i n t e r p r e t i n g the Timaeus. This book w i l l now
pursue two general lines of enquiry. First, there is a great deal more to be
122 Plato's Philosophy of Science
said about the development of Plato's astronomy and how t h a t relates to
the development of Greek astronomy. We need to examine the notorious
'contrary power' of Timaeus 38d i n some detail, as some argue this runs
contrary to the idea of regular circular motion, and this debate raises some
i m p o r t a n t issues of interpretational strategy. We also need to look at the
development of Plato's cosmology i n more detail, r e l a t i n g i t to the develop­
ment of Greek cosmology, and examining the possible nomological reasons
Plato may have for teleology. These w i l l be the tasks of the next two
chapters.
A further point here relates to the realism/instrumentalism debate
about Plato. H e t h e r i n g t o n has argued t h a t as the visible w o r l d for Plato is
imperfect, citing Republic V I I , and t h a t 'reality exists not i n the visible
72

w o r l d b u t rather the w o r l d of ideas', Plato cannot be considered a realist

73

i n the orthodox sense. The fact t h a t Plato distinguishes forms and particu­
lars cannot be decisive against his being a scientific realist i n a reasonable
sense. M a n y modern scientific realists are realists about universals, and
would accept at least some of w h a t Plato m i g h t be thought to say about the
contrast between universals and particulars. As emphasised earlier, i t is
also i m p o r t a n t to treat Plato as a proto-realist rather t h a n to t r y to
assimilate h i m to modern realism. The i m p o r t a n t question for scientific
realists t h o u g h is not w h a t is most r e a l , b u t whether theories about the
74

physical w o r l d can be considered to be true or not. Even i f t h a t is problem­

atic i n the Republic because of a theory of the imperfection of the physical
w o r l d , or some inherent i r r e g u l a r behaviour (and t h a t would need some
further argument), i t is not a problem any longer i n the Timaeus.
The second general line of enquiry w i l l be to investigate the conse­
quences t h a t regular and stable celestial motion have for how we view the
Timaeus. One m i g h t object t h a t m y position on celestial motion and the
relation of reason and necessity attributes too m u c h i n the way of regular­
i t y and stability to the physical w o r l d of the Timaeus, and by allowing
reason so m u c h sway, renders too much of the physical w o r l d comprehen­
sible. I do not argue t h a t the view I have given i n this chapter is the only
possible i n t e r p r e t a t i o n of the relation of reason and necessity, merely t h a t
i t is one t h a t accords w i t h regular circular motion. W h a t I shall investigate
i n the final part of this book is whether we can align key passages i n the
Timaeus w i t h the idea of regular and stable circular motion, and produce
a viable overall reading. I believe this is possible, and the result is a
reading w h i c h is rather more epistemologically optimistic t h a n i n many
accounts.
There are various questions here t h a t w i l l need to be considered. W h a t
are the implications of there being exactitude i n at least one aspect of the
physical world? W h a t is the epistemological status of these stable heav­
ens? I t w o u l d seem t h a t i n the first place we can know t h a t the heavens
are stable and i n the second t h a t we are able to make some precise
calculations concerning the heavenly bodies. Perhaps, then, the astro-
75
 4. Celestial Motion in the Timaeus 123
nomy of the Timaeus shares some of the optimism of the Theaetetus and
the Philebus towards knowledge of the sensible. The whole account of the
w o r l d given by the Timaeus is presented as a m y t h or likely story. Is i t not
t h e n the case t h a t the theory t h a t the heavens move regularly is itself only
likely, such t h a t our calculations can only be as likely as t h a t theory?
Whatever Plato's reasons for t a k i n g this approach i n the Timaeus, one can
certainly point to places i n other dialogues where the r e g u l a r i t y of the
heavens is asserted w i t h o u t the l i k e l y story' qualifier, most notably Laws
822a ff., 893b ff. and 966e ff., and so too throughout the Epinomis. There
is also, though, the question of flux. There m i g h t be two problems here.
First, one m i g h t worry as the Theaetetus does about how things t h a t are
i n radical flux can be known. Secondly, one m i g h t w o r r y how i t is t h a t the
heavens can have regular motion and be stable i f the things t h a t constitute
t h e m are i n radical flux. I f an alternative reading can be established t h a t
makes the idea of regular and stable celestial motion sit w e l l i n relation to
these issues, we w i l l t h e n have to deal w i t h the question of w h a t Plato
intended i n the Timaeus. F i r s t though, let us look i n some more detail at
his astronomy and cosmology.
 5

Plato and the Development of

Greek Astronomy

I n this chapter I argue for two conclusions w h i c h do not, at first sight,

appear to sit well together. First, I argue t h a t although i t shows some
i m p o r t a n t advances on its predecessors, the astronomy of the Timaeus is
actually quite crude and poor and has generally been overrated by the
commentators. Secondly, I argue t h a t the astronomy and cosmology of the
Timaeus was of paramount importance to the development of these disci­
plines i n ancient Greece. I t marks the coming together of two previously
disparate traditions, those of observational astronomy and philosophical
cosmology, and the beginning of an enormously fruitful t r a d i t i o n i n astro­
nomy. I n w h a t way t h e n is the astronomy of the Timaeus poor? There are
h i g h l y evident discrepancies between w h a t the model of the Timaeus
predicts and phenomena well k n o w n i n fourth-century Greece. So m u c h so
t h a t one has to wonder about Plato's competence i f we take this to be a
l i t e r a l model of the heavens. Commentators have therefore sought to close
the gap between w h a t Plato was clearly aware of and w h a t the model can
account for, by arguing t h a t the astronomy is more sophisticated t h a n i t
first appears. Paradoxically, I believe this makes matters worse because
i n doing so the model becomes ad hoc and disorganised, losing the empha­
sis on regular, circular and ordered motion. The harder we try, worse i t
becomes. I argue t h a t we ought to abandon this approach altogether. I t is
m u c h better t h a t we allow the discrepancies to stand (there is evidence
t h a t Plato was aware of them) and treat the model, as I believe Plato
intended, as a prototype. I t is a first attempt at accounting for the motions
of the heavens i n terms of combinations of regular circular motions, and
as such i t indicates the basic principles one m i g h t employ i n cosmology and
i n solving the problems of the motions of the heavens. W h i l e the idea of
combinations of regular circular motion, especially i n conjunction w i t h an
offset ecliptic is a major advance i n astronomy, i t is by no means meant to
be the last word. One m i g h t add that, i f Plato thought t h a t the model of
the Timaeus d i d account for a l l the phenomena, t h e n there has to be a
question as to w h y he issued a challenge to ' a l l who are engaged on these
studies' to find the regular, circular and ordered motions t h a t account for
the phenomena. So while the model of the Timaeus may be a relatively
 5. Plato and the Development of Greek Astronomy 125
crude prototype, i t marks the coming together of observational astronomy
and cosmology and lays the basis for perhaps the most successful pro­
gramme of research undertaken by the ancients i n the science.

I . T h e Republic a n d the m y t h of E r
To see how Plato's astronomy advances, we need to compare the astronomy
of the Republic w i t h t h a t of the Timaeus. The Republic's m y t h of E r gives
us a picture of the heavens as a spindle (616c ff.), and provides the
following information: 1

Position Object Appearance Band width Speed

(vs. stars)
Outermost Fixed stars Spangled Broadest N/A
2nd Saturn Yellow 8th 5th
3rd Jupiter White 7th 4th
4th Mars Ruddy 3rd 3rd
5th Mercury Yellow 6th 2nd =
6th Venus 2nd whitest 2nd 2nd =
7th Sun Brightest 5th 2nd =
Innermost Moon Lit by 7 4th Fastest

The w i d t h of the r i m s of each of the bowls i n the spindle is most plausibly

interpreted as a reference to the relative distances of the planets, and the
speeds are the orbital speeds of the planets. M u c h of this is interesting and
useful. We have an order for the distances of the planets related to t h e i r
relative speeds. Some of the physiologoi (Anaxagoras, Hippocrates of Chios
and Democritus) considered the planets to be l e f t behind' (hupoleipesthai),
such t h a t the planet left behind most is the slowest. Here we have the
2

view, typical i n Plato, t h a t the planet l e f t behind' most is i n fact swiftest

i n its own motion. We have the moon l i t by the sun, and there is at least
3

a recognition of a relation between the orbits of Mercury, Venus and the

sun. 4

A t t e m p t s to associate the widths of the bands w i t h the amount of

deviation a planet has from the ecliptic, or possible sizes of epicycles fail,
and i t is far more l i k e l y t h a t this is some sort of representation of spacing
between the planets, though w h y these relations are chosen is not at a l l
clear. One m i g h t say t h e n t h a t the model of the Republic is not m u c h of
5

an advance on t h a t of Philolaus (see Figs 12 and 13). The counter-earth

6

has been abandoned, though so has the idea of a mobile earth. As I argue
i n the next chapter, however, the Greeks had good reasons for t h e i r
geocentrism. There is the basis for a decent theory of the phases of the
moon, as the moon is now l i t by the sun and is orbiting the earth (and not
the central fire). B u t the basic principle of the planets h a v i n g simple
circular motion and a l l l y i n g i n the plane of the earth's equator remains
the same. 8
126 Plato's Philosophy of Science

Fig. 12. Philolaus. All the heavenly

bodies revolve around the central fire,
which is hidden from the earth by the
counter-earth. The bodies are
(working outwards) the central fire,
the counter-earth, earth, moon, sun,
the five planets and the stars.

Fig. 13. Republic, myth of E r . All

7

the heavenly bodies revolve around

the earth, at various constant speeds
and distances, all motion again being
in one plane.

Does the Republic refer to the retrogression of the planets? I t is 9

possible t h a t i t does. The m a i n question is the meaning of epanakuk-

loumenon at Republic 617b2 i n the description of the motion of Mars. The
case for a reference to retrogression is t h a t at Timaeus 40c5 epanakukleseis
almost certainly does refer to retrograde m o t i o n . W h y though is M a r s
10

singled out w h e n a l l the planets undergo retrogradation? B u r n e t suggests

an emendation of the text, such t h a t Republic 617b2-3 would t h e n read
epanakukloumenon malista ton allon, 'retrogrades more t h a n the others',
w h i c h is reasonable i n astronomical terms as M a r s has a greater arc of
retrogression t h a n the other planets. B u r n e t supports this w i t h the fact
t h a t Theon of Smyrna has the addition w h e n he quotes this passage, and
that:

In the first place he is apparently quoting from Dercyllides, who first

established the text of Plato from which ours is derived. I n the second place
 5. Plato and the Development of Greek Astronomy 127
malista ton allon is exactly fifteen letters, the normal length of omissions
from Plato's text.11

Dicks argues t h a t Mars is singled out because i t 'makes more t h a n more

t h a n one circuit of the heavens between one loop and the next' and so he
12

takes epanakukloumenon to mean m a k i n g an additional circle. K n o r r

13

emphasises the phrase 'as i t appeared to t h e m ' and suggests t h a t this

suggest a specific t i m e at w h i c h the heavens were observed, w h e n M a r s
was undergoing retrograde m o t i o n . O n the very l i m i t e d account t h a t the
14

m y t h of E r gives us, i t is probably not possible to t e l l whether t h i s is a

reference to retrogression or not. Whether or not the Republic refers to
retrograde motion, i t cannot account for i t . The only possibility is the
actions of the daughters of necessity. The key passage is Republic 617c5-
dl:

Clotho with her right hand grasped and turned the outer revolution of the
spindle, leaving intervals of time between, while Atropos with her left
hand did similarly with the inner. Lachesis grasped each i n turn with
each hand.

So Clotho moves the fixed stars and so gives the general motion of the
heavens, while Atropos moves the sun, moon, planets slowly i n the oppo­
site direction. We m i g h t take i t as significant t h a t Clotho and Atropos
create motion i n opposite directions w i t h opposite hands, especially as
Plato refers to motion i n the heavens as to the r i g h t or left. So Lachesis,
15

i f we assume t h a t she can touch a l l the appropriate rims using either hand,
could either advance or r e t a r d each of them. There is a question t h o u g h as
to exactly w h a t Lachesis grasps. Even i f we assume t h a t she only grasps
the inner revolutions, so t h a t the stars do not have sudden backwards and
(extra) forwards movements, she is indiscriminate between the inner
revolutions, t h a t is between the planets, w h i c h do exhibit retrogression,
and the sun and moon, w h i c h do not. I would suggest t h a t the n a t u r a l
reading here is t h a t Lachesis alternately takes inner and outer rings i n
both hands. Something more sophisticated and d i s c r i m i n a t i n g w i l l be
required, especially as there is no suggestion whatsoever t h a t Lachesis
could generate the variations i n latitude observed d u r i n g retrogression.
The actions of the daughters of necessity look more oriented to keeping
moving a model w h i c h rotates on a pivot and so slows down gradually (as
a real spindle would), thus giving the deviations of Republic 530b t h a n
a n y t h i n g as complex astronomically as retrogression. The Timaeus t h e n
makes significant advances on the Republic. We not only have the ideas of
regular and stable circular motion, b u t also the ideas t h a t the motions of
the planets are composites of such motions and t h a t the plane of one of
these motions is offset from the earth's equator.
128 Plato's Philosophy of Science

I I . T h e Timaeus model
The Timaeus makes some i m p o r t a n t advances on the model of the Repub­
lic. I shall begin by r u n n i n g t h r o u g h some of the basic phenomena of
astronomy, to show some of the advantages of the Timaeus model over t h a t
of the Republic. I f one tracks the motion of the sun against the background
of the fixed stars d u r i n g a year, one gets a plot like t h a t shown i n F i g . 14.
This motion relative to the background of the fixed stars can be repre­
sented i n various ways. The modern definition of ecliptic is simply the p a t h
of the sun relative to the fixed stars. B y definition, the sun cannot deviate
from the ecliptic. For us the earth-sun relationship is special, as we believe
t h a t the earth orbits the sun, and t h a t the moon, b u t not the other planets,
orbits the earth. For the Greeks, though, a l l these bodies orbit the earth,
and so there is no reason to single out the sun for special treatment. So for
the Greeks, the zodiac is the band w i t h i n w h i c h a l l seven heavenly bodies
move. For some of the Greeks the sun moves i n a simple circle parallel to
and i n the middle of the zodiac, for others, notably Eudoxus and Callippus,

Fig. 14. Plot of the sun's motion. I n

order to make this sort of plot one
observes the point on the horizon at
which the sun sets (or rises) and then
observes which of the fixed stars rises
(or sets) at this point.

Fig. 15. The sun's motion. I n a sort

of pseudo-3D, we have the sphere of
the fixed stars and the motion of the
sun around the earth. The plane of
the sun's motion, at approximately
23 V2 to the earth's equator is known
0

as the plane of the ecliptic. The

6

standard ancient value was 24° (one

fifteenth of a circle). What value Plato
used, or whether he used a value, are
unknown.
 5. Plato and the Development of Greek Astronomy 129
i t deviates slightly from such a path. We can represent the motion of the
sun and the position of the zodiac as i n F i g . 16.
There is a further i m p o r t a n t phenomenon i n relation to the sun w h i c h
the ancient Greeks were well aware of. The sun does not set (or rise) on
the same point on the horizon each day. The point at w h i c h i t sets moves
between a n o r t h e r l y and a southerly m a x i m u m . So from w i n t e r solstice
this point moves n o r t h to a m a x i m u m of 23 V2 n o r t h of west (inclination of
0

the ecliptic). We t h e n have the longest day and shortest night, t h a t is

summer solstice (for the n o r t h e r n hemisphere). The point at w h i c h the sun
sets t h e n moves south. W h e n i t coincides w i t h due west, we have a day of
equal day and night, or equinox ( a u t u m n equinox). This point t h e n moves
to a southerly m a x i m u m , w h i c h produces the shortest day and longest
night, or w i n t e r solstice. I t t h e n moves n o r t h again t h r o u g h a second
(spring) equinox to the northerly m a x i m u m again. The points at w h i c h the

Fig. 16. Sun's motion and zodiac.

The Greeks had varying estimates of
the size of the zodiac though common
at this time was 12 . As will be
0 17

familiar from astrology, the zodiac

can be divided into twelve houses
and then significance is attached to
the positions of the sun, moon and
planets relative to those divisions.

Fig. 17 Simplified motion of the

sun. One can simplify Fig. 15 down to
two dimensions as here.
130 Plato's Philosophy of Science
sun reaches its m a x i m u m northerly or southerly position and t h e n begins
to head i n the opposite direction were k n o w n to the Greeks as tropai,
t u r n i n g points.
These are some of the basic phenomena of astronomy, w h i c h were
reasonably w e l l k n o w n to the ancient Greeks. Let us now look at Plato's
model i n the Timaeus. I t can be represented as shown i n Fig. 18. The circle
of the same rotates once every twenty-four hours, and generates the
motions of the fixed stars. The circle of the different is set at an angle to
the same, and is split into seven, for the sun, moon and five planets. The
different generates motion relative to the fixed stars along the ecliptic.

Fig. 18. Plato's two-sphere

model. Apseudo-3D
representation. Sun, moon and
planets move at varying speeds
and at varying distances from the
earth but all their motions not
due to the same are i n one plane.

Fig. 19. Simplified representation

of the Timaeus model. This
emphasises that all planetary motion
is in the plane of the different which
is equivalent to the ecliptic.
 5. Plato and the Development of Greek Astronomy 131
Relative to previous models of the heavens, the major advance here is the
idea t h a t the motions of the heavens can be resolved into combinations of
regular, circular motion offset at angles to each other. I f one chooses
appropriate orbital speeds, t h e n so far the model looks quite good, and
indeed i t can give a very good, though by no means perfect representation
of the movements of the sun. Clearly this is a major advance on the
Republic's m y t h of Er. A l t h o u g h the sun, moon and planets have t h e i r own
motions, there is n o t h i n g to indicate there t h a t these are i n any way
inclined to the motions of the fixed stars. Indeed such an idea simply w i l l
not f i t w i t h the image of a spindle. W h i l e the Pythagoreans may w e l l have
k n o w n of the ecliptic, and indeed may well have divided up the e a r t h into
the relevant tropical and polar zones, they were not able to introduce an
offset ecliptic into t h e i r cosmology.18

So the Timaeus model, like t h a t of the Republic, can account (at least i n
a qualitative manner) for motions of the fixed stars and the periods of orbit
of sun, moon and planets. I n addition, the Timaeus model has the offset
ecliptic and so can account for the motion of sun, moon and planets as offset
to t h a t of the stars. The model of the sun w i l l also be able to give an account
of the seasons, the v a r y i n g height of the sun, its variations i n setting points
and the fact t h a t i t exhibits tropai. These are clearly significant improve­
ments on the model of Republic and t h a t of Philolaus. Let us now t u r n to
the problems w i t h the Timaeus model.

I I I . Retrogression and deviation i n latitude

I w i l l refer to the view of the Timaeus whereby the stars have one regular
circular motion and the sun, moon and planets are assumed to have two
regular circular motions (that of the stars and t h e i r own), w i t h the e a r t h
central and static, and no other motions whatsoever, as the 'bare model'.
19

There are serious astronomical difficulties w i t h this model. I t cannot

account for the retrograde motion of the planets, the deviation of the moon
and the planets from the plane of the different, the relation of Venus and
M e r c u r y to the sun, the frequency of l u n a r and solar eclipses, the fre­
quency of occultations of planets and stars, the inequality of the seasons
or variations i n the apparent size of heavenly bodies. As i t would seem t h a t
Plato was aware of at least some of these phenomena, attempts have been
made to find a more sophisticated astronomy i n the Timaeus t h a n is
allowed by the bare model. A n i m p o r t a n t question i n w h a t follows w i l l be
whether we should accept the principle t h a t Plato believed t h a t the model
of the Timaeus could deal w i t h a l l the astronomical phenomena he was
aware of. A second critical m a t t e r is this. The bare model is very strong
cosmologically, as i t adheres strictly to the principle of regular circular
motion. I n terms of astronomy, however, i t is weak i n being unable to
account for many phenomena t h a t Plato was aware of, indeed more t h a n
is generally recognised. I n order to give the Timaeus a reasonable astro-
132 Plato's Philosophy of Science

Fig. 20. The zodiac. Here the

ecliptic (the path of the sun) is
the dotted line, while the two
plain lines denote the boundaries
of the zodiac, the band within
which the planets move.

Fig. 21. Simplified zodiac. For

simplicity, we can represent the
ecliptic and the zodiac as here.
I— ——— —;—;—r-—;— — — —.
: : ; : : We can then speak of longitude as
parallel to the ecliptic/zodiac, and
—;— :
—\—i—i—\—\—i—•—!— latitude as up or down deviation
from the line of the ecliptic. The
zodiac can be divided into twelve
parts, or houses.

nomy i n these terms, i t is necessary to abandon t h e bare model a n d

introduce a number of a r b i t r a r y motions w h i c h are either non-regular or
non-circular or both. I n w h a t follows I argue t h a t we need to do t h i s to a
greater extent t h a n is generally recognised, and this not only compromises
the principle of regular circular motion b u t threatens to wreck i t com­
pletely. There is t h e n a considerable tension between the astronomical and
the cosmological merits of this model, and as we shall see later, this raises
some i m p o r t a n t problems of i n t e r p r e t a t i o n both for Plato and for ancient
astronomy and cosmology i n general.
Let us begin w i t h the questions of retrogression and deviation. We have
seen t h e p a t h t h a t t h e s u n follows throughout a year. The moon and t h e
planets do not follow this path, b u t deviate to either side of it, and move
relative to t h e fixed stars t h r o u g h a band around t h e p a t h o f t h e s u n
k n o w n as t h e zodiac. W h i l e t h e sun's m o t i o n is r e l a t i v e l y simple, t h a t
of t h e five planets k n o w n t o t h e ancient Greeks is not. N o t only do t h e y
deviate i n l a t i t u d e from the ecliptic, t h e y undergo a phenomenon k n o w n
as retrogression. They progress r e l a t i v e to t h e fixed stars, b u t t h e n
become stationary, regress, become s t a t i o n a r y a n d t h e n move o n once
more.
Planetary motion, and retrograde motion i n particular, shows u p one of
the key differences between geocentric Greek astronomy a n d modern
heliocentric explanations. For us, the planets do not stop and t h e n start
 5. Plato and the Development of Greek Astronomy 133

Fig. 22. Retrogression and

deviation from the ecliptic. A
schematic representation of planetary
retrograde motion relative to the
background of the fixed stars.
it
it
it
it *

Fig. 23. Patterns of retrograde

motion. We can represent this on a
zodiac diagram. Note that the 'normal'
path of the planet is not along the
ecliptic and that retrograde motion
involves deviation in latitude as well
as regression in longitude. The second
diagram gives another pattern of
retrogression.20

moving again. They appear to do so against the background of the fixed

stars because both the earth and the planet are i n motion around the s u n . 21

They have different sizes and speeds of orbit, and at certain times these
factors combine to produce the appearance of retrograde motion for a
viewer from the earth. To anyone who holds to a geocentric and geostatic
view, however, retrograde motion has to be a real motion of the planets,
and so complex schemes have to be devised to account for such motion. The
reason t h a t the planets deviate from the ecliptic is t h a t the orbits of the
planets around the sun are nearly i n the same plane as t h a t of the earth,
b u t not q u i t e .
22

The bare model of the Timaeus cannot account for retrograde m o t i o n or

deviation i n latitude of the planets. The planets have only one m o t i o n
relative to the fixed stars, moving around the circle of the different at a
constant velocity, m a k i n g retrogression impossible. The planets do not
move out of the circle of the different, m a k i n g deviation i n l a t i t u d e
impossible. A l l the planets should follow the p a t h of the sun according to
the bare model, w i t h no need for a zodiac as a band around the p a t h of the
sun.
The moon also deviates from the ecliptic, although i t does not have any
retrograde motion. The reason for this deviation is t h a t the moon's orbit
around the earth is not i n the same plane as the earth's orbit around the
sun. The inclination is quite small, as indeed are the inclinations of the
134 Plato's Philosophy of Science
Fig. 24. The moon's motion
through the zodiac. Greeks take
the moon's deviation in latitude as 6°
rather than the modern figure of 5 . 0 23

Fig. 25. Plato's two spheres. Here

we have one sphere for the motion of
the same and one sphere for the
motion of the different, so one sphere
for the daily rotation and one to
generate motion along the ecliptic.

orbits of the planets to the orbit of the earth, w h i c h is w h y the zodiac is a

relatively narrow band. I t is i m p o r t a n t to note t h a t the bare model cannot
account for this deviation i n latitude of the moon either, for the same
reason as w i t h the planets. This is of considerable significance as we shall
see soon, as i t plays an i m p o r t a n t role i n the frequency of eclipses.
Before we move on, let us briefly look at how these matters are dealt
w i t h by Eudoxus. Plato's model for sun, moon and planets of the heavenly
bodies can be represented as i n Fig. 25. Eudoxus' model for the sun and
the moon involves three spheres (Fig. 26). Eudoxus' model for any of the
five planets is more complex (Fig. 27). As is well k n o w n , Eudoxus' model
allows for both retrogression and deviation from the ecliptic for the plan­
ets. F i r s t , let us conceive Plato's model i n mathematical terms. One sphere
here generates the motion of the fixed stars, the other sphere, w i t h its axis
set at an angle to the first, generates motion along the ecliptic. Eudoxus'
model for the planets has two extra spheres. T h e i r function is to generate
a figure k n o w n as a hippopede (Fig. 28). W h e n the four motions are
combined, w i t h a suitable choice of axes and speeds, t h e n the r e s u l t i n g
motion gives an interesting representation of the motions of the planets.
One gets both deviation from the ecliptic and retrograde motion (Fig. 29).
W h i l e this is far from a perfect representation of the heavens, and there
are indeed serious i n t r i n s i c problems, i t is a major advance over the
Timaeus model.
 5. Plato and the Development of Greek Astronomy 135
i

Fig. 26. Third sphere for the sun

and the moon. Two spheres as ever
generate daily motion and motion
along the ecliptic. Another sphere
generates some deviation from the
ecliptic. The angle at which it is set
24

to the ecliptic gives the maximum

deviation. 5

Fig. 27. Eudoxus' four

spheres. Here we have the same
first two spheres as with Plato,
but this time we have two more
to generate the hippopede.

Fig. 28. Hippopede. The possible

shape generated by the two spheres
Eudoxus adds to Plato's model. 26

Fig. 29 Resulting motion. The

resulting motion of the four spheres 6 2
then looks like this.
4
136 Plato's Philosophy of Science

IV. V e n u s , M e r c u r y a n d the c o n t r a r y p o w e r
There is a further problem for the bare model w h i c h is t h a t i t cannot
account for the relation of the sun, M e r c u r y and Venus. M e r c u r y and Venus
are always to be seen relatively near to the sun i n terms of angular
separation, while the other three planets (Mars, J u p i t e r and Saturn) may
be seen at any angle relative to the sun. The reason for this i n modern
terms is t h a t M e r c u r y and Venus are inferior planets, t h a t is they orbit the
sun inside the orbit of the earth, while M a r s , J u p i t e r and S a t u r n are
superior planets and orbit the sun outside the orbit of the earth (Fig. 30).
The consequence of this is t h a t M e r c u r y and Venus m u s t be seen as
relatively close to the sun when observed from earth, while M a r s , J u p i t e r
and S a t u r n can be seen at any angle relative to the sun.

Fig. 30. Inferior and superior planets. Mercury and Venus have smaller orbits
than the earth and are known as inferior planets; Mars, Jupiter and Saturn have
larger orbits and are known as superior planets.

Fig. 31. Possibilities for inferior planets. Mercury and Venus, as they 'pass in
front of' the earth, can be seen only at certain angles relative to the sun. Wherever
the earth is in its orbit Venus and Mercury will be seen relatively close to the sun.

Fig. 32. Maximum angular separation of inferior planets. The maximum

value for the angular separation of an inferior planet from the sun can be seen from
this diagram. For Mercury it is around 23°, for Venus around 48°.
 5. Plato and the Development of Greek Astronomy 137

Fig. 33. Possibilities for superior planets. Mars, Jupiter and Saturn, as they
'pass behind' the earth, can be seen at all angles relative to the sun. The problem
for any geocentric cosmology is to explain why this should be so, and indeed there
was a considerable struggle to do so in a plausible manner (Copernicus was
possibly the first to do so).

I do not w i s h to a t t r i b u t e to Plato a distinction between inferior and

superior planets. He does recognise t h a t there is something different about
the orbits of M e r c u r y and Venus, though, and t h a t these planets ought to
be grouped w i t h the sun. Certainly this seems to be the case not only i n
the Timaeus, b u t i n the Republic, 27
and the Epinomis as well, at 986e, and
at990b5ff.:

The fourth orbit and revolution, and also the fifth, have a speed nearly equal
to that of the sun, and neither slower nor greater on the whole. I t is necessary
that of the three that with the mind most adequate to the task leads.

The moon goes through her orbit most rapidly, bringing firstly month and
full moon. Secondly i t is necessary to understand the sun, the turnings he
brings through the whole of his orbit, and the followers he brings with him.

M e r c u r y and Venus are t h e n always found i n the vicinity of the sun. A t

some times they move more swiftly t h a n the sun t h r o u g h the zodiac, at
others more slowly, and can appear on either side of the sun, appearing to
overtake i t and t h e n be overtaken. As Plato puts i t at Timaeus 38d:
28

The morning star [Venus] and the star sacred to Hermes [Mercury] he placed
in an orbit equal in speed to that of the sun, but possessing a contrary power
(enantian ... dunamin) to it, whence i t is that the sun, the star of Hermes
and the morning star overtake and are overtaken by each other.

This raises the rather vexed issue of the nature and extent of the enantia
dunamis, the 'contrary power'. The bare model cannot account for either
the M e r c u r y and Venus problem, or retrogression, though Plato appears to
be aware of b o t h . As the only extra device we have available is the
29

contrary power, there is a great t e m p t a t i o n to seek an i n t e r p r e t a t i o n of the

contrary power such t h a t i t can deal w i t h both of these problems. Yet the
contrary power is only explicitly applied to M e r c u r y and Venus. According
to Vlastos, this is because Plato is:
138 Plato's Philosophy of Science
explicitly denying us here an account of the 'others' (i.e. of Mars, Jupiter and
Saturn), saying that this he might perhaps (isos) do adequately (axios) at
some future time (38d6-e2): declining to go into the particulars of the
motions and periods of the outer planets (to no one of which does he make so
much as a single individual reference), he has no occasion to speak of their
retrogradation. 30

This I find unconvincing. Merely to say t h a t M e r c u r y and Venus have the

contrary power is (notoriously!) not to give much of an account of t h e m ,
and certainly is not an account of t h e i r retrogression ( i f this w h a t the
contrary power is) t h a t cannot be equally tersely applied to the other
planets. I have grave reservations about this extension of the contrary
power. I t is not clear t h a t the contrary power is about retrogression at a l l .
One could have an account of the relation of Mercury, Venus and the sun
whereby M e r c u r y and Venus at times progress more slowly t h r o u g h the
zodiac t h a n the sun, and so are overtaken by i t , w i t h o u t M e r c u r y and
Venus ever being stationary or h a v i n g retrograde motion relative to the
fixed stars. T h a t would cover a l l Plato says at Timaeus 38d, and i t is
notable t h a t there is no reference to retrogression here. Indeed, the vast
majority of the motion of Venus and M e r c u r y is progressive. Against this,
i f there is a means by w h i c h a planet's progression can be slowed i t is no
great leap to create stationary points and retrogression. However, i t is also
quite possible t h a t Plato treats the relation Venus and M e r c u r y to the sun
and retrograde motion as two distinct problems. The fact t h a t Plato uses
31

the phrase 'contrary power' only for M e r c u r y and Venus i n relation to the
sun and for no other celestial bodies while he has a serviceable phrase for
retrogression is indicative t h a t he recognised t h a t there were two separate
problems here.
As to the nature of the contrary power, Plato is quiet. One view of this
is t h a t Plato merely notes the behaviour of Mercury and Venus relative to
the sun and each other b u t does not claim, on the bare model of the
Timaeus, to be able to account for such behaviour. I shall r e t u r n to this
possibility later, b u t first let us look at two other interpretations of the
nature of the contrary power. 32

K n o r r proposes t h a t the contrary power refers to a 'periodic oscillating

motion', w h i c h can account not only for the orbits of M e r c u r y and Venus
b u t for retrograde motion as w e l l . This he links to the roles of the
33

daughters of necessity i n the m y t h of Er, i n particular Lachesis who

occasionally grasps the r i m s of the hemispheres and t u r n s them. He finds
that:

The passages we have surveyed in the Timaeus offer no explicit support for
this proposal even i f they are entirely consistent with it; but the Republic
does. 34

W h i l e this is a highly ingenious proposal, the Timaeus is not consistent

 5. Plato and the Development of Greek Astronomy 139
w i t h the m y t h of Er. I n the m y t h of E r the cosmos t u r n s on a spindle, and
the actions of Clotho and Atropos are i n t e r m i t t e n t , w h i c h w o u l d f i t w e l l
w i t h the Republic's position on celestial motion, b u t not w i t h t h a t of the
^Imaeus.
For the Timaeus model, Clotho and Atropos would be redundant. As for
the role of Lachesis, as we have seen, one would have to be extraordinarily
generous to get retrograde motion (and no other strange effects) out of
Lachesis' actions, as there is no discrimination between w h i c h celestial
objects she affects or how she affects them. T h a t is a problem whether we
consider the contrary power to deal w i t h retrogression, or merely the orbits
of M e r c u r y and Venus.
A different approach is t h a t of Cornford, who has the more general idea
of self-motions for the moon and planets, the contrary power being one of
these self-motions. According to Cornford, as the planets have souls, they
can i n i t i a t e self-motions. So a l l the planets have an axial rotation, the
moon has the ability to go faster t h a n the speed of the different while M a r s ,
J u p i t e r and S a t u r n have the ability to go slower, and a l l of the planets,
except the sun and the moon retrogress. I n relation to retrogression he
comments:

This is the 'contrary tendency' (enantia dunamis) explicitly ascribed to Venus

and Mercury, but also shared by Mars, Jupiter and Saturn. I t involves
variations in the speed of each planet, and intermittent counter-revolution
accelerated to the point of bringing the main motion to a standstill and
temporarily reversing its sense. (None of these self-motions distorts in any
way the circular track of he planet's proper motion. So the planets do not
'stray' from one path to another, Laws 821, Epin. 982c.) 35

Cornford constructs w h a t at first sight looks a plausible via media. The

Timaeus model can account for retrograde motion and for M e r c u r y and
Venus, and we get a m i n i m a l deviation from regular circular motion, as
planetary motion remains strictly circular even i f i t is no longer regular.
So the Timaeus model can account for a l l the phenomena Plato is aware
of, and the tension between the cosmological and astronomical merits is to
some extent resolved. Can this compromise hold? I n the next section I
argue t h a t there are i m p o r t a n t phenomena t h a t Plato was aware of w h i c h
are not accounted for by Cornford's model.

V. E c l i p s e s a n d o c c l u s i o n s
There is a difficulty w i t h the bare model of the Timaeus, and indeed
Cornford's model as well, w h i c h is not often considered b u t is critical to a
proper evaluation of the Timaeus' astronomy. Quite simply, these models
w i l l produce far too many eclipses of the sun and the moon, and Plato
cannot have failed to be aware t h a t solar and l u n a r eclipses do not occur
each m o n t h as these models imply. For these purposes, we can reduce the
140 Plato's Philosophy of Science

Fig. 35. Phases of the moon.

bare model to one plane, as we only need to consider the relations of the
earth, sun and moon (Fig. 34). F r o m this i t w i l l be evident t h a t a great
number of eclipses and occlusions w i l l occur as sun, moon and a l l the
planets are i n the same plane. To understand the reasons for the regularity
of eclipses of sun and moon, we need to understand how they are related
to the phases of the moon (Fig. 35). As the moon is l i t by the sun, how m u c h
of the moon we can see is dependent on the relative positions of earth,
moon and sun. I t should be clear t h a t l u n a r eclipses can occur only close
to full moon, and solar eclipses only close to new moon. They do not occur
every new and full moon due to the inclination of the moon's orbit to the
ecliptic. The points at w h i c h the moon passes t h r o u g h the plane of the
ecliptic (the only t i m e when eclipses can occur) are subject to complex
change. Eclipses occur when new or full moon coincides w i t h passing
t h r o u g h the plane of the ecliptic.
 5. Plato and the Development of Greek Astronomy 141
F r o m an understanding of the phases of the moon, i t is evident t h a t a
solar eclipse can only occur at new moon and a l u n a r eclipse can only occur
at full moon. I f the moon's orbit of the earth is i n exactly the same plane
as the earth's orbit of the sun, t h e n there w i l l be an eclipse of the sun every
new moon and an eclipse of the moon every full moon. The basic problem
here is t h a t Plato does indeed have the sun and the moon i n the same
plane, as they both move on the circle of the different. I t was k n o w n at
Plato's t i m e t h a t eclipses occur only at new or full moon, though not why.
Empedocles may well have k n o w n the causes of eclipses i n the sense of the
alignment of sun, moon and earth, b u t nothing suggests t h a t he could
explain w h y they occurred at specific t i m e s . I n fact the moon's orbit of the
36

earth is rather complex. The plane of its orbit is slightly inclined to the
plane of the earth's orbit of the sun, and the points at w h i c h the moon
passes t h r o u g h the plane of the earth's orbit are subject to change. Only
w h e n these points occur on, or very near to the line between the earth and
the sun do total, or p a r t i a l eclipses occur.
This is a rather neglected problem w i t h the astronomy of the Timaeus,
and none of the proposed more sophisticated models deal w i t h i t . Yet the
3 7

frequency of eclipses was something t h a t was of great importance both for

the Babylonians and for early Greek astronomy. One m i g h t w e l l argue t h a t
eclipses are a more evident and more i m p o r t a n t phenomenon t h a n plane­
t a r y retrogression or the M e r c u r y and Venus problem. There is not an easy
solution to this problem for the more sophisticated models. Those of K n o r r
and Cornford have no deviation from the plane of the different, and even
i f the contrary power could deal w i t h deviation, i t cannot be extended to
the sun, as i t is the sun t h a t other objects have this power contrary to, and
there is no t e x t u a l evidence for extending i t to the moon. One is left w i t h
supposing some self-motion for the moon to generate deviation.

Fig. 36. Moon, sun and month.

Month can be defined as either the
time taken by the moon to return
to the same position relative to the
fixed stars or as the time taken to
complete a set of phases, i.e. return
to the same position relative to the
sun. As the sun moves during this
time, the latter is slightly longer.
142 Plato's Philosophy of Science
Plato's model of the moon, however, does have its plus points. We do get
a correct theory of the nature of t o t a l eclipses out of i t , even i f there is a
problem about t h e i r frequency. A l t h o u g h Plato does not explicitly give us
a theory of the phases of the moon, a solid basis for the correct theory has
been laid, and Timaeus comments at 39c3-4 t h a t ' M o n t h occurs whenever
the moon completes her own circuit and overtakes the sun.' This w o u l d
seem to indicate t h a t Plato at least recognises t h a t the phase cycle of the
moon is related to the relative positions of sun, moon and earth. Certainly
t h a t seems to be the case at Epinomis 978e.
I t is certainly the case t h a t Eudoxus made an attempt to rectify some of
the problems here. As we have seen, he added two extra spheres to each of
the planets i n an attempt to generate retrograde motion. Eudoxus also
added one sphere each for the sun and the moon i n order to generate some
deviation from the ecliptic. Two spheres as ever generate daily motion and
motion along the ecliptic. Another sphere generates some deviation from
the ecliptic. The angle at w h i c h i t is set to the ecliptic gives the m a x i m u m
38

d e v i a t i o n . Callippus adds two further spheres for sun and moon to

39

account for variations i n orbital velocity. I f you do not make some

40

addition such as Eudoxus' t h i r d sphere, and as w i t h epicycles and h i p -

popedes there is no evidence for this i n Plato, either you cannot account
for deviation i n latitude i n general and the frequency of eclipses, or one
m u s t suppose some form of non-regular and non-circular motion.
W i t h the bare model, i n addition to the problem w i t h eclipses, the moon
w i l l produce occultations of the other planets rather too often. So too the
41

planets w i l l produce occultations of each other i n sequences, and a l l the

planets w i l l produce occultations of the fixed stars on the same plane as
t h a t of the different, b u t no others. Again, we can simplify t h i s down to one
plane (Fig. 37). T h a t such phenomena had been studied is attested by

Fig. 37. Bodies on the plane

of the different. Each inner
body will occlude each outer
body far too often, i.e. the moon
will occlude all the other bodies
roughly once a month. Only
those stars which fall on the
plane of the different will be
occluded.
 5. Plato and the Development of Greek Astronomy 143
Aristotle at Meteorologica 343b8 ff. and b28 ff. where he refers to the
Egyptians on these problems i n relation to his theory of comets, as well as
to Greek observations. He also refers to an occlusion of M a r s by the moon
at De Caelo 292a3 ff., and mentions similar accounts of other stars w h i c h
are given by the Egyptians and Babylonians. There is a similar problem
w i t h the planets and the fixed stars, and from Timaeus 40c i t is evident
t h a t Plato is quite aware of the fact t h a t the planets at times obscure each
other. W h e n planets come to be i n the same section of the zodiac, they are
said to be i n conjunction. There are several possibilities for w h a t happens
w h e n planets pass each other. There may be a direct eclipse, there may be
a p a r t i a l eclipse, they may pass close enough t h a t to the naked eye the two
planets appear as one brighter, larger planet or they may be sufficiently
separated t h a t they merely pass each other. I t is clear that, as w i t h the sun
and the moon, i n the Timaeus model there w i l l be far too many times w h e n
the planets are absolutely i n line w i t h each other and no times w h e n they
have the same longitude (measured along the ecliptic) b u t different l a t i ­
tudes (again relative to the ecliptic) w h i c h are necessary for the other types
of phenomena. I t would be surprising i f Plato were not aware of this, and
Timaeus 40c6 ff. may indicate a fair amount of knowledge of these phenom­
ena:

The dances of these stars and their juxtapositions (parabolas) with one
another, the reversals and advances of their own cycles, which of the gods
come into line (sunapsesin) with each other and which into opposition, which
cover each other relative to us, and for what periods they each disappear and
again re-appear (katakaluptontai kai palin anaphainomena) ...

I f we take this passage to reiterate the same point, i t is simply about

eclipses and occlusions. However, i f there is a contrast here between
parabolas and sunapsesin, t h e n we m i g h t take parabolas to indicate a
close passing and sunapsesin to indicate being absolutely i n l i n e . I f so, 42

Plato is aware of the difference between at least some types of phenomena

associated w i t h planets passing each other, b u t i t is clear t h a t his model of
the Timaeus is not able to account for i t . More radically, instead of t a k i n g
katakaluptontai kai palin anaphainomena as 'disappear and again re­
appear', again a reiteration, one m i g h t t r y something along the lines of
V e i l themselves and again become bright'. I f so, there are some interesting
possibilities. This m i g h t be a reference to the fact t h a t the planets have
v a r y i n g apparent size and brightness, though 'veil themselves' w o u l d
suggest they disappear from view, and 'appear more b r i g h t l y ' w o u l d not be
an attested usage of anaphainesthai. I t may be a more specific reference
to w h a t happens w h e n t w o planets are very close to being absolutely i n
line w i t h the earth (same longitude, slightly different latitude), w h e n they
appear as one b r i g h t e r object, though parabolas may have already referred
to t h i s phenomenon. More intriguingly, i t m i g h t instead be a reference to
the variations i n apparent brightness shown by M e r c u r y and p a r t i c u l a r l y
144 Plato's Philosophy of Science
Venus as they move i n close p r o x i m i t y to the s u n . As they overtake or are
43

overtaken by the sun they disappear from view (are lost i n the brightness
of the sun) only to reappear on the other side of the sun. I f any of these
possibilities are correct, t h e n Plato seems to know a fair amount about
these phenomena, and certainly rather more t h a n his model of the Ti­
maeus could account for. Note also t h a t i f the t h i r d possibility is correct
Plato keeps separate his references to retrogression (40c5) and to the
Mercury, Venus and sun problem (Fig. 38), and gives us a good catalogue
of the problems for astronomy i n the Timaeus.
Cornford's model cannot account for the frequency of eclipses, a phe­
nomenon Plato could not fail to have been aware of. So at least for the
44

moon, Cornford's via media fails. Depending on Plato's knowledge, i t may

w e l l fail for the planets as well. T h a t the planets deviate i n l a t i t u d e may
be inferred from the fact t h a t when they pass each other, at times there is
sufficient difference i n latitude for t h e m not to produce an occultation or
merge into one b r i g h t entity, but to pass w i t h clear space between them.

Fig. 38. Venus, Mercury and

the sun. Both of the inferior
planets (here Venus) are seen
relatively close to the sun, and as
Plato is aware, overtake and are
overtaken by the sun. When
Venus gets sufficiently close to
the sun, as at A, i t will no longer
be visible, and will re-appear on
the opposite side of the sun at B.
Similarly it will vanish again at
C and re-appear at D. I t may be
this phenomenon which Plato
refers to at Timaeus 40c8-9.
 5. Plato and the Development of Greek Astronomy 145
So too i t is evident from simply t r a c k i n g the p a t h of the planets against
the background of the fixed stars. This need not be very sophisticated, as
all one has too realise is t h a t the planets do not produce occupations of the
same stars on successive passes t h r o u g h the zodiac. One m i g h t make a
case on the grounds of Timaeus 40c t h a t Plato knew of either or b o t h of
these phenomena, though such a case would not be conclusive.
A further consideration here is the nature of retrograde motion. As
emphasised i n an earlier section, retrogression does not mean t h a t a
planet simply reverses its path, and often involves a significant deviation
i n l a t i t u d e to produce the loop and zig-zag shapes. Plato's t e r m for retro­
gression, i f t h a t is w h a t i t is, may be significant here. A t Republic 617b2
45

he uses epanakukloumenon, and at Timaeus 40c5 he uses epanakuklesis.

This means 'circles back on i t s e l f . I t is quite possible t h a t this is a
reference to the loops produced i n retrogression, and i f so Plato m u s t have
realised t h a t there is deviation i n latitude by the planets. I t is of course
difficult to produce a w a t e r t i g h t case here, and I do not pretend to do so.
I t is quite possible t h a t Plato was aware of deviation i n l a t i t u d e by the
planets, and on balance I would say i t is likely t h a t he was aware. I f so,
there are further problems for Cornford's model w h i c h cannot account for
any deviation i n latitude.

V I . Variations i n planetary velocity a n d distance

As mentioned earlier, Plato's model for the motions of the sun is good b u t
by no means perfect, even i n relation to ancient knowledge of the heavens
(Fig. 39). The bare model cannot account for the unequal l e n g t h of the
seasons, something t h a t was k n o w n of i n Plato's time. For the definition
of the seasons, there are four critical days i n the year. W i n t e r solstice, the
day of longest night, spring equinox, a day of equal day and night, summer
solstice, the day of longest day, and vernal equinox, the other day of equal
day and night. The seasons are the periods from solstice to equinox or
equinox to solstice, b u t they are not of equal length. The Babylonians d i d
not recognise this, b u t Euctemon and M e t o n h a d by the fourth century, and
they h a d measured values for the inequality of the seasons. B y around46

330 B C , Callippus had a remarkably good estimation of the i n e q u a l i t y of

the seasons. The following information is from the Eudoxus papyrus. This
does not give the values for the final quarter of the year. Those given are
calculated on the basis of a 365-day year. N o r are there figures for the first
quarter for Democritus or Eudoxus, those given here being Neugebauer's
conjectures. Neugebauer makes a tentative assumption t h a t Eudoxus d i d
not k n o w of the inequality of the seasons, as he made no effort to account
for t h i s . Equally, we m i g h t say he d i d know b u t could not account for this,
47

and conjecture say 90, 92, 9 1 , 92.

146 Plato's Philosophy of Science
Democritus Euctemon Eudoxus Callipp
and Meton
Summer solstice to
autumn equinox (91) 90 (91) 92
Autumn equinox to
winter solstice 91 90 92 89
Winter solstice to
spring equinox 91 92 91 90
Spring equinox to
summer solstice (92) 93 (91) 94

Fig. 39. Plato and the equality of

the seasons. Plato's model is far too
simple to account for the inequality of
the seasons. All we need to consider
here is the relation between sun and
earth. If the sun moves in a circle at a
steady speed there will be perfect
equality of the seasons.

Spring Winter

Fig. 40. Hipparchus'

model. An example of what
one must do to retain simple
circular motion and account
for the inequality of the
seasons. Hipparchus' model
retained regular circular
motion but offset the earth
from the centre of the sun's
orbit to generate the
inequality of the seasons.

Summer Autumn
 5. Plato and the Development of Greek Astronomy 147
A g a i n we can simplify matters down to a plot of the motion of the sun
on the different. I n order to generate any inequality i n the l e n g t h of the
seasons here the orbit of the sun would have to be made out of r o u n d (Fig.
39) or there would have to be some v a r i a t i o n i n the orbital velocity of the
sun. The contrary power is not a possibility here as i t is a power contrary
to t h a t of the sun, so we would need to assume some form of self-motion
v a r y i n g the speed of the sun.
Callippus, perhaps on the basis of the observation of the inequality of
the seasons by Euctemon and Meton, goes beyond Eudoxus and adds
another two spheres for the sun and two more for the moon to generate
variations i n velocity for the sun and m o o n . Schiaperelli conjectures t h a t
49

the extra pairs were hippopedes, w h i c h w i t h appropriate values w o u l d give

reasonable estimates of the variations i n velocity. 50

The bare model of the Timaeus cannot account for variations i n the
apparent size of the celestial bodies. D i d Plato k n o w of such variations?
According to Simplicius in De Caelo 504.17 ff.:

These [unrolling spheres] of Eudoxus' school do not save the phenomena, not
only those that were found later, but also those known before (ta proteron
gnothenta) and recognised by them. 51

The phenomena w h i c h Simplicius goes on to refer to are t h a t Venus and

M a r s appear at times m u c h brighter t h a n at others, t h a t there is v a r i a t i o n
i n the size of moon, and there is also the m a t t e r of annular eclipses. The
moon can produce solar eclipses while at v a r y i n g distances from the earth.
W h e n i t is near i t covers the whole sun, w h e n i t is far away, a small r i n g
of the sun is visible around the moon at t o t a l eclipse, the phenomenon of
annular eclipse. Whether these phenomena were k n o w n to Plato is a
m a t t e r of conjecture. The closest t h i n g to direct evidence we have is
Epinomis 978d7, where the moon is said to appear larger and smaller (e
tote men meizon, tote d'elatton), w h i c h would not appear to be a reference
to shape i n r e l a t i o n to phase, especially as the next line appears to give us
such a reference. Again, one could suppose a self-motion for t h i s , t h o u g h
52

t h a t w o u l d be another breach of regular circular motion and a new

departure for self-motions (which so far have been for longitude, t h e n
latitude, and now distance). I f we accept t h a t Plato knew of variations i n
apparent size or brightness, and' t h a t is quite possible, t h e n there w o u l d
need to be variations i n the distances of the heavenly bodies. I f so, and we
p u t this i n Plato's terms, we have now u t t e r l y wrecked regular circular
motion as we have a t t r i b u t e d deviations from i t i n a l l six motions w h i c h
Plato recognises; left and r i g h t (around the plane of the different), up and
down (deviation i n latitude), forward and back (deviation i n distance).

1. A t Timaeus 38d, we have the problem of w h y M e r c u r y and Venus and

the sun overtake and are overtaken by each other. As argued above, I
148 Plato's Philosophy of Science
believe we should take this to be a separate problem for Plato from t h a t of
planetary retrogression.
2. A t Timaeus 39cd, we have the problem of the times of orbit of M a r s ,
J u p i t e r and Saturn. This may be slightly more complex t h a n is generally
recognised, as immediately prior to this Plato has Timaeus say t h a t m o n t h
is the t i m e the moon takes to come back to the same place i n relation to
the sun, thus giving a synodic month. Plato may t h e n w a n t the synodic
periods of the planets (possibly as well as the sidereal periods i f the great
year is to be calculated, as t h a t is done relative to the same). Note i n
relation to this t h a t the Epinomis (990a) criticises Hesiod as one who is
not a t r u e astronomer, being someone who has only 'observed settings and
risings' rather t h a n studied orbits.
3. A t Timaeus 40cd, the nature of the problems is more debatable, b u t
I w o u l d argue t h a t we have the problems of planetary retrogression, of
juxtaposition and eclipses, of the times of eclipses and perhaps the disap­
pearance and reappearance of Mercury and Venus as they pass the sun.
4. A t Epinomis 978d7 we have a possible reference to v a r i a t i o n i n the
size of the moon.
5. A t Timaeus 47a5 we are told that: 'Sight of day and night, of months
and the revolving years, of equinoxes and solstices (isemeriai kai tropai), 53

have caused i n v e n t i o n of number.'

We can thus reasonably attribute at least a recognition of the phenomena

of solstice and equinox to Plato. Plato also refers to solstices i n the Laws
(767cd and 945e) as a means of m a r k i n g out the political year and refers
to the t u r n i n g s of the sun at Epinomis 990b. So too at Phaedo 98a Socrates
says t h a t he was determined to find out about:

the sun, the moon and the other stars, their speeds relative to each other,
their turnings (tropon) and what else they underwent.

A l t h o u g h he does not directly say so, i t is beyond credence t h a t Plato was

not aware t h a t l u n a r and solar eclipses do not occur every m o n t h . There
are m a n y phenomena w h i c h were k n o w n of by those interested i n astro­
nomy at the t i m e of Plato or j u s t after, so i t is quite possible t h a t he was
aware of the problem of the inequality of the seasons, other evidence on
variations i n size and brightness, or the deviation of the moon and planets
from the p a t h of the sun.

V I I . Prototype or finished article?

The consequence of a close examination of the astronomy of the Timaeus
is t h a t we appear to have an a w k w a r d dilemma. The evidence i n favour of
regular celestial motion i n the Timaeus is very strong, as we saw i n the
previous chapter. This though leads to a poor astronomy, and surely Plato
 5. Plato and the Development of Greek Astronomy 149
cannot have been so ignorant of astronomy to t h i n k this a good model. A
good astronomy t h o u g h would not only compromise regular circular mo­
tion, b u t w o u l d wreck i t . Cornford's attempt to construct a v i a media of a
reasonable astronomy w i t h m i n i m a l deviation from regular circular mo­
t i o n also fails as i t cannot account for the frequency of eclipses, something
t h a t Plato was certainly aware of, and several other phenomena t h a t he
may w e l l have been aware of. The choice between a good astronomy and a
good cosmology t h e n appears to be a stark one.
A t t e m p t s to make the astronomy of the Timaeus more sophisticated by
reading into i t later developments i n Greek astronomy are also h i g h l y
suspect. Van der Waerden's suggestion t h a t Plato held an early form of the
epicycle theory lacks evidence. Plato never mentions epicycles w h e n he
54

could describe this critical device for circular motion astronomy i n a few
lines, and there are more plausible interpretation of the contrary power. I f
Plato d i d use epicycles, t h e n w h y do we find no trace of t h e m i n Eudoxus,
Callippus and Aristotle when they could solve some of the problem of
concentric sphere astronomy? We m i g h t , on the same grounds a t t r i b u t e
hippopedes to Plato. As epicycles can solve some quite sophisticated prob­
lems, w h a t degree of knowledge, and on w h a t grounds, do we a t t r i b u t e to
Plato? Zeyl has commented t h a t w h e n the circle of the different is divided:

The planes of the seven smaller circles are not necessarily parallel to the
ecliptic or to each other, but vary within the outer limits of the Zodiac. 55

This w o u l d solve the problem of general deviation i n l a t i t u d e (but not

retrogression), b u t I do not see any evidence t h a t Plato held this h i g h l y
i m p o r t a n t idea of having the motions of the planets offset to one another, 56

w h i c h was only introduced m u c h later, nor do Eudoxus, Callippus or

Aristotle make use of the idea when i t would considerably simplify and
improve t h e i r systems.
So we m i g h t t r y a different approach. Ought we to accept the principle
t h a t the Timaeus model was adequate for a l l the phenomena k n o w n to
Plato? I t is by no means clear t h a t we should. I f we consider this as a more
general principle for ancient astronomy (everyone believed t h e i r own
model could, etc.), t h e n i t w i l l certainly not do. The most pertinent case
w o u l d be t h a t of Eudoxus. Most commentators would agree he almost
certainly w o u l d have k n o w n of the inequality of the seasons, b u t could not
account for i t , and we have Simplicius' evidence on Venus and M a r s ,
annular eclipses and the size of the moon, a l l of w h i c h were t a k e n to i m p l y
v a r i a t i o n i n distance but could not be accounted for. He may w e l l have
k n o w n t h a t his system could produce reasonable results for Mercury,
J u p i t e r and S a t u r n b u t r a n into serious difficulties w i t h Venus and failed
to produce any retrogression for M a r s . I t is h a r d to believe t h a t Eudoxus
57

did not realise at least some of the shortcomings. One m i g h t argue some­
t h i n g s i m i l a r for C a l l i p p u s , t a k i n g due account o f h i s proposed
150 Plato's Philosophy of Science
improvements, and indeed Simplicius tells us t h a t no one up to Autolycus
could account for the phenomena he lists i n relation to variations i n
distance. I would suggest t h a t the great majority of ancient astronomers
were aware of discrepancies between t h e i r models (whether merely
mathematical or physical) and the motions of the heavens. This of course
w o u l d be one of the major factors d r i v i n g t h e m forward. We need t h e n
some evidence or argument to support this principle i n Plato's case, and
t h a t does not appear to be forthcoming.
There is a further interpretational issue here. Cornford's argument for
i r r e g u l a r motion begins by citing Taylor on the contrary power. Taylor
comments:

The Timaeus does not tell us why the two planets and the sun in turns gain
on one another. No explanation could be offered by a man who assumed all
three to be revolving with uniform velocities i n the same sense and with the
same period in concentric circular orbits. 58

Cornford replies:

This seems to me to be a reason for concluding that Timaeus does not make
all these assumptions, which would render the phenomena not merely
inexplicable but impossible. 59

There are phenomena w h i c h Eudoxus and Callippus w o u l d have k n o w n

about w h i c h are impossible on t h e i r models. Do we t h e n conclude t h a t they
d i d not adhere to regular circular motion? Or are we better off under­
standing t h e i r models as improvements on previous regular circular mo­
t i o n models, w i t h E u d o x u s a n d C a l l i p p u s aware of some of t h e i r
deficiencies?
I t seems to me t h a t Cornford's view falls between two stools and
illustrates some of the difficulties of interpretation involved here. O n the
one hand, we have the bare model w h i c h cannot account for retrogression,
p r o x i m i t y and the frequency of eclipses. Alternatively, one can produce a
good correspondence between the Timaeus model and observations k n o w n
at the t i m e by a general application of the contrary power b u t lose the idea
of regular circular motion. Indeed one can make the accuracy of such a
self-motion model a r b i t r a r i l y good. By i n v o k i n g self-motion for linear
retrogression Cornford saves observation b u t loses regular circular mo­
tion, b u t t h e n for deviation from the plane of the different he saves regular
circular motion b u t loses correspondence w i t h observation (and indeed
makes the phenomena impossible!). A t this point he needs an argument as
to w h y one should change strategy here, some criterion for l i m i t i n g the
application of self-motion. That, though, is h i g h l y problematic once we go
beyond the confines of Timaeus 38d, as there is no t e x t u a l reason for any
such l i m i t a t i o n and self-motion could account for a n y t h i n g .
60

Plato could very easily have gone along the route of citing a generalised
 5. Plato and the Development of Greek Astronomy 151
self-motion wherever the bare model failed to meet up w i t h observations.
This could have been done w i t h a m i n i m u m of effort. A l l he has to do is add
two or three words about the contrary power, or self-motion w h e n he
mentions retrogression and eclipses. Plato, however, does not make this
explicit w h e n he easily have done so and wisely leaves options for further
investigation open. I would emphasise t h a t a l l Timaeus 3 8 d l - 6 says is t h a t
M e r c u r y and Venus have some power i n addition to t h a t w h i c h the sun has,
and t h a t may w e l l be only a means of m a r k i n g out the question of the
p r o x i m i t y of M e r c u r y and Venus to the sun from other astronomical
problems. 61

I f we were to accept t h a t Plato thought the Timaeus model to be

adequate for the phenomena he was aware of, t h e n i t would seem t h a t one,
or some combination of the following would be the case: 62

1. Plato blundered i n believing his model was adequate for the phe­
nomena of w h i c h he was aware.
2. Plato was ignorant of some basic and evident observational astron­
omy, w h i c h was well k n o w n to those around h i m .
3. Plato believed i n some (non-regular, non-circular) self-motions and
either d i d not realise or was not concerned by the extent to w h i c h this
would sabotage the emphasis on regular circular motion i n the Timaeus
and Laws.

Yet there is n o t h i n g w h i c h compels us to accept this principle, w h i c h seems

implausible i n relation to many other ancient thinkers. W h a t do we do i f
we abandon this principle? A far better option is simply to assume t h a t
Plato intended the model of the Timaeus to be a prototype w h i c h sets out
some basic principles of cosmology. I t would be a prototype of models w h i c h
employ combinations of regular circular motion and motions for the sun,
moon and planets offset to those of the fixed stars. As such, i t is strong
cosmologically though weak i n terms of astronomy. Plato may well have
been aware t h a t this simple model cannot account for some major phenom­
ena, b u t we have to take into account t h a t this is a very early stage i n
model building. A l t h o u g h the model of the Timaeus is weak i n terms of
astronomy, i t is a significant advance on the model w h i c h Plato gives i n the
Republic, w h i c h does not offset the motion of the planets. I t w o u l d also be
an advance on the model offered by Philolaus. 63

Some points i n favour of this idea. The general evidence for regular
circular motion i n the Timaeus would seem very strong, as we saw i n the
last chapter. Plato's concern, i n general, is more w i t h teleological cosmo­
logy. Another w o r r y here is that, as Vlastos has pointed out, i f Plato d i d
employ self-motions, this really does not speak well of h i m . As Vlastos
comments: T h e explanatory value of such an ancillary hypothesis would
have been bogus.' Indeed, any use of the contrary power/self-motion looks
64

entirely ad hoc, and has no explanatory or predictive power. I w i l l r e t u r n

152 Plato's Philosophy of Science
to the question of the relation of Meno's paradox to astronomy and cosmo­
logy i n a l i t t l e while, b u t here I w i l l j u s t point out t h a t any position created
by a generalisation of the contrary power/self-motion w o u l d be an exceed­
i n g l y unhealthy hypothesis.
I f the model of the Timaeus is a prototype, t h e n the role of the Platonic
challenge reported by Simplicius is clear. There are problems i n astronomy
w h i c h the bare model of the Timaeus is not able to solve, w h i c h should be
solved i n terms of regular circular motion. Plato wisely leaves those
difficulties open for further investigation, b u t how one should set about
this challenge is clear. I f Plato believed the model of the Timaeus to be a
final one, or d i d not believe i n regular circular motion, t h e n the challenge
makes l i t t l e sense.
Why m i g h t Plato be optimistic t h a t such phenomena as retrogression
and the relation of M e r c u r y and Venus to the sun m i g h t be solved by
regular circular motion? Perhaps because Plato's model for the sun does
something similar. I f we consider the setting points of the sun on the
horizon d u r i n g a year, they progress to a northerly m a x i m u m , h a l t , move
to a southerly m a x i m u m , h a l t and progress again. T h a t problem is solved
by i m a g i n i n g two regular circular motions for the sun i n combination. One
m i g h t describe the retrograde motion of the planets and the motions of
M e r c u r y and Venus relative to the sun i n a s i m i l a r manner, and hope t h a t
further regular circular motions w i l l solve these problems.
The model of same and different, being effectively t h a t of the m i n d of a
visible god, has a great many other functions to perform i n terms of
epistemology, ethics and psychology as we shall see later. I t is not solely
an astronomical model and so we should not be overly surprised t h a t i t is
strong on cosmological principle but weaker on astronomical detail.
Cornford believes t h a t he can b r i n g his view of the Timaeus i n line w i t h
the Laws and Epinomis. O n the question of circular motion, t h a t is indeed
so. However, Cornford believes t h a t each of the planets has an i n t e r m i t ­
tent self-motion w h i c h creates retrograde motion. As we shall see i n the
next chapter, this most certainly cannot be brought into line w i t h the view
of the Laws and the Epinomis, which are most insistent t h a t the celestial
gods always have acted and always w i l l act i n an entirely consistent way,
and simply do not act i n a n y t h i n g like an i n t e r m i t t e n t manner.
Finally, we get a good t r a n s i t i o n into the w o r k of Eudoxus. W i t h Plato
we have regular circular motion and a two-sphere model for sun, moon and
planets. Eudoxus can be seen as picking up the Platonic challenge w i t h
three-sphere models for sun and moon and four-sphere models for the
planets. I t is also possible to argue t h a t Eudoxus attempts to remedy the
major defects i n Plato's prototype. Thus he can (at least qualitatively)
account for retrogression, for deviation from the ecliptic, for the relation of
Mercury, Venus and the sun, and, as he has three-sphere models for both
sun and moon, he has the foundations for a m u c h improved theory of the
frequency of eclipses. 65
 5. Plato and the Development of Greek Astronomy 153
I f so, are we not better off considering Plato's model as a cosmologically
strong prototype where he is aware of at least some of the astronomical
deficiencies? I t would be a prototype of models w h i c h employ combinations
of regular circular motion and employ such motions for the sun, moon and
planets offset to those of the fixed stars. Plato may well have been aware
t h a t this simple model cannot account for retrogression, p r o x i m i t y and
eclipses, and here we have to take into account t h a t this is a very early
stage i n model building. A l t h o u g h the model of the Timaeus is weak, i t is
an advance on the model w h i c h Plato gives i n the Republic and on the
model offered by Philolaus. I t is also i m p o r t a n t to note t h a t i f the model of
the Timaeus was a prototype, Simplicius' report of the challenge Plato set
the astronomers makes sense, w h i c h i t does not i f Plato believed himself
to have solved a l l the astronomical problems.

V I I I . T h e a s t r o n o m i c a l a n d cosmological t r a d i t i o n s
One m i g h t argue t h a t i n the Presocratic period there were t w o quite
separate traditions of the investigation of the heavens. There was a
t r a d i t i o n of observational astronomy, and of the meteorological calendars,
the parapegma. There was also a t r a d i t i o n of philosophical cosmology,
66

perhaps beginning w i t h Thales and certainly counting many of the most

eminent Presocratic philosophers among its number. W h i l e one m i g h t
point to Anaximander's rings and Pythagorean speculation concerning the
h a r m o n y of the heavens, perhaps c u l m i n a t i n g i n the cosmology of Philo­
laus, there is no really effective u n i o n of cosmology and astronomy. Cosmo­
logical models are not so far advanced as to take into account any b u t the
most elementary observations. The heavens go around the earth, and
there is a distinction between the planets and the fixed stars, b u t beyond
t h a t there is no attempt to make cosmological models f i t w i t h the results
of observational astronomy i n any detailed manner.
I t is perhaps w i t h Plato and Eudoxus t h a t we find the first wedding of
these traditions, something t h a t is clearly hugely i m p o r t a n t for the history
of astronomy and cosmology. One m i g h t object to giving such a role to
Plato, arguing t h a t Eudoxus really is the critical character. A standard
objection here, t h a t on the evidence of Republic V I I Plato rejected or
denigrated empirical astronomy, and so could not have brought about a
synthesis of the philosophical and observational traditions, we have al­
ready dealt w i t h i n Chapter 2. The Timaeus shows awareness of m a n y of
the empirical problems of astronomy. A weakened form of this objection
w o u l d be t h a t Plato d i d no empirical w o r k himself. This m i g h t be met i n
two ways. F i r s t , one m i g h t accept t h a t Plato was not p r i m a r i l y an astrono­
mer, b u t argue t h a t he made a significant philosophical and theoretical
contribution. Secondly, one m i g h t argue t h a t d u r i n g his own career Plato
does make some significant advances i n astronomy. Clearly Plato is not an
154 Plato's Philosophy of Science
observational astronomer, b u t the Timaeus w o u l d seem to indicate t h a t he
was at least conversant w i t h the major astronomical phenomena.
There is of course a vexed question here of any influence between Plato
and Eudoxus. I f there was any influence, who influenced w h o m , on w h a t
subjects, and to w h a t extent are a l l open questions. There is also an issue
67

for the d a t i n g of the Timaeus associated w i t h Eudoxus' dates. Owen argues

t h a t there is no influence from Eudoxus on the Timaeus and t h a t this
implies an early d a t i n g . C i t i n g Apollodorus, Owen places Eudoxus'
68 69

death at 356-353, and states t h a t he must have produced his system w e l l

before this, as he left Athens for Cnidus some years before his death. The 70

effect of a l l this is to push back the date of composition of the Timaeus, i n

Owen's o p i n i o n to perhaps before 369 a n d t h e composition of the
Theaetetus. I reject the dating of Eudoxus based on the notoriously
unreliable testimony of Apollodorus w h i c h w o u l d have his death i n 356-
353. Gisinger places Eudoxus' death i n 342, and both de S a n t i l l a n a
71 72 73

and Laserre give 337, after Plato's death i n 348. There is also the problem
74

of accommodating the evidence of Pliny, who talks of Eudoxus d a t i n g

75

Zoroaster as l i v i n g 6,000 years before the death of Plato, i m p l y i n g t h a t

Eudoxus outlived h i m . W h i l e i t may be 'commonly agreed' t h a t Eudoxus
came to the Academy i n 368, the alternative datings allow for Eudoxus
76

to have formulated his full theory after Plato composed the astronomical
passages i n the Laws or even after his death i n 348. I t may w e l l be t h a t
77

Eudoxus came to the Academy i n 368, b u t w h e n Eudoxus began his w o r k

on concentric sphere astronomy and w h e n he finished i t are b o t h quite
open questions, and we need to recognise t h a t i t is quite possible t h a t Plato
only k n e w of i t i n a developmental stage. There is n o t h i n g t h a t compels us
to suppose any influence was i n one direction only, and i t may w e l l be t h a t
Eudoxus d i d not complete his development of his homocentric sphere
system u n t i l after Plato's death. I t is possible t h a t Eudoxus, armed w i t h
his homocentric spheres, convinced Plato t h a t the motions of the heavens
could be resolved into regular circular motions and this precipitated
Plato's move between the Republic and the Timaeus. Alternatively, Plato's
prototype of the Timaeus may have been the spark w h i c h led to Eudoxus'
w o r k on homocentric spheres. I n the absence of any further evidence, we
have to go w i t h the existing texts. So Plato gives us the first extant account
of a p a t h for the planets offset from the daily motion of the fixed stars, and
t h a t is a significant step forward i n astronomy. Owen argues t h a t Plato's
use of the t e r m 'contrary power' i n relation to M e r c u r y and Venus means
t h a t he was not conversant w i t h Eudoxus' system at the t i m e of w r i t i n g
the Timaeus. Even i f this is so, Eudoxus has plenty of t i m e yet to formulate
his system. Owen is inclined to find some Eudoxan influence i n a move
78

from i r r e g u l a r to regular celestial motion between the Timaeus and the

Laws. 79
However, I have argued at length t h a t this change occurs between
the Republic and the Timaeus. F r o m the evidence t h a t we have, i t is not
clear whether or how Eudoxus influenced Plato, or indeed whether or how
 5. Plato and the Development of Greek Astronomy 155
Plato influenced Eudoxus. I f his influence is as Owen suggests, t h e n i t was
certainly ipre-Timaeus. Even i f he d i d not influence the composition of the
Timaeus, the doubts concerning his dates mean t h a t we cannot draw any
f i r m conclusions about the date of this work.

IX. Plato and Eudoxus

One way of a t t e m p t i n g to break any t i g h t l i n k between Plato and Eudoxus
is to argue t h a t while Plato was a realist, or engaged i n metaphysical
speculation about the nature of the heavens, Eudoxus was an i n s t r u m e n ­
talist, merely interested i n finding a mathematical scheme t h a t w o u l d f i t
the motions of the heavens. Thus one can find Dreyer commenting:

As Eudoxus made no attempt to connect the movements of the various

groups of spheres with each other, it seems probable that he only regarded
them as geometrical constructions suitable for computing the apparent
paths of the planets. 80

So too H e a t h remarks:

It would appear that he did not give his spheres any substance or mechanical
connection; the whole system was a purely geometrical hypothesis, or set of
theoretical constructions calculated to represent the apparent paths of the
planets and enable them to be computed. 81

As W r i g h t has argued, typically Eudoxus is thought to have been engaged

82

i n 'geometry' rather t h a n 'physics' because he gives no account of the

substance of the homocentric spheres and no account of how they m i g h t be
interconnected. W r i g h t suggests that, relative to other considerations
w h i c h I shall look at i n a moment, this evidence is relatively weak. I w o u l d
go further and say t h a t this evidence does not discriminate between
i n s t r u m e n t a l i s t and possible realist positions at all. One does not have to
be an A r i s t o t e l i a n realist, w i t h spheres of aether, to be a realist on this
matter. I f one considers the planets to be intelligent entities, t h e n there is
no need to give a n y t h i n g more t h a n the appropriate mathematics for each
planet. As intelligent Platonic entities, they w i l l of course be following the
best sort of motion so n a t u r a l l y they w i l l be executing (combinations of)
regular circular motions.
Here i t is i m p o r t a n t to recognise the nature of how the mathematics are
presented. There are sufficient parameters to determine fully the motions
of a l l the planets. This is considerably beyond w h a t an i n s t r u m e n t a l i s t
needs to give. I f the sole criterion were t h a t of calculational simplicity, t h e n
Eudoxus could have the earth revolve and so remove one sphere from each
of the planets and t h a t of the fixed stars, resulting i n a considerable
simplification of the system. Or again for simplicity, the motion of the fixed
stars could be considered to be common to a l l the planets, thus saving
156 Plato's Philosophy of Science
seven spheres. This is not the path t h a t Eudoxus takes, however. He
83

gives a full description of the motions for each of the planets. I f Eudoxus
is a Platonic realist on this matter, each planetary intelligence w i l l need
to be cognisant of all the motions t h a t i t needs to make, though this is not
required simply for the calculation of the motions of the heavens. I f each
planetary intelligence has enough information to govern its own motions,
t h e n the question of the interaction of the spheres becomes redundant.
There are t w o senses to this, how the spheres for each planet interact to
produce t h a t planet's movement, and how the systems for each planet
m i g h t interact w i t h one another. There is no need for the interaction of the
spheres for one planet. The planet knows w h a t motion i t ought to execute,
as a combination of circular motions. There is no interaction between the
systems, and none is needed. Each planet knows w h a t i t is doing, and its
motion is independent of all the other planets. I f Eudoxus was a Platonic
realist, t h e n he has no need to specify the substance of the homocentric
spheres, nor has he any need to specify how the spheres interact.
So the fact t h a t he does neither is equally compatible w i t h Platonic
realism as i t is w i t h an i n s t r u m e n t a l i s t approach. This, of course, is not to
say t h a t on these grounds Eudoxus was a realist i n the Platonic manner.
The i n s t r u m e n t a l i s t interpretation here though, as Lloyd has remarked,
is a questionable argument from silence, and here I hope to have shown
t h a t t h a t silence can j u s t as well be taken i n another way. There is no other
reason to suppose t h a t Eudoxus was an instrumentalist. As W r i g h t ar­
gues, i f Eudoxus were only interested i n the representation of the
phenomena and i n prediction t h e n the accuracy of the system would have
h a d paramount importance, perhaps w i t h the simplicity of calculation a
secondary consideration. I f one was looking merely for an i n s t r u m e n t a l
solution i n astronomy, I would suggest t h a t one would not be i n the least
bit interested i n the models produced by either Plato or Eudoxus. This is
because they are simply not as good at representing the phenomena as
other methods open to the Greeks at the time, or even as good as the
Babylonian methods of extrapolation from the data. Eudoxus' system, as
we have seen, h a d significant drawbacks, and i t is h a r d to believe t h a t he
was not aware of at least some of these. As W r i g h t puts i t :

From a purely predictive point of view there is no reason at all to avoid

openly ad hoc modifications of that model, la Ptolemy, for the sake of bringing
predictions closer to observations. For instance, standard moves within the
project should have included trying non-uniform or eccentric motion, and
centres other than the earth. But there is no trace of this sort of thing in the
records of the enterprise. 84

I t is difficult t h e n to see Eudoxus as an instrumentalist, and so he may

have been influenced by Plato.
 5. Plato and the Development of Greek Astronomy 157

X . Meno's p a r a d o x a n d a s t r o n o m y
Let us r e t u r n to the question of Meno's paradox and its resolution. I n
previous chapters we have seen the use of a solution to Meno's paradox i n
relation to the shape of the u l t i m a t e particles. W h a t is the problem t h a t
Plato faces i n astronomy? There is certainly another underdetermination
problem here. We could adopt any number of hypotheses w h i c h save the
phenomena of the motions of the heavens. Clearly one can formulate
infinitely m a n y such hypotheses i f there are no constraints on these
hypotheses other t h a n f i t t i n g the data. So where does one start i n this
indefinite multitude? I n line w i t h Plato's solution to underdetermination,
we ought to start w i t h a healthy hypothesis. So neither the Platonic
challenge nor the Timaeus (nor indeed the Laws and Epinomis) w i l l allow
any hypothesis w h i c h does not postulate regular, circular and ordered
motion. Effectively Plato and Eudoxus do 'some s t r a i g h t ' l i n e drawing, only
this is now done i n three dimensions and for moving objects, and move
directly to the simplest, most elegant mathematical generation. As we saw i n
the previous section, i n fact neither Plato nor Eudoxus can 'save the phenom­
ena' as well as other calculational devices that were available to the Greeks.
On this matter, Wright makes a comment w i t h which I would wholeheartedly
agree: Tf prediction was his primary aim, Eudoxus should have given up on
his spheres early i n his search.' One strong suggestion as to w h y he did not
85

must be that he was not an instrumentalist and considered homocentric

spheres to be a very healthy hypothesis i n the Platonic manner.
There is another aspect of Plato's astronomy w h i c h is perhaps under­
appreciated - the drive for a unified account of the heavens. Previous views
had been either been too simplistic i n t h e i r u n i t y and so had failed to
provide a scheme w h i c h could hope to represent the celestial motions
properly (so Philolaus, Anaximander), or, while saving the phenomena,
had been decidedly fragmentary and ad hoc and so h a d not offered a
unified conception of the heavens (so the parapegmata t r a d i t i o n , the
Babylonians). Plato is the first to recognise the complexity of the heavens
and offer (at least the prospect of) a unified theory of the motions of the
heavens. W h i l e Plato's prototype model of the Timaeus falls short of this
aim, an empirically adequate and unified picture of the heavens is an
i m p o r t a n t a i m of the following astronomical t r a d i t i o n . This brings out
another aspect of the solution to Meno's paradox and to the Platonic
challenge. Part of being a healthy hypothesis is h a v i n g the ability or
likelihood to be able to cohere w i t h other hypotheses. I n relation to this I
would stress again t h a t the Platonic challenge asks for motions w h i c h are
regular, circular and ordered. 86

I n the history of science, some ideas become so commonplace we are apt

to forget t h e i r significance, and so t h e i r provenance. I t appears t h a t those
ideas have always been there, because t h e i r contraries are so strange to
us. Such is the case, I suggest, w i t h the r e g u l a r i t y and order of celestial
158 Plato's Philosophy of Science
motion, and a unified picture of the heavens. Since Plato these have been
the basic assumptions of astronomy. We should not forget t h a t was not
necessarily the case before Plato, not only w i t h the Babylonians who
referred to the planets as 'sheep' (bibbu) because of t h e i r w a n d e r i n g
courses, and h a d fragmentary methods of depicting the motions of the
heavens, b u t so too w i t h some Presocratics, and indeed as I have argued
w i t h earlier Plato. As we shall see i n the next chapter, while Plato
considers the heavenly bodies to be gods, he steers very firmly away from
the idea t h a t these gods act i n any i n t e r m i t t e n t manner or t h a t they do
a n y t h i n g other t h a n w h a t is good, such t h a t they act more like laws of
nature and less like the gods of classical mythology.
I t is clear from the nature of Plato's solution of Meno's paradox, i n
combination w i t h his views on the best sort of motion, w h a t the healthiest
hypothesis concerning the heavens would be. T h a t Plato is answering
Meno's paradox/underdetermination questions i n the Timaeus is, I t h i n k ,
clear from his treatment of the questions of the number of cosmoi, the
shape and size of atoms and the number of species. I t would t h e n be very
surprising i f Plato were to deviate from regular circular motions for the
heavens w i t h o u t very good cause. This is a good reason for adopting an
i n t e r p r e t a t i o n of the Timaeus w h i c h holds on strongly to regular, circular
and orderly motion b u t allows Plato to acknowledge some of the astro­
nomical difficulties associated w i t h an analysis solely i n terms of regular
circular motion, and set a problem for solution i n the Platonic imperative.
So we should not take the Timaeus to be literal and finished cosmology, but a
prototype attempting to instantiate a solution to Meno's paradox. This is w h y
the contrary power/self-motion would be an unhealthy hypothesis, and so
would seem to be a highly unlikely interpretation of what Plato means at
Timaeus 38d4. Again, contra Vlastos, what inspires Plato's advocacy of
regular, circular ordered motion is not a 'metaphysical fairy tale'. I t answers
to real epistemological problems of how to go about doing astronomy.
One final comment on Plato and astronomy. One of the general weak­
nesses of ancient science is t h a t many of its practitioners presented t h e i r
w o r k as finished and dogmatic systems. As suggested i n Chapter 1, i t is a
modern rather t h a n an ancient response to say 'We do not know yet, there
are problems, b u t we hope to find an answer along the following lines'. One
very i m p o r t a n t exception to this may be Plato and the t r a d i t i o n he begins
i n astronomy. The weaknesses of the astronomical model of the Timaeus,
87

w h i c h are considerable and greater t h a n is generally recognised, are offset

by the methodological and cosmological insights, and the fact t h a t the
model is (and is intended to be) the prototype for a whole new t r a d i t i o n of
astronomy and cosmology. I t is overwhelmingly likely t h a t Plato recog­
nised at least some of the defects of the Timaeus model, j u s t as Eudoxus
and Callippus would have been aware of the problems w i t h t h e i r own
models.
 6

Plato and the Development of

Greek Cosmology

One of the most remarkable aspects of Presocratic thought is the r a p i d

increase i n the sophistication of its cosmological models. F r o m Thales and
the e a r t h floating on water t h r o u g h to the first complex cosmology of the
Timaeus, there is a steady b u t swift overcoming of conceptual obstacles.
W h y t h a t should be so is beyond the scope of this book, t h o u g h I w o u l d look
to the Greek formulation of the idea of a cosmos as an ordered e n t i t y
comprehensible to h u m a n reason, and the rejection of mythological and
supernatural explanations i n favour of philosophy and debate. Though 1

not as spectacular, there is also an advance i n observational astronomy,

and u l t i m a t e l y a u n i o n of this t r a d i t i o n w i t h t h a t of cosmological specula­
tion. The task of this chapter w i l l be to assess Plato's role i n the develop­
ment of Greek cosmology. F i r s t , I argue t h a t he takes up the Presocratic
concerns w i t h the position and stability of the earth. Plato can t h e n be seen
to consider the application of these concerns about the earth to the cosmos
as a whole, along w i t h a new concern, the i n t e g r i t y of the earth and the
cosmos. As Plato's thought develops, we can see some more sophisticated
answers being produced on the nature of the cosmos. These developments
m a r k an i m p o r t a n t advance i n Greek cosmological t h i n k i n g . One aspect of
Plato's cosmology w h i c h has received severe criticism, however, is his
description of the cosmos and the celestial bodies as animate, intelligent,
ensouled deities. Pedersen comments:

I n many ways, the Timaios reflects an animistic and anthropomorphic

philosophy foreign to the Ionian philosophers' fundamental attitude, al­
though there are a few similar ideas. 2

So too we can find B u r k e r t saying:

Plato thought it an inescapable conclusion that the orderly motion of the

stars is due to their having souls; it is a voluntary, chosen order. Here
sophisticated Greek science harks back to the pre-scientific way of thinking
and comes to rest in i t . 3

A t first glance Plato w o u l d indeed seem to be m a k i n g a backward step

160 Plato's Philosophy of Science
here, r e t u r n i n g to precisely the sort of animistic or supernatural explana­
tions the early Greeks h a d done so well to reject. M a t t e r s are not quite so
simple though, and i n fact Plato is very careful about w h a t he attributes
to the cosmos and celestial bodies i n an attempt to explain w h y t h e i r
behaviour is regular. He has to do this w i t h very few resources, and his
attempt is interesting, ingenious, and, i n fact, very helpful for the future
progress of Greek astronomy and cosmology.

I . P l a t o a n d the P r e s o c r a t i c s
I f we accept Aristotle's report, Thales produced the first cosmology:

Others say that the earth rests on water. For this is the most ancient account
we have received, which is said to have been given by Thales of Miletus, that
it stays in place through floating like a log or some other such thing (for none
of these rest by their nature on air, but do on water), as though the same
argument did not apply to the water under the earth as to the earth itself. 4

Quite possibly i n reply to the sort of difficulty Aristotle points out come a
new generation of cosmologies w h i c h rely on air to support the e a r t h . A i r
5

does not fall and so is perhaps more suitable as a supporting m e d i u m t h a n

water. Aristotle also tells us:

Anaximenes, Anaxagoras and Democritus say that i t is the flatness (of the
earth) that is the reason why i t stays still, for i t does not cut the air below
but covers i t as flat bodies are able to do. 6

Thales and Anaximenes are prime examples of w h a t Furley w o u l d call a

parallel as opposed to a centrifocal conception of the cosmos. I n a parallel
7

cosmology, heavy objects fall i n one direction, and do so i n parallel lines.

Thus we have the problem of w h y the earth does not fall i n a s i m i l a r
manner. One can contrast this w i t h the centrifocal cosmos. Aristotle is the
8

classic example here as everything i n n a t u r a l motion moves towards (or

moves away from) the centre of the cosmos. The cosmos here has no
direction of fall b u t does have a central point. Plato's cosmology is undoubt­
edly a precursor to Aristotle's, though Plato lacks the strictly centri-focal
dynamics w h i c h make the Aristotelian cosmos classically centri-focal. 9

There are of course precursors to Plato on this m a t t e r too. One m i g h t look

to Parmenides, especially Fragment 8 42-5, on the w e l l rounded sphere
balanced i n a l l directions, though how the meditation on being of Frag­
ment 8 relates to the more explicit cosmology of Fragments 10, 12 and
Aetius I I , 7, 1 is not at a l l clear. Alternatively one m i g h t look to the
Pythagorean cosmology of Philolaus w i t h its concentric orbits around the
central f i r e , though again both lack centri-focal dynamics as far as we
10

know.
There is some controversy over the nature of Anaximander's cosmology.
 6. Plato and the Development of Greek Cosmology 161
The question is whether he gives a full e q u i l i b r i u m argument for the
stability of the earth, w h i c h would i m p l y a central earth and no direction
for the cosmos, thus giving at least a precursor of a centrifocal cosmos, or
whether he only gives a p a r t i a l e q u i l i b r i u m account. I n the l a t t e r case the
e a r t h is supported by air, such t h a t we are back w i t h the parallel cosmos,
but the flat top and bottom of the drum-shaped earth are kept at r i g h t
angles to the line of fall of the cosmos by the e q u i l i b r i u m of the d r u m . The
outcome of t h i s debate is i m p o r t a n t for how we value the full e q u i l i b r i u m
argument for the stability of the earth t h a t we find i n the Phaedo. Aristotle
tells us:

There are some, such as Anaximander among the ancients, who say that (the
earth) rests on account of its likeness. I t is fitting that what is established in
the centre and has equal relations to the extremes should not move up, down
or to the side. I t is not possible for it to move in opposite directions at the
same time, and so necessarily i t remains still. 11

This w o u l d appear to a t t r i b u t e a full e q u i l i b r i u m theory to Anaximander.

The key piece of counter evidence though, as cited by Furley, comes from
Simplicius:

Anaximander thought that the earth remains in place because of the air that
supports it and because of its equilibrium and uniformity. 12

I side w i t h Panchenko and Schofield against Furley here i n giving greater

weight to the position Aristotle describes, especially as i t is supported by
13

H i p p o l y t u s . A t Phaedo 108e Plato has Socrates say:

14

I am now persuaded that, firstly, if the earth is round and in the centre of the
heavens, then i t requires neither air to prevent i t falling (pesein) nor any
other necessitation of this sort, but the uniformity of the heaven itself in
every way and the equipoise of the earth itself is sufficient to restrain it. For
something which is equipoised and is placed in the middle of something
homogenous cannot yield to being moved aside in any way (oudamose
klithenai), but will remain steadfast.

There is no doubt t h a t this is a full e q u i l i b r i u m argument, and Plato

almost certainly has a spherical earth i n m i n d here. I t is possible t h a t
15

there is another layer of sophistication here, one t h a t is only required once

one gives up the parallel cosmos. Plato usespiptein and klinein to describe
w h a t the e a r t h w i l l not do. Piptein has connotations similar to the E n g l i s h
verb to fall. I t can be used of a city, of a government, or of a soldier i n battle,
and so can carry a sense of downfall, demise or r u i n a t i o n . So Plato's 16

worries here may t h e n go beyond w h y the earth does not drop, to w h y the
e a r t h doesn't fall apart, i f n o t h i n g holds i t together. W i t h a parallel cosmos,
the i n t e g r i t y of the earth, at least i n one direction, is a simple matter.
Pieces of e a r t h would move downwards, b u t are supported either by water
162 Plato's Philosophy of Science
or air and so cohere together. Even i f i n t e g r i t y is not a concern here (and
admittedly this evidence is rather t h i n ) , i t is i n the Republic.
The Phaedo gives us the conception of the free floating and immobile
earth. The m y t h of E r t h e n makes a significant conceptual leap to begin to
t a l k of the i n t e g r i t y of the cosmos and how the cosmos m i g h t be supported.
Here the cosmos needs external bonds, ta hupozomata (616c2), like the
brace t h a t passes under or around the h u l l of a t r i r e m e , rather t h a n being
self-sufficient. Or i f ta hupozomata are i n t e r n a l to the h u l l , t h e n the
17

cosmos needs i n t e r n a l bracing to hold i t together. I t is notable t h a t here

18

the cosmos is not referred to as a l i v i n g entity and does need some bracing,
w h i l e i n the Politicus, Timaeus and later works the cosmos is a l i v i n g e n t i t y
( m o r t a l i n the Politicus, i m m o r t a l i n the Timaeus) and needs no such
b r a c i n g . Another i m p o r t a n t contrast we saw i n Chapter 4. Where the
19

cosmos t u r n s on a pivot (Republic, Politicus) and does not have an i m m o r ­

t a l soul (no soul, Republic, m o r t a l soul, Politicus) there is i r r e g u l a r
celestial motion and degeneration. The planets too have no souls and
require the daughters of necessity to move them.
The Timaeus t h e n has some significant developments on these views.
The cosmos is a l l there is, so i t can have no external bracing nor can i t rest
on a pivot. W h i l e I have so far stressed a point r e l a t i n g to celestial motion
here, the move from a supported cosmos to a free one w h i c h exhausts a l l
there is looks to be a significant conceptual leap, analogous to the move
from a supported to a free-standing earth i n Presocratic thought. The
i n t e g r i t y of the cosmos (and indeed of a l l the celestial bodies) is explained
by t h e i r h a v i n g souls, so no bracing is required, and as these souls are
i m m o r t a l , there is no degeneration. So too the motions of the cosmos and
the celestial bodies are generated by t h e i r souls. W h i l e there is a sense i n
w h i c h we can easily refer to top and bottom of the cosmos i n the m y t h of
E r (as the cosmos is on the knees of necessity) Timaeus 62c ff. makes i t
very clear t h a t such terms as above and below are not suitable. One m u s t
make due allowance for the highly metaphorical language of the m y t h of
E r and the fact t h a t cosmology is not Plato's p r i m a r y concern there, b u t
s t i l l there are i m p o r t a n t differences i n the structure of the cosmos from the
Republic to the Timaeus.

I I . Plato a n d geocentrism
Plato gives us a cosmos w i t h a central, stable and static earth. T h a t the
e a r t h is static, at least i n the usual sense, I take to be the upshot of the
debate about the rotation of the earth i n the Timaeus. The key passage
20

is at Timaeus 40b8-cl:

Earth, our nurse, is packed around (illomenen) the axis which stretches
through the universe, was contrived to be guardian and craftsman (demiour-
 6. Plato and the Development of Greek Cosmology 163
gon) of night and day, first and eldest of the gods who have come into being
with the heavens. 21

The i n i t i a l problem here is t h a t illomenen (or eillomenon) can mean

'packed around' or 'revolving around'. I f Plato h a d a proper motion of the
22

e a r t h i n m i n d , there are several words w h i c h he could have used to express

himself more clearly. W h a t the earth is required to do to produce a
23

coherent cosmology m i g h t be inferred from other passages. Timaeus 3 6 e l -

3 tells us t h a t soul was:

woven everywhere from the centre to the extremity of the heavens, and
enveloped in a circle from the outside, revolving on itself.

One would expect t h e n t h a t the earth would revolve w i t h this and so there
would be no relative motion between the earth and the fixed stars, i f no
other factors were involved. However, Timaeus 39cl-2 quite reasonably
says:

So for these reasons night and day came to be, the period of the single and
most intelligent revolution.

For t h a t to be the case, there must be relative motion between the earth
and the movement of the same. One should also note t h a t the earth is
described as the 'first and eldest of the gods w h i c h have come into exist­
ence w i t h the heavens' (Timaeus 40c2-3). The other gods, i.e. the stars and
planets, a l l have motions befitting t h e i r intelligences, so perhaps the earth
ought to as well. I f i t does have motion, this ought to be a simple motion
befitting the eldest. This counts heavily against the view of B u r n e t and
Taylor t h a t the earth moves up and down on the central axis of the
cosmos. As Taylor recognises, such a motion could be used to explain the
24

frequency of eclipses. However, this motion would be a r b i t r a r y and

25

non-circular, and h a r d l y fitting for the e a r t h . I t could not account for

26

deviation or occlusions i n general, as moving the earth up or down would

make the sun, moon and a l l the planets appear to deviate the same side of
the ecliptic at once. I do not see t h a t we can support the idea of a motion
up and down an axis, nor w o u l d such a motion be consistent w i t h views
expressed i n the rest of the Timaeus.
The standard resolution of this matjter is t h a t the earth is considered to
have a motion equal and opposite to t h a t of the same and so is effectively
static. This is a simple motion (regular circular motion i n one place) and
27

w o u l d produce the required contrast w i t h the same to produce n i g h t and

day. W h i l e the effect on the earth is clear enough, whether one calls this a
motion of the earth is another matter. Aristotle comments:

There are also those who hold that the earth is in the centre and is wound
164 Plato's Philosophy of Science
around, i.e. moved (illesthai kai kineisthai) about the axis through of the
universe, as is written in the Timaeus. 28

I take the report by Plutarch, who reports Theophrastus, t h a t Plato i n his

old age regretted giving the earth the central place i n the cosmos, and
thought i t ought to go to some w o r t h i e r body, to be unreliable. There is
29

n o t h i n g i n Plato's later w r i t i n g s to suggest this, nor is there any evidence

from his contemporaries to support P l u t a r c h . Plato t h e n does n o t h i n g to
30

offset the geocentric t r e n d of Greek cosmology, and l i t t l e to offset the

geostatic trend. H o w should we view that? First, we m i g h t consider w h a t
the alternatives here were. According to A r i s t o t l e , a l l those who thought
31

the cosmos to be unique and finite believed t h a t the earth was at the centre
of the cosmos. O f course the atomists also believe t h a t our earth is at the
centre of our cosmos, even i f there are indefinitely many other cosmoi. The
only dissenters here were some of the Pythagoreans.
W h y d i d the Greeks accept geocentrism? I f one is beginning cosmology,
the most n a t u r a l assumption to make is t h a t the e a r t h is static and
central. There are generally perceptible consequences of being i n motion,
and as the earth shows none of these, i t is reasonable to conclude t h a t i t is
static and i t is the heavens w h i c h move. I f we look at the heavens over
32

the course of a night, the stars appear to circle around us and so i t is

reasonable to assume t h a t the earth is central. So one m i g h t argue t h a t
w h i l e the i n i t i a l impulse to geocentrism is understandable, the Greeks
never developed a n y t h i n g more sophisticated. Perhaps this is because of a
religious prejudice on t h e i r part, most explicit w h e n the Stoic Cleanthes
believed t h a t Aristarchus should be charged w i t h i m p i e t y for suggesting
t h a t the earth moved. Perhaps this is because of an anthropocentric
33

prejudice, t h a t m a n should be at the centre of the cosmos, possibly derived

from theology or mythology. I t may be the case t h a t any or a l l of these
factors influenced the Greeks, b u t we should not allow this to obscure the
fact t h a t the Greeks had good arguments for geocentrism. The basic fact
34

t h a t there are no perceptible effects of the motion of the earth is a very

strong consideration, and remains so up u n t i l Galileo and the development
of a physics w h i c h can deal w i t h a mobile e a r t h . This is one reason w h y
35

Copernicus' heliocentric hypothesis was not accepted for over sixty years. 36

I I I . T h e G r e e k s and geocentrism
A further major argument i n favour of geocentrism i n the ancient w o r l d is
the question of parallax. I f the earth is i n motion, t h e n parallax effects,
due to the change of position of the observer should be evident. The Greeks
were aware of parallax effects, but could detect none due to the motion of
the e a r t h . This remained a problem beyond the Copernican revolution,
37

as stellar parallax was not detected u n t i l Bessell i n 1838. To explain away

the lack of parallax, heliocentrists had to argue for a m u c h larger cosmos
 6. Plato and the Development of Greek Cosmology 165

F i g . 41. The problem of parallax.

where parallax effects were too small to be measured by current i n s t r u ­

ments. Often such a cosmos would be implausibly large by the standards
of the t i m e . F i g . 4 1 illustrates the problem. I f the earth orbits the sun i n
38

one year, t h e n there is a significant difference i n its positions at six-month

intervals. This ought to be detectable relative to the fixed stars, w h i c h
should appear to be i n slightly different positions. I n fact this is difficult,
as the amount the earth moves is very small compared to the distance of
the stars, and parallax was not detected u n t i l 1838. A further difficulty
here lies i n choosing w h i c h star w i t h w h i c h to make parallax measure­
ments. Clearly i t is i m p o r t a n t to choose one relatively near, such as A l p h a
Centauri, where one is t a l k i n g i n terms of single figures of l i g h t years,
rather t h a n one t h a t is tens or hundreds of l i g h t years away, or even
further. Prior to the establishment of a reliable cosmological distance
ladder, this is not a straightforward t a s k . One consequence of this is t h a t
39

w i t h each improvement i n observational accuracy, advocates of a mobile

earth h a d to increase t h e i r estimates of the size of the cosmos such t h a t
parallax w o u l d be too small to be detected.
There is the further difficulty t h a t i f the earth is i n motion, w h y does
the moon follow it? G r a v i t y is not a possible answer here, and for Plato the
like-to-like principle cannot be invoked as the moon is a star and so is
made m a i n l y of fire, and so w i l l not be attracted to the earth. For Aristotle
the moon following a mobile earth is inconceivable. Would changing to
heliocentrism account for the phenomena better or allow for a simpler and
more elegant cosmology for the ancients? The answer on both counts is no.
166 Plato's Philosophy of Science
The phenomena can be met j u s t as w e l l w i t h a geocentric cosmology. There
is nothing, before the i n v e n t i o n of the telescope, w h i c h demands heliocen-
t r i s m . The cosmological model of Philolaus, apart from its observational
40

drawbacks, is not simpler t h a n Plato's models, and is actually a complex

model i n its special requirements. I f we look to later developments, t h e n i f
Aristarchus' model is to be able to save the appearances, t h e n the m o t i o n
of the e a r t h would have to be far more complex t h a n simply a circular orbit
around the sun and a daily rotation. The demands of the inequality of the
seasons and the precession of the equinoxes would make the motion of the
e a r t h extremely complex i f the phenomena were to be saved. I t should also
be noted t h a t Copernicus' system was not significantly simpler t h a n the
Ptolemaic m o d e l , and i t was not u n t i l Kepler's w o r k on elliptical orbits
41

sixty years later t h a t the Copernican hypothesis became viable i n this

respect.
Finally, there is a problem w h i c h i n m y view has been underrated i n
assessments of the ancients. This is the question of the i n t e g r i t y of the
earth. The problem is t h a t i f the i n t e g r i t y of the earth is based on its
position, t h e n t h a t i n t e g r i t y disappears as soon as the e a r t h is p u t i n
motion, or otherwise moved from t h a t position. Aristotle of course has a
42

very strong account of the i n t e g r i t y of the earth. The theory of n a t u r a l

place and n a t u r a l motion means t h a t earth and water w i l l move to the
centre of the cosmos unless hindered from doing so, and so the earth w i l l
be at the centre of the cosmos, and w i l l have a great deal of i n t e g r i t y and
coherence. However, i f the earth were i n motion i t would no longer have
this integrity, nor would heavy objects fall i n the usual manner. A r i s ­ 43

totle's account of the i n t e g r i t y of the earth and phenomena w h i c h we would

explain i n terms of gravity leads inexorably to geocentrism. So too Plato's
e q u i l i b r i u m account of the i n t e g r i t y of the earth i n the Phaedo w o u l d
require a central and static position for the earth, as indeed w o u l d A n a x i -
mander's. I f one is going to relate i n t e g r i t y to position, and i n the absence
of a conception of gravity this has to be an option, t h e n a central position
appears the most plausible. This i n t u r n leads almost inexorably to a
central and static earth. This is one reason w h y I argue t h a t the lack of
a n y t h i n g like a modern conception of gravity runs very deep i n Greek
cosmology. One of Copernicus' major difficulties i n m a k i n g heliocentrism
plausible lay i n accounting for the i n t e g r i t y of a mobile earth. He has to
give up the A r i s t o t e l i a n account t h a t went w i t h Ptolemaic cosmology, and
falls back on a rather implausible version of a like-to-like p r i n c i p l e . 44

I f we look to the Timaeus, t h e n Plato's cosmogony is strongly geocentric

due to the like-to-like principle. The subsequent dynamics w h i c h are
generated by the like-to-like principle are not as strongly centri-focal as
Aristotle's i n principle, as i f the earth were moved the dynamical centre
point w o u l d move too, b u t as there seems to be no reason for the earth to
move i n practice Plato's later view is nearly as strongly centri-focal as
Aristotle. I f the earth were moved for Plato, I suggest i t w o u l d r e t u r n to
 6. Plato and the Development of Greek Cosmology 167
the centre again, under the influence of the like-to-like principle and the
constrictive principle generated by the motion of the same, though t h a t
m i g h t be quite a slow process, u n l i k e i n Aristotle. B u t t h e n we r e t u r n to
the Phaedo principle - something equipoised i n something u n i f o r m has no
reason to move. T h a t principle i n alliance w i t h the cosmogony means the
e a r t h w i l l always be central for Plato. Perhaps this also encapsulates the
difference between modern and ancient attitudes on this matter. We ask
w h y the earth should be central and stable, the Greeks ask w h y i t would
not be. The Greeks, Plato included, h a d good reasons for t h e i r geocentrism.

IV. T h e a n i m a t e h e a v e n s
Plato considers the heavenly bodies to be gods, to be animate, to be
intelligent and to have souls, and so too the cosmos as a whole. Coming
after the m a t e r i a l i s m of the atomists, is this not a regressive move i n
Greek cosmology? After all, Plato is well aware of the views of the atomists
and Anaxagoras' theory t h a t the celestial bodies are hot stones. So w h y 45

does Plato adopt this position? As we saw i n Chapter 1, Plato has a very
low opinion of w h a t m a t t e r can do alone and unaided. Even i f m a t t e r is
properly formed (has i n t r i n s i c order), t h e n physical principles (such as the
like-to-like principle of Timaeus 62c-e) are not sufficient to explain how the
cosmos acquires or sustains order. For Plato, regular and orderly behav­
iour requires further explanation, and so the motions of the heavens
require explanation. Laws 967b puts i t like this:

Those who engaged in these matters accurately would not have been able to
use such wonderfully accurate calculations i f these entities did not have
souls.

I t is w o r t h looking closely at w h a t sort of entities Plato postulates here,

w h a t they do and w h a t they are supposed to explain. Timaeus 30b de­
scribes the cosmos i n these terms:

Thus in accordance with the likely account i t is necessary to say that this
living, ensouled intelligent cosmos has in truth come about through the
foresight of god.

The Timaeus states t h a t the cosmos does not have many of the things we
w o u l d usually associate w i t h an a n i m a l (or even, w i t h i n either a Pagan or
a C h r i s t i a n t r a d i t i o n , w i t h a god). F r o m Timaeus 33b onwards we are told
t h a t this 'animal' is perfectly spherical, has neither eyes nor ears as there
is n o t h i n g external to see or hear, there is no air for i t to breathe, no need
of any organ to receive food or to excrete the remains, as i t is entirely
self-sufficient and n o t h i n g comes i n or goes out. As i t needs neither hands
to defend itself nor feet to stand on, and has no need of legs or feet to propel
itself, i t has no limbs, and at 34b we are told i t is a god. To say the least,
168 Plato's Philosophy of Science
this is a somewhat strange animal, and certainly could not be considered
to be an anthropomorphic god, even i f i t does have intelligence and soul. 46

There may well be assorted criticisms of some Presocratics i m p l i c i t here

too, notably perhaps the Pythagoreans. Quoting Aristotle's lost On the
Pythagoreans, Stobaeus tells us:

The universe [according to the Pythagoreans] is unique, and from the infinite
it draws in time, breath and void, which distinguishes the places of separate
things. 47

So too the idea of a unique, self-sufficient and exhaustive cosmos cuts

against the views of the atomists. There are advantages t h a t come out of
this idea of the cosmos though. I f the cosmos is like an animal, t h e n we
have an analogue for the i n t e g r i t y and the i n t e r n a l organisation of the
cosmos, and the ongoing order of the cosmos. A typical ancient contrast
here is between the body w i t h life and m i n d and w h a t happens to the body
after death. W h a t does this god do though? A l l i t does is revolve uniformly
48

i n one place. A t Timaeus 34a we are told t h a t the cosmos was given:

A motion proper to its body, that of the seven motions which is best suited to
reason and intelligence. Therefore he made i t move in a circle, revolving of
itself uniformly and in the same place, and he took from i t all trace of the
other six motions and kept it free from their wanderings 4 9

Here I w o u l d emphasise t h a t this is entirely regular circular m o t i o n w i t h

no deviations. This is so for the fixed stars and the planets as well. The
stars are spherical and intelligent like the cosmos (40a) and are divine and
l i v i n g creatures (40b). As w i t h the cosmos, they have motions befitting
t h e i r intelligence, and so we are told at Timaeus 40a:

To each of these he gave two motions, one being uniform and in the same
place, always thinking the same thoughts concerning the same things, the
other being a forward motion obeying the revolution of the same and similar.
With regard to the other five motions, they were motionless and still, in order
that each might attain the greatest possible perfection.

The planets are brought into being i n order to generate time. We are t h e n
t o l d at Timaeus 38e:

When each of the bodies required to help in the production of time had been
bought into its proper motion, as generated living creatures having their
bodies bound with living bonds and having learnt their assigned duties, they
began to move according to the motion of the different.

As I have argued, we have no reason to suppose t h a t the planets deviate

i n any way. They follow the same and the different, where the fixed stars
follow the same and the cosmos rotates i n one place. The essential point
 6. Plato and the Development of Greek Cosmology 169
here t h e n is t h a t the celestial bodies have no freedom of action (or no desire
to deviate from regular circular motion). They have the intelligence to
carry out t h e i r assigned duties, and not to do a n y t h i n g else. I t is t h e i r
intelligence w h i c h explains their regular and orderly behaviour. T h a t they
move, and do so w i t h o u t any external compulsion, is explained by t h e i r
having souls. For Plato, the soul is a principle of motion. So at Laws 895c
we find:

Athenian Stranger: I f we should see that this motion (self-generating

motion) had come about in something composed of earth, water or fire,
whether separate or in combination, what should we say was the state of
that thing?
Cleinias: You are asking me is whether we should speak of this thing as
being alive?
AS: Yes.

Cleinias: Emphatically so.

So too at Laws 897b:

I f the whole course and motion of the heavens and everything that are in
them are similar in nature to the motion and revolution and calculations of
intelligence (nous), and work in a related manner, then clearly we must say
that the best soul has charge of the whole cosmos and leads it along these
selfsame courses. 51

A l t h o u g h the cosmos and the heavenly bodies have life, soul, d i v i n i t y and
intelligence, they have these i n a highly circumscribed and attenuated
manner, from a modern point of view. I f this is describable as vitalism, i t
52

is a highly depersonalised one where the attributes of animate beings t h a t

are required for the description of the cosmos have been carefully sorted
from those t h a t are n o t . This allows us to distinguish quite sharply
53

between Plato and the mythological and magical traditions. The key
matter here is t h a t these souls/gods are not capricious. They always act
for the best so w i l l always act i n the same manner. There is i n principle
n o t h i n g unpredictable or irregular about t h e i r behaviour. This is some­
54

t h i n g t h a t the Epinomis is very clear on at 982d ff. The stars have been
acting i n this manner for a very long t i m e , and do not ever decide to do
55

a n y t h i n g different. I n terms of theology, i t is significant t h a t Plato's god

always acts for the best, i n sharp distinction w i t h many previous Greek
conceptions of the gods.
Why does Plato not simply express this i n terms of physical laws, or j u s t
leave the behaviour of the heavens as a set of regularities? Plato's proto-
realism w i l l not let h i m rest w i t h the idea of bare regularities. The idea of
physical law though is something t h a t has been h a r d won and was not
something t h a t was simply available for use by the Greeks. W h y t h a t
should be so and w h y Plato answers i n the way he does I shall discuss i n
detail i n the next section. I n the meantime i t is w o r t h noting t h a t a typical
170 Plato's Philosophy of Science
seventeenth-, and indeed eighteenth-century view is t h a t i n some manner
god underwrites the laws of nature and is responsible for the regular
behaviour of matter. Being a C h r i s t i a n god, he always acts i n a good
56

manner, and so we see a regular and well arranged world. Whether we can
know the laws w h i c h god has i n m i n d or are l i m i t e d to knowledge of
regularities depended on the nature of god (the v o l u n t a r i s m debate) and
our r e l a t i o n to h i m . Thus Berkeley urged an early form of i n s t r u m e n t a l -
ism, w h i l e N e w t o n and Boyle were more optimistic. This was i m p o r t a n t
because the mechanical philosophy, i n l i m i t i n g matter to extension and
motion, h a d removed a l l the scholastic and occult qualities w h i c h m i g h t be
used to explain w h y bodies behave i n an orderly manner. Even those who
dissented from the mechanical philosophy, such as the Cambridge Platon-
ists, were keen to emphasise regular behaviour. Thus we find More
57

saying:

I affirm with Descartes, that nothing affects our sense but such variations of
matter as are made by differences of motion, figure, position of parts, etc.,
but I dissent from him in this, that I hold it is not mere and pure mechanical
motion that causes all the sensible modifications of matter, but that many
times the immediate director therof is the spirit of nature, one and the same
everywhere, and acting always alike upon the like occasions, as a clear
minded man, and of a solid judgement, gives always the same verdict in the
same circumstances.

Plato is not alone t h e n i n the history of scientific though i n using divine

entities to explain regular behaviour. The key question is whether these
58

entities are law abiding or not, and for Plato they most certainly are.
Lennox argues t h a t Plato's teleology is u n n a t u r a l . I n the sense t h a t
59

teleology is imposed on the w o r l d by the demiurge, and so is not a n a t u r a l ,

original feature of the world, t h a t is so. I t is also i m p o r t a n t to recognise
t h a t Plato's demiurge is a god subject to n a t u r a l law and regular behav­
iour, as are his demi-gods, and so there is a sharp contrast w i t h the caprice
of the gods of Greek m y t h . 60

V. Nomological r e a l i s m
Plato's views on the explanation of regularities are w o r t h t h i n k i n g of i n
terms of the cave allegory. The prisoners i n the cave are chained facing the
w a l l , and see only the shadows of the objects being carried along the road
behind them. The state of the prisoners m i g h t be described as H u m e a n . 61

They may associate the voices they hear w i t h the passing shadows w h i c h
they see, as ok Republic 515b:

I f the prison had an echo from the opposite wall, then whenever a passer by
spoke, do you think they would they suppose anything other than that the
voice belonged to the shadow?
 6. Plato and the Development of Greek Cosmology 171
They cannot though make any further investigation. So too they can
associate one object or a sequence of objects w i t h w h a t objects w i l l follow,
b u t again can investigate no further t h a n this. A t Republic 516cd Socrates
says t h a t perhaps there was:

Some honour and praise among them and prizes for seeing most sharply the
passing things, and being best able to remember which of them came before,
which after and which together, and so had the greatest power in predicting
which of these was to come.

As emphasised i n Chapter 2, however, for Plato we both can escape and

ought to escape from the cave. Clearly there are many other epistemologi­
cal issues w h i c h Plato wishes to focus on here, b u t one point m i g h t be p u t
l i k e this. I f the w o r l d were such t h a t there were only regularities, or we
only had access to regularities, t h e n we would not be able to escape the
cave. I m p o r t a n t here is Plato's distinction between empeiria, experience or
practice, and techne, art or skill. As we saw i n Chapter 2, often w h e n Plato
contrasts the two, empeiria is coupled w i t h tribe, l i t e r a l l y r u b b i n g or
threshing, though i n these contexts i t carries a sense of learning by routine
or perhaps t r i a l and error, and may w e l l carry certain pejorative connota­
t i o n s . Plato supposes t h a t there is always a techne. However, i f there
62

were not something more t h a n regularities, or as Plato puts i t 'what

usually happens', t h e n there would be nothing to base the techne on. So we
find at Gorgias 500e4:

Cookery seems to me to be not a skill (techne) but an empirical knack

(empeiria), while medicine investigates the nature of those it treats and the
reason for what it does, and has some account to give of each of these things,
does medicine. But the other, concerned with pleasure, at which its whole
treatment is aimed, works towards this in an entirely unskilful (atechnos)
manner, not having investigated the nature of nor the reason for pleasure,
unreasoning (alogos) in every way, and so, one might say, making no distinc­
tions, and by routine and knack (tribe kai empeiria) i t preserves only a
memory of what usually happens, and that is how it is able to provide
pleasures.

So too at Phaedrus 270b, this t i m e the contrast being medicine and

rhetoric:

In both cases i t is necessary to analyse a nature, of the body on the one hand,
and of the soul on the other, i f you intend to have not only practice and
experience (me tribe monon kai empeiria), but skill (techne), when attempt­
ing to bring about health and bodily strength by means of medicine and diet,
or to instil will-power and virtue by means of discourses and rules for the
conduct of life.

I do not w i s h to a t t r i b u t e to Plato an understanding of the modern

172 Plato's Philosophy of Science
distinction between regularities and laws, nor an understanding of the
modern conception of physical law. He does though require something
beyond regularities, and as I have argued for other issues, this narrows
the gap between Plato and some modern positions. Vlastos may w e l l
believe t h a t the laws of nature are 'radical contingencies ... for w h i c h no
further reason can be given', and so he sees a need to excuse Plato by
63

commenting t h a t there was 'no H u m e i n Plato's past'. There are m a n y64

modern philosophers of science though who would disagree w i t h H u m e ,

and object to the dissolution of the 'causal into the casual', as, I suspect,
w o u l d P l a t o . Vlastos admits there are problems of supporting counter-
65

factuals on a regularity theory, but there are also other serious difficul­
66

ties. 67

1. There is the problem of vacuous laws. I f laws are simply t r u e

universal generalisations, any universal generalisation (however vacuous
or silly) counts as a l a w i f i t has no counter instances.
2. There is the problem of non-instantial laws. Newton's first l a w ('All
bodies on w h i c h no net external force is acting either r e m a i n at rest or
move i n u n i f o r m velocity i n a straight line') is non-vacuous b u t has no
instances. I f one amends the regularity theory to 'universal generalisation
w h i c h have instances', i n order to exclude vacuous laws, Newton's first l a w
is r u l e d out.
3. There is the p r o b l e m of m i s s i n g values. Laws are q u i t e general
( F = m a ranges over a l l values of F, m and a) and so range over i n f i n i t e
instances, only some of w h i c h can have been actualised. R e g u l a r i t y
t h e o r y can only support laws for those instances w h i c h have been
actualised.
4. There is the problem of accidental generalisations. ' A l l dogs w i t h
three legs and one ear are called lucky', i f true, is as m u c h of a l a w as
a n y t h i n g offered by physics.
5. There is a p r o b l e m for some 'epistemic' r e g u l a r i t y theories have
t h e i m p l i c a t i o n t h a t there are no u n k n o w n laws. W h e t h e r s o m e t h i n g is
a l a w depends upon w h e t h e r scientists t r e a t generalisations as laws.
Scientists cannot t r e a t w h a t t h e y have not discovered, so there are no
u n k n o w n laws.

This is not to say t h a t the realist position is problem free, the latest
theories r u n n i n g into problems of identification, t h a t is giving an account
of the relation of necessity t h a t makes l a w rather t h a n regularity, and
inference, t h a t is i f we hold t h a t necessity, however construed, holds
between universals, w h a t is i t t h a t allows us to say a n y t h i n g about
p a r t i c u l a r s . Interestingly for w h a t follows, Dretske offers a legal analogy
68

to explicate the inference r e l a t i o n . Vlastos w r i t i n g i n 1969 held t h a t the

69

regularity theorists were more populous, and while t h a t has probably

 6. Plato and the Development of Greek Cosmology 173
changed i n the last t h i r t y years, the i m p o r t a n t point is t h a t the realist
position is s t i l l a viable one. 70

There is another i m p o r t a n t aspect to Plato's distinction between em-

peiria and techne, to do w i t h the development of science. One m i g h t argue
t h a t firstly h u m a n k i n d developed technology, the ability to manipulate
and predict phenomena w i t h o u t being able to produce a proper explana­
t i o n of those phenomena. The first ages of m a n k i n d can be defined i n
relation to the dominant technology, thus stone age, bronze age, i r o n age,
etc. Only after this do proper scientific explanations develop. Exactly w h e n
and where is too large an issue here, though personally I w o u l d place this
w i t h the Milesians and the creation of the idea of the cosmos and the
rejection of mythological and supernatural explanations. One distinguish­
i n g feature of a society t h a t m i g h t be said to possess science t h e n w o u l d be
a consciousness of the distinction between science and technology. That, I
suggest, is first clearly formulated by Plato, and t h a t is an i m p o r t a n t point
i n the development of Greek science. While his distinction between em-
peiria and techne is clearly not an exact match for the modern distinction
between science and technology, i t does show a recognition of something
beyond m a n i p u l a t i o n and prediction based on 'experience and g r i n d i n g
away' (empeiria kai tribe), which we cannot give a proper account of. We 71

m i g h t also differ from Plato i n w h a t we believe to be on the techne side of

this distinction, b u t the key phrase (Gorgias 500e) would seem to be t h a t
medicine at least 'investigates the nature of those i t treats and the reason
for w h a t i t does, and has some account to give of each of these things'.

VI. Civil law and physical law

We can now examine why, i n order to explain regular and orderly behav­
iour, Plato attributes soul and intelligence to the heavenly bodies. He does
not have at his disposal a conception of gravity or action at a distance, or
indeed a notion of physical l a w i n the modern sense at all. The like-to-like
principle can explain the regular behaviour of matter, b u t as we saw i n
Chapter 1 i t is insufficient to explain the order and the continuing appar­
ently intelligent behaviour of the heavens. I n order to produce and sustain
good order intelligence is required. So where the proto-physical l a w of the
like-to-like principle fails we need to t u r n to something else. The Greeks
also h a d the idea of nomos, l a w ( i n the civil sense) or convention, w h i c h
was often contrasted w i t h phusis, nature. Thus the idea of a l a w of nature
i n this sense would be something t h a t m i g h t well look rather strange to
the Greeks, and indeed where this phrase occurs i n Plato and Aristotle i t
looks rather paradoxical. WTien other ides fail, though, one m i g h t t r y to
72

apply the idea of civil l a w to nature. However, there is an immediate

problem. W h y should nature obey something like civil law? I t is a m a t t e r
of choice, of intelligence and m o r a l character for humans to obey such laws.
Thus Plato gives the cosmos and the celestial bodies some semblance of
174 Plato's Philosophy of Science
intelligence and soul i n order t h a t they may be able to obey laws. I n order
to get perfect regularity, those entities must not only be able to make the
choice to obey the laws of celestial motion, b u t also have to be perfectly
rational, t h a t is gods, i n order t h a t they do so a l l the t i m e . Timaeus 47bc
is of considerable importance here, as i t contrasts the entirely unwander-
ing (pantos aplaneis) revolutions of god w i t h the wandering (peplane-
menas) revolutions i n ourselves. Something else m u s t be added here,
w h i c h is t h a t the principles of celestial motion m u s t be the best sort of
principles otherwise perfectly good souls would not freely choose to obey
them. Here t h e n we see another reason for Plato's teleology. These princi­
ples need to be good i n order to be obeyed and the celestial bodies need to
be good to obey t h e m w i t h o u t any further compulsion.
Another i m p o r t a n t factor to come out of this is the ongoing w o r k of
intelligence i n the cosmos. As we have seen, i t is one t h i n g for the cosmos
to come i n t o an organised state, quite another for i t to continue i n t h a t
state. Plato has criticisms of the atomists on this matter, b u t here we
perhaps see an i m p o r t a n t reply to Anaxagoras, by showing j u s t how
intelligence may continue to operate i n the cosmos and contribute to its
organisation and stability. The l a t t e r is i m p o r t a n t as for Plato the physical
w o r l d is i n some sort of flux. Just how radical t h a t flux is w i l l be a topic for
the next chapter, as w i l l the relations of the heavens to t h a t flux. The fact
t h a t the heavenly bodies have divine souls w i l l be i m p o r t a n t i n this context
for the ongoing stability of the heavens. One m i g h t also note t h a t i t is the
heavens where reason scores its greatest victory over necessity, as is
m a r k e d by the t r a n s i t i o n from the works of reason to those of necessity at
Timaeus 47e. T h a t is not surprising perhaps as we have the continued
influence of intelligence.
This is not of course Plato's only attempt to explain consistent behaviour
w i t h something more t h a n a regularity theory. I f we look to Phaedo
102b-105b, t h e r e is t h e ' b r i n g i n g on' r e l a t i o n s h i p for h o t / f i r e a n d
cold/snow, odd/3 & 5 and even/2 & 4 . So too i f we look to Timaeus 61d
73 74

ff., we find an account of the properties and behaviour of atoms related to

t h e i r shapes. I t is interesting to note t h a t i n each of Plato's three accounts
of w h y this should be so he begins to address the modern problems of
identification (what is the necessity involved i n these relationships?) and
inference ( i f relations hold between universals, w h y do particulars follow
them?). I n the Phaedo, the nature of the necessity is the relation between
forms (though i t is not entirely clear i n the Phaedo w h a t t h a t is), and
inference is validated by the model of participation of particulars i n
forms. 75

I n the Timaeus we have a different sort of necessity based on geometri­

cal atomism. A l t h o u g h the Phaedo does not use geometrical examples, i t
is not h a r d to see how i t might. W h a t is different about the Timaeus is t h a t
i t connects fire w i t h swiftness and ability to cut of necessity v i a the
geometrical structure of the fire a t o m . W i t h the account of the heavens
76
 6. Plato and the Development of Greek Cosmology 175
i n the Timaeus and Laws, the nature of necessity is again an issue, as is
why i n d i v i d u a l gods w i l l act i n a rational and regular manner.
One can of course raise objections against a l l of these accounts, b u t a l l
I w a n t to point out here is Plato's struggle towards nomological realism i n
the absence of an idea of physical law. To some extent I would agree w i t h
Vlastos, w h e n he suggest t h a t Plato's comment to Leucippus and Democri­
tus m i g h t w e l l have been:

I f you must have rational necessity in nature, you cannot get it from
regularities which are matters of brute fact. The only kind of necessity
known to me is that of mathematics and dialectic. 77

I n relation to this i t is interesting to see how much of the fundamental

structure of the physical w o r l d is given a mathematical, or more accurately
a geometrical basis i n the Timaeus. T h a t i n itself is a significant move i n
Greek cosmology, w h i c h I shall discuss i n more detail i n the context of
geometrical atomism. W h i l e Plato never manages a proper conception of
physical law, the geometric cosmology may well have been influential i n
its later formulation. The Renaissance becomes interested i n the applica­
t i o n of mathematics to nature, and we can find Galileo saying:

Philosophy is written in that vast book which stands forever open before our
eyes, I mean the universe; but i t cannot be read until we have learnt the
language and become familiar with the characters in which it is written. I t
is written in mathematical language, and the letters are triangles, circles
and other geometrical figures, without which means it is humanly impossi­
ble to comprehend a single word.

One way i n w h i c h the scientific revolution may have come to the idea of
physical l a w is v i a the though t h a t i f nature is fundamentally geometrical,
t h e n perhaps i t obeys laws something like of geometry. Plato's geometrical
cosmology was a significant step forward, even i f he was not able to exploit
to the full a l l the possibilities t h a t i t offered.

V I I . A n a l o g u e s for p h y s i c a l p r o c e s s e s
I t is sometimes said t h a t the Greeks had an organic picture of the world,
w h i l e we have a mechanistic one. Clearly t h a t is an overgeneralisation,
78

b u t there is a significant difference i n style of explanation between the

ancients who d i d hold a more organic w o r l d view and modern views. Since
the scientific revolution of the seventeenth century, and the overthrow of
scholasticism and the n a t u r a l magic t r a d i t i o n , we have become accus­
tomed to scientific explanations h a v i n g a certain form. Rejecting the
scholastic modes of explanation and occult qualities, we explain i n terms
of physical causation and mathematically framed laws of n a t u r e . As we 79

consider the w o r l d to be u l t i m a t e l y made up of small particles, u l t i m a t e

176 Plato's Philosophy of Science
physical explanation is i n terms of those particles and t h e i r properties. 80

We use mechanical models as analogues for physical processes. We also

use physical and mechanical analogues to describe biological processes.
This is perhaps most clearly seen w i t h the seventeenth-century view t h a t
animals were no more t h a n very complicated pieces of clockwork, clock­
w o r k being the key metaphor for the mechanical philosophy of the seven­
t e e n t h century. C l e a r l y t h i s sort of e x p l a n a t i o n has become more
81

sophisticated, b u t w h a t I shall t e r m the direction of explanation remains

the same. We explain biological processes i n physical and mechanical
terms.
There are though other possibilities, and we m i g h t reverse this direc­
t i o n explanation, and use biological analogues and for physical processes,
as Plato does w i t h the cosmos and the celestial bodies. Aristotle is perhaps
a good example of how biological analogues can i n t r u d e into physics i n a
more subtle manner, and this is something w h i c h is continued by the
Stoics. W i t h modern science having moved away from the use of biologi­
82

cal analogues, the essential question is the following. Is i t evident, i n the

context of fourth-century Greece, t h a t one should be using mechanical
rather t h a n biological analogues? The answer to t h a t is quite clearly no.
There are several reasons for this. I f one wishes to explain order and
regularity, mechanisms are not a good option for the Greeks. As F u r l e y has
argued, mechanism, and particularly clockwork, are paradigms of regular­
i t y and predictability to us. The mechanisms available to the Greeks
83

though, say the cart and the winch do not offer such an example. I t is not 84

surprising t h e n t h a t even where the Greeks m i g h t be described as mate­

rialists, they are not mechanists. I t is w o r t h noting t h a t even the atomists
use biological analogues for physical processes as well. As we saw i n
Chapter 1, Sextus tells us t h a t Democritus bases his argument for a
like-to-like principle out of the vortex on both animate and i n a n i m a t e
things. A n i m a l s congregate w i t h animals of the same and s i m i l a r l y i n the
case of inanimate things, such as seeds t h a t are being riddled and pebbles
on the sea-shore. 85

Biological analogues were not the only Greek alternative to mechanical

ones. As L l o y d has argued, the Greeks have quite a variety of analogues
for the cosmos. These can be grouped under the ideas of the cosmos as an
artefact, as a political e n t i t y or as a l i v i n g entity, but even t h e n there is a
great deal of variety w i t h i n the groups. I t is i m p o r t a n t to appreciate t h a t
86

the Greeks can get a good deal from these analogues. They feel the need
to explain the u n i t y of cosmos, the current and ongoing order of the cosmos
and the nature of processes w i t h i n the cosmos. Thus the idea of the cosmos
as something analogous to an a n i m a l or a polis would seem quite reason­
able to t h e m . I t is easy to forget t h a t these are significant questions for
87

the Greeks. For the C h r i s t i a n t r a d i t i o n i t is God who creates and sustains

the u n i t y of the cosmos, and gives i t its current and ongoing order. For
modern science, the cosmos is not a unified e n t i t y i n the same way. This is
 6. Plato and the Development of Greek Cosmology 111
due to many factors, not least a more sophisticated understanding of the
nature of space and b i g bang cosmogony. However, t h e u n i t y of the cosmos
is a problem for those who believe i n m u l t i p l e worlds, as one has to give
some sort of account as to w h a t individuates the cosmoi. The ancient
atomists would possibly answer t h a t by referring to the membrane w h i c h
is supposed to form around each cosmos. 88

W h a t t h e n have we got i n the way of objections to Plato's idea of the

cosmos a n d the celestial bodies as animate and intelligent entities? Plato
is careful to l i m i t the characteristics of these entities i n such a way t h a t
they always behave i n a regular manner. H i s view, then, is not regressive
relative to t h e Greek advances i n terms of n a t u r a l explanation versus
mythology and supernatural explanation. We can h a r d l y blame h i m for not
employing ideas such as g r a v i t a t i o n and physical l a w w h i c h were not
available to h i m . Plato makes an interesting effort at explaining the
regular behaviour of the heavens w i t h w h a t he has available, w h i c h are
biological analogues a n d t h e notion of civil law. O u t of t h a t come t h e
nomological reasons w h i c h Plato has for teleology. Again, i n context bio­
logical analogues would have appeared at least as viable a possibility as
mechanical analogues, even given t h a t the Greeks h a d mechanical ana­
logues i n a reasonable form, a n d i t is less t h a n clear t h a t they did. Plato
gets a good deal of explanatory mileage out of these analogues on questions
w h i c h are of considerable importance for Greek cosmology. We are r i g h t at
first reading to be worried by Plato's account of the animate and intelligent
heavens. However, i f we place his ideas carefully i n context, and assess the
resources t h a t were available to h i m , i t is possible to see Plato m a k i n g a
very reasonable attempt to explain w h y the heavens move as they do.
A t t h e beginning of the twenty-first century, we perhaps ought to be
more open to alternatives to the mechanical conception of the world. W h y
so? Because the last century saw the gradual dissolution of the classical
mechanical picture of the world. This begins w i t h the rise of thermo­
dynamics, and the idea of entropy and the irreversibility of m a n y physical
processes i n the latter part of the nineteenth century. Relativity broke
89

down some mechanical conceptions of matter w i t h the interconversion of

m a t t e r a n d energy. Q u a n t u m mechanics dealt further blows to t h e me­
chanical picture w i t h the idea of inherent probabilism and i n d e t e r m i n i s m ,
and the idea of wave/particle duality. The u l t i m a t e difficulty we have found
is t h a t the micro w o r l d is simply not like the macro world. W h i l e i t was
possible to use macroscopic analogues for the micro w o r l d (particles being
like b i l l i a r d balls being a typical example) i t is no longer possible to do so.
The physical micro w o r l d does not operate along mechanical lines.
I a m not suggesting here t h a t biological analogues for physical proc­
esses have made or w i l l make a comeback. I merely w i s h to point out t h a t
the mechanical models for physical processes w h i c h were predominant
from t h e seventeenth to the nineteenth century are not the be a l l and end
a l l of science. Doubtless they were a viable a n d plausible possibility at
178 Plato's Philosophy of Science
t h e i r t i m e , and they facilitated a good deal of progress, b u t science is now
moving beyond them. I n due course, we may come to understand w h y we
found such mechanical analogues so alluring, at a specific stage i n the
development of science, relative to the analogues w h i c h succeed t h e m . I n
the meantime, we should be as generous to the ancients. There was good
reason w h y biological analogues for physical processes should have been
attractive to t h e m .

V I I I . Plato a n d subsequent astronomy

Even i f Plato's ideas about the nature of the cosmos and the celestial bodies
were reasonable i n context, one m i g h t s t i l l be worried about the effect t h a t
they h a d on subsequent astronomy and cosmology. I n fact, w i t h the context
of Greek geocentrism, they were i m p o r t a n t i n allowing some significant
advances. W h a t explanations of the motions of the heavens do the Greeks
have? Effectively they have three sorts of answer, the Platonic, the Aristo­
t e l i a n and the vortex. Plato's answer is t h a t the planets have the ability to
move themselves, and regulate t h e i r own motions. The heavenly bodies
are gods, are animate and have souls. As we have seen, i t is the intelligence
of these entities t h a t guarantees t h e i r regular behaviour, and they are
animate and ensouled to t h a t extent and no more. A l t e r n a t i v e l y there is
Aristotle's answer of nesting spheres of aether, w i t h the full complement
of d r i v i n g spheres and counteracting spheres. This looks somewhat more
mechanical i n its orientation, and certainly the planets are more con­
strained by w h a t is around them. They are though executing t h e i r n a t u r a l
motion and are moving due to t h e i r love of the p r i m e mover, and are
described by Aristotle as being d i v i n e , so this view is not as distinct from
90

Plato as t h a t of the atomists and Anaxagoras, who hold t h a t purely

m a t e r i a l celestial bodies w i t h no active properties of t h e i r own are d r i v e n
r o u n d by a v o r t e x .
91

One can see from this brief survey t h a t the idea of a force such as g r a v i t y
controlling the motions of the heavens is a long way off for the Greeks.
Even i f the Greeks had been i n possession of a conception of g r a v i t y
though, considerable difficulties lay i n the way of t h e i r u t i l i s i n g such an
idea to explain the motions of the heavens. W h y so? For the Greeks there
is a complexity of the motions of the heavens w h i c h cannot be explained
by a single force emanating from a homogeneous spherical earth. T h a t
complexity is generated inherently by geocentrism allied to a static earth.
I n such cosmologies a l l the motions of the heavens have to be real motions.
None of t h e m can be apparent motions due to motion of the observer on
the earth. Even i n a model as simple as Plato's, how can one explain the
different motions generated by the same and the different as due to a
single force i n a spherical and homogeneous earth? The position r a p i d l y
becomes more difficult as the complexity of the model grows. The addition
of extra spheres for the sun and moon and the hippopede generating
 6. Plato and the Development of Greek Cosmology 179
spheres for the other planets i n the models of Eudoxus and Callippus
makes i t even more difficult to see a l l of these motions i n terms of a single
force emanating from the earth.
I t should also be evident t h a t like-to-like principles and other ideas
w h i c h the Greeks use to account for w h a t we would t h i n k of as gravita­
t i o n a l phenomena w i l l not be of any help here either. Aristotle's model is
h i g h l y ingenious i n t h a t i t gives a coherent picture of the cosmos tied to a
general theory of motion and n a t u r a l place (which does give an account of
many g r a v i t a t i o n a l phenomena), even i f i t does require a special account
of celestial motion. However, there are problems w i t h accommodating
further astronomical sophistication w i t h i n the model. A t this stage there
is already too m u c h sophistication for the vortex theory, and t h a t may be
so even w i t h Plato's simple model of the Timaeus. W h i l e a single vortex
m i g h t give a first account of the motions of the heavens, i t is h a r d to see
how (or indeed why) two could combine to give the motions of the same and
the different. W h e n one moves on to the Eudoxan system w i t h 27 circular
92

motions, the idea of vortices is no longer plausible. The later i n v e n t i o n of

more the sophisticated devices of epicycle, eccentric and equant pose even
greater problems, and create difficulties for the Aristotelian answer as
well.

The epicycle
The epicycle (Fig. 42) is the basic device of later Greek astronomy. B y a
suitable choice of speeds of rotation retrograde motion can be produced.
One advantage over the hippopede is t h a t different, and more accurate
shapes of retrograde motion can be generated.

Fig. 42. Epicycle.

180 Plato's Philosophy of Science
The problem here, for both the application of gravity and for Aristotle,
is w h y does a planet revolve around an i m a g i n a r y point (and w h y does t h a t
point itself revolve around the earth?). A further problem for Aristotle, and
indeed one t h a t is inherent i n the concentric sphere scheme, is t h a t
epicycles generate differences of distance from the planet to the centre of
the cosmos.

The eccentric circle

One of the difficulties w i t h concentric sphere astronomy was t h a t a l l the
celestial bodies were held at a constant distance from the earth, b u t some
show considerable variation i n apparent size and brightness. W i t h the
eccentric circle, the earth is now moved from the centre such t h a t there is
v a r i a t i o n i n distance to the planet (Fig. 43). The eccentricity here is greatly
exaggerated to demonstrate this effect. We can add i n an epicycle again to
generate retrograde motion (Fig. 44).

Fig. 43 Eccentric circle. Fig. 44. Eccentric circle with

epicycle.

The critical problem here is t h a t we have a split between the physical

centre of the cosmos, the earth, and the rotational centre of the cosmos, the
eccentric point. I t is t h e n h a r d to see how a force emanating from the e a r t h
could control the motions of the planets, or w h y there should be a force
emanating from the point i n space w h i c h is the rotational centre.

Theequant
The planets do not move w i t h uniform motion, even w h e n they are not
retrogressing. So too there are variations i n the speeds of the sun and
moon. One way of generating variations i n angular velocity from combina-
 6. Plato and the Development of Greek Cosmology 181
tions of regular circular motion is by the equant device (Fig. 45). We have
again an eccentric so the earth is offset from the centre of the cosmos, and
an epicycle. We also have the equant point, about which there is constant
angular velocity. The angular motion of the planet w i l l t h e n be constant
viewed from the equant point, but not from the earth, nor from the centre
of the deferent circle. Here the planet w i l l travel from A to B i n the same
t i m e as i t travels from B to A. F r o m the equant point t h a t is a split of 180°
for each section, but clearly t h a t is not so viewed from the earth. A g a i n the
eccentricity here is greatly exaggerated to demonstrate the effect. Note
also t h a t the epicycle w i l l both affect the shape of the orbit and vary the
planet's angular velocity.

Fig. 45. Equant.

The equant poses similar problems to the eccentric. Now we have

another point i n space, the equant point, and i t is h a r d to see how t h a t
could control the motion of the planets. Thus i t is very difficult to conceive
of a force emanating from the physical centre of the cosmos or the rota­
tional centre of the cosmos which w i l l control the orbits of the planets. This
is a problem t h a t cannot be solved u n t i l the orbits of the planets are
simplified and the physical and rotational centres of the cosmos are i n the
same place again.
The position of Ptolemy is of interest here. While his cosmology is
generally thought of as Aristotelian, there are also some significant Pla­
tonic elements i n i t , especially i n relation to planetary m o t i o n . I n the
93

Suntaxis and indeed also the Tetrabiblos, there is no mention of any

celestial spheres. The later Planetary Hypotheses may present a slightly
different picture. Ptolemy presents the 'nesting spheres' hypothesis i n this
w o r k (which is absent from both the Suntaxis and the Tetrabiblos),
whereby the spheres of each of the celestial bodies f i t together perfectly
182 Plato's Philosophy of Science
w i t h no overlap or void between t h e m . Ptolemy is adamant i n the
94

Planetary Hypotheses t h a t there are no reacting spheres. Indeed, i n his

95

new system there is no need for them, as he conceives of each of the

heavenly bodies as h a v i n g its own source of m o t i o n . Each of the heavenly
96

bodies moves itself w i t h o u t any assistance or interference from any other

body. This means t h a t there is no transference of motion between spheres
97

(Ptolemy specifically rules this o u t ) and so the motions of the outer

98

heavenly bodies are not transferred to the inner ones. I t would seem t h e n
99

t h a t far from being a hindrance to progress, Plato's ideas about the motions
of the planets made up the only scheme w i t h i n w h i c h Greek astronomy
and cosmology could make progress. T h a t would certainly seem to be so
w i t h i n geocentrism (and as we have seen, the Greeks have good reasons
for t h e i r geocentrism), b u t would also be t r u e for Aristarchus as w e l l , for
w i t h o u t the idea of elliptical orbits, i t is not possible for the sun to control
the orbits of the planets.
I t is i m p o r t a n t to recognise t h a t the move to gravity as the controlling
factor i n planetary astronomy was a long one t h a t came i n stages. W h i l e
Copernicus hypothesised a sun centred cosmos, he retained the idea of
regular circular motion. W h i l e he h a d ingenious means for replacing
equants, he needed to r e t a i n epicycles and eccentrics i n order to account
for the non-circular orbits of the planets. So while Copernicus gave us
heliostatism, t h a t is a cosmology where the sun does not move, he d i d not
give us heliocentrism, a cosmology where the sun is both the physical and
rotational centre of the cosmos, as the eccentric device splits the physical
and rotational centres. So too Ptolemy gave a geostatic b u t not a geocentric
cosmology as he too used the eccentric. Kepler's proposal of elliptical orbits
for the planets is perhaps the critical stage, as now the physical and the
rotational centres of the cosmos are t h e n reunited (Figs 46 and 47). I t is
only at t h i s point t h a t one can begin to t h i n k i n terms of a force emanating
from the sun w h i c h m i g h t control the orbits of the planets. This idea indeed
occurs to Kepler, b u t he chooses magnetism as opposed to gravity as the
controlling force. I t is not u n t i l later i n the seventeenth century t h a t we
get gravity as the controlling force, and later s t i l l t h a t we get the critical
1/d formulation.
2

Fig. 46. Kepler's first law.

Kepler's ellipses reunite the
physical and rotational
centres of the cosmos, which
are both at one of the foci of
the ellipse. This greatly
simplifies cosmology. That
one uses only one ellipse per
planet, and not an ad hoc
collections of epicycles,
eccentrics and equants also
simplifies astronomy.
 6. Plato and the Development of Greek Cosmology 183
Fig. 47. Kepler's second
law. Kepler's speed law
generates the variations in a
planet's orbital velocity as a
natural consequence of the
elliptical orbits. Equal areas
are swept out i n equal times
by the line from the planet
to the focus of the ellipse.
Planets then move more
swiftly as they approach the
sun, less swiftly as they
recede from it.

IX. Plato and subsequent cosmology

I shall conclude this chapter by looking briefly at Plato's influence on
subsequent astronomy and cosmology. One reason for giving details of the
w o r k of Eudoxus and Callippus i n the previous chapter has been to show
j u s t how immediately fruitful the challenge set by Plato was. They tackle
precisely the problems t h a t Plato sets i n the Timaeus i n precisely the way
he w o u l d like. So too we have seen t h a t the next phase of astronomy,
involving epicycles, eccentrics and equants, also held to the Platonic
challenge and so too was extremely f r u i t f u l . There is no doubt about the
100

importance of Eudoxus and Callippus i n the history of astronomy, nor of

those who replaced the concentric sphere scheme w i t h epicyclic astronomy
such as Hipparchus, Apollonius and Ptolemy
Some, however, have been highly critical of Plato. B y demanding circu­
l a r motion (solely on the basis of metaphysical speculation), they argue,
Plato set astronomy on the wrong track for 2,000 years, only for Kepler to
reject Platonism and p u t things r i g h t again. This w o u l d be a harsh
judgement for several reasons. F i r s t , i t is the fate of v i r t u a l l y a l l scientific
ideas eventually to become outdated and so i n a sense become a bar to
progress. T h a t the Platonic challenge produced progress at the outset, and
indeed for m a n y years is indisputable. I t is also the case i n the history of
science t h a t some ideas seem so n a t u r a l and evident t h a t i t is easy to forget
t h a t someone actually introduced t h e m and at the t i m e t h a t was an
i m p o r t a n t step. I t is i m p o r t a n t i n relation to this to remember t h a t the
Platonic challenge asks for regular and ordered motions as w e l l as circular
motions, and i m p l i c i t i n this is the idea of a unified theory of the heavens.
Since Plato there has been no question t h a t we should be looking for a l l of
these, except circular motion. I f the objection is merely t h a t Plato got i t
wrong, so m u c h is evident, and w h a t is more i m p o r t a n t is whether circular
motion was a plausible and viable possibility at the time.
Secondly, Ptolemaic astronomy and circular motion were indeed re­
jected i n the Copernican Revolution. However, there is a considerable
184 Plato's Philosophy of Science
amount of analysis required i n sifting which elements of the Platonic
challenge were rejected and which elements may have proved inspira­
t i o n a l i n i n s t i g a t i n g and d r i v i n g forward the Copernican revolution. There
is no simple rejection of Platonic principles here. Copernicus straight­
forwardly accepts regular circular m o t i o n : 101

The movement of the celestial bodies is regular, circular and everlasting - or

else compounded of circular movements. 102

He applies this so strictly t h a t he w i l l not accept the equant. Thus he says

of Ptolemy's system:

These theories were not adequate unless certain equants were also con­
ceived; i t then appeared that a planet moved with uniform velocity neither
on its deferent nor about the centre of its epicycle. Hence a system of this sort
seemed neither sufficiently absolute nor sufficiently pleasing to the mind ...
I often considered whether there could be found a more reasonable arrange­
ment of circles, from which every apparent inequality could be derived and
in which everything would move uniformly about its proper centre, as the
rule of absolute motion requires. 103

One m i g h t also compare Copernicus'following criticism of his predecessors

w i t h Laws 898ab on celestial motion being part of one p l a n and one order:

They do not employ the same principles, assumptions or demonstrations for

the revolutions and the apparent movements ... moreover they have not been
able to discover or to infer the chief point of all, i.e. the form of the world and
the certain commensurability of its parts. But they are in exactly the same
fix as someone taking from different places hands, feet, head and other limbs
- shaped very beautifully but not with reference to one body and without
correspondence to one another - so that such parts make up a monster rather
than a man. 104

So too Copernicus makes reference to a 'best and most orderly demiour-

gos', and to 'godlike circular motions'. Copernicus' precise m o t i v a t i o n
105 106

and intentions are a m a t t e r of considerable debate. W h a t is generally

agreed, though, is t h a t Copernicus' m a i n concern was not w i t h the empiri­
cal adequacy of the Ptolemaic system, b u t rather w i t h its structure, and
107

t h a t Copernicus was influenced to some extent by Neoplatonic t h o u g h t . 108

I t is clear t h a t Copernicus adhered very strictly to regular circular mo­

tion, 109
and t h a t this far from being a bar to progress may w e l l have
precipitated the chain of thought leading to the heliocentric hypothesis.
I f we look to Kepler, his Neoplatonism is even more m a r k e d . One can 110

point i n p a r t i c u l a r to his attempt to derive the ratios of the orbits of the

planets from the five Platonic solids. Kepler though rejects circular mo­
tion, and is often held to have thus broken w i t h Platonism. One question
we m i g h t ask here is simply, W h a t shape should Plato have used? I f one is
 6. Plato and the Development of Greek Cosmology 185
going to have more t h a n a purely descriptive astronomy one m u s t take
some view on the basic motions of the heavens, and t h a t brings us back to
Meno's paradox. W h a t hypothesis should we propose? F r o m an indefinite
number of possible motions circular motion must have seemed by far the
most plausible and viable possibility, especially from a geocentric view­
point, and as we have seen the Greeks had very good reasons for t h e i r
geocentrism. The stars appear to move i n v i r t u a l l y perfect circles, so a good
first guess at the motions of the planets has to be circular motions or a
combination of circular motions. E l l i p t i c a l orbits for the planets are i n fact
quite h a r d to conceive of even w i t h heliocentrism. This is because the
orbits are not actually perfect ellipses, and because the eccentricity of the
ellipses varies considerably from planet to planet. The orbits are not
elliptical because of perturbation (Figs 48 and 49).

Fig. 48. Two bodies. I f we

have a two body system with
no external influences then we
get a straightforward elliptical
orbit. The only forces operating
are those between the orbiting
and the orbited bodies.

Fig. 49. Perturbations. When we add another orbiting body, if we consider only
the forces between the bodies and the sun, then we have two perfect ellipses, as
here. However, we also have to take into account forces operating between the two
orbiting bodies, which will distort their elliptical orbits. These distortions are
known as perturbations.
186 Plato's Philosophy of Science
W h e n Kepler eventually breaks w i t h circular motion, i t is i m p o r t a n t to
recognise w h a t he gains by this. As explained i n the previous section, there
is the i m p o r t a n t reunification of the physical and rotational centres of the
cosmos, w h i c h allows the idea of a force emanating from the sun to control
the motions of the planets. A l l i e d to this is the rejection of the Ptolemaic
devices of epicycle, eccentric and equant w h i c h h a d to be applied i n a
relatively ad hoc manner to each planet. W h a t this achieves is an enor­
mous simplification of the heavens, and a unification of the account of the
motion of each of the celestial bodies, and again one m i g h t look to Laws
898ab and its demand t h a t there be one p l a n and order to the cosmos. B y
breaking the letter of the Platonic challenge i n rejecting circular motion
Kepler is able to preserve its spirit or regular, ordered motion and one p l a n .
Indeed t h a t spirit is very m u c h alive today, but has become so m u c h a
background assumption t h a t we rather take i t for granted. I t is also
interesting to consider Kepler i n relation to Timaeus 54a, where Timaeus
says t h a t anyone who can produce more beautiful shapes for the basic
triangles w o u l d be welcomed as a friend rather t h a n a foe. So too here
perhaps. I strongly suspect t h a t Plato would welcome Kepler as a friend
rather t h a n a foe.
 7

Geometrical Atomism - Flux

and Language

Plato's geometrical atomism is i n m y view somewhat undervalued. Pyle's

magisterial Atomism and its Critics gives l i t t l e mention to i t , and concen­
trates on the t r a d i t i o n of Leucippus, Democritus and Epicurus. This is a
1

shame, as geometrical atomism is an i m p o r t a n t and interesting theory i n

itself, as well as h a v i n g i m p o r t a n t relationships to the earlier theories of
Leucippus and Democritus and later theories such as those of Descartes.
I t is interesting t h a t Plato presents geometrical atomism using a letters
and syllables analogy, as Aristotle tells us t h a t for the atomists the
differences between bodies are caused by the atoms and are due to:

shape, arrangement and position; being, they say, differs only in 'rhythm,
touching and turning', of which 'rhythm' is shape, 'touching' is arrangement,
and 'turning' is position; for A differs from N in shape, A N from NA in
arrangement, and Z from N in position. 2

We shall also see something very a k i n to Empedocles' four-element theory

under attack, as w e l l as the ideas of the atomists and the Milesians. As we
shall see, there are interesting parallels between the use of the letters and
syllables analogy i n the Timaeus and other late works, notably the
Theaetetus and the Philebus.
E x a m i n i n g geometrical atomism w i l l also involve questions concerning
the nature of the receptacle, the degree to w h i c h the w o r l d can be said to
be i n flux and w h a t we can say about the w o r l d around us. So far, I have
argued t h a t i n the Timaeus the heavens move i n an entirely regular
manner and are entirely stable. One m i g h t feel t h a t runs against the
3

general tenor of some passages i n the Timaeus and creates a significant

tension, i n particular w i t h 27d where a radical disjunction is made be­
tween w h a t always is and never becomes, and w h a t always becomes and
never is. There are several possible strategies for resolving this tension.
First, we m i g h t look to Timaeus 29c, where after arguing t h a t we can only
expect a l i k e l y account of a likeness Timaeus says t h a t i f we cannot
produce an entirely consistent account we should not be surprised b u t
should gladly accept the most likely account and not seek any further. So
we m i g h t shrug our shoulders and accept t h a t such tension is inevitable.
188 Plato's Philosophy of Science
W i t h passages where several interpretations are possible, we m i g h t look
to reduce this tension by i n t e r p r e t i n g t h e m so as to produce the greatest
overall harmony. We saw an example of this i n Chapter 4 where I gave an
interpretation of the relation between reason and necessity w h i c h would
allow for regular and stable celestial motion. A second strategy w o u l d be
to argue t h a t there is sufficient difference between the nature of the
heavens and the rest of the cosmos to allow for the stability of the heavens.
Clearly the fact t h a t a l l the celestial bodies are gods w i l l be significant
here, t h o u g h there are some other i m p o r t a n t differences as well. I f this
fails to release the tension entirely, t h e n we can fall back on the first
strategy h a v i n g at least dissolved some more of the tension. The t h i r d , and
by far the most radical strategy, is to counter-attack. I f one can harmonise
a sufficient number of passages w i t h the idea of regular and stable
celestial motion, t h e n one m i g h t become suspicious of the orthodox inter­
pretation of any r e m a i n i n g recalcitrant passages. T h a t is something I shall
explore i n Chapter 9. I n this chapter and the following one I shall argue
t h a t i t is possible to produce a reading of geometrical atomism, the
receptacle and flux t h a t underpins the stability of the heavens and allows
us to t a l k i n an informative manner about the physical cosmos.
As commented i n the general introduction, I take a different line on the
relation of the receptacle to geometrical atomism to most commentators.
Commonly, the receptacle is treated on its own, or geometrical atomism is
interpreted such t h a t i t fits conclusions d r a w n from an analysis of the
receptacle. As w i l l become clear i n this chapter and the next, I believe there
are serious problems w i t h the receptacle w h i c h can be resolved by the
introduction of geometrical atomism and a reformulation of the nature of
the receptacle. So i n effect I take geometrical atomism as the dominant
theory to w h i c h we need to fit an i n t e r p r e t a t i o n of the receptacle, rather
t h a n vice versa.

I, T h e receptacle and language

There are many problems w i t h the Timaeus' account of the relation of the
basic constituents of the w o r l d to each other and to the receptacle, b u t here
I w i s h to focus on three. There m i g h t be insufficient stability to u n d e r w r i t e
the stability of the heavens, there m i g h t be problems w i t h how we ought
to refer to a n y t h i n g physical, and there is the m a t t e r of the relation of the
account of the Timaeus on these matters to t h a t given i n the Theaetetus.
A t 49b Timaeus introduces the problem of the relation of language to
things t h a t are i n flux, i n i t i a l l y t a k e n to be earth, air, fire and water by:

Speaking of each of these, to say which ought really to be termed water

rather than fire, and which by any name rather than each and all, so as to
employ language which is sound and secure, is hard (pisto kai bebaio
chresasthai logoi, chalepon). We see, so we suppose, water congealing and
4
 7. Geometrical Atomism - Flux and Language 189
becoming stones and earth, and this same thing when dissolved or dispersed
becoming air, and air becoming fire by combustion ... and thus, i t appears,
they pass into each other in a cycle of birth.

The problem we face i n a t t e m p t i n g to construct a secure logos is t h a t the

elements t h a t we refer to appear to change into one another. A l o n g w i t h
5

others, I take this to be a problem concerning identifying references. I f we 6

identify some stuff as water, and i t changes into air, should we now identify
the same stuff as air? I f we do, w h a t happens to any distinction between
the elements - w h a t name, 'rather t h a n one and all' ought to be applied to
each? I f so, this w i l l t h e n function as a critique of anyone h o l d i n g any of
these substances to be basic, which w i l l certainly include Empedocles,
anyone who holds t h a t there are fire atoms and even the Milesians. The
key passage is perhaps the following, w h i c h has been the subject of m u c h
debate concerning both its translation and interpretation. A t 49d ff. Timaeus
tells us:

Since, then, each of these never appear the same, which of them can we
assert with confidence and without shame to be this - whatever it is - and
not something else? I t is not possible, but much the safest way to speak of
them is as follows. Whatever we perceive as always changing into something
else, like fire, in each case we should not call fire 'this' but 'the suchlike' (hos
pur, me touto alia to toiouton hekastote prosageurein pur). Nor should we call
water 'this' but always 'suchlike', nor anything else as this i f i t had some
stability, among the things we indicate by the expressions 'this' or 'that', and
think we indicate something; for they slip away and do not wait to be
described as 'this' or 'that' or any term which attributes permanence to
them. 7

The central debate concerns the phrase, me touto alia to toiouton hekastote
prosageurein pur. The t r a d i t i o n a l reading is t h a t touto 'this' and to
8

toiouton 'suchlike' are competing predicates for the subject pur, fire. The
phrase t h e n concerns ways i n w h i c h we may t a l k of fire, one proper and
one improper. So we ought not to call fire 'this', b u t we can call i t 'suchlike'.
The alternative reading takes touto and toiouton to be competing subjects
for the predicate pur, the sense now being t h a t words such as fire w h i c h
we now apply to transient phenomena are better applied to more stable
entities. So i f we are to use 'fire' properly, we should only use i t to refer to
entities w h i c h are 'this', and not those w h i c h are 'suchlike'.
B o t h of these interpretations are acceptable renderings of the Greek. 9

WThich we choose to accept, however tentatively, w i l l be dependent on more

general considerations. As we shall see, this is not the only issue related
10

to the question of flux where there is such a choice. As indicated i n the

introduction to t h i s chapter, m y concern is whether these passages can
support the regular and stable motions of the heavens. So too as G i l l
comments, the decision w h i c h reading to adopt:
190 Plato's Philosophy of Science
reflects a decision about the status Plato grants to physical phenomena. On
the traditional reading he legitimates talk about such objects. On the alter­
native he proscribes i t . 11

So on strategic grounds I opt to explore the first of these alternatives. This

is not to say t h a t the alternative reading is not a viable one, nor is to say
there is not a great deal of interest to be said for the alternative reading,
b o t h i n relation to the rest of the Timaeus and other works, notably the
Theaetetus. I do have concerns about the alternative reading i n t h a t i t
seems i n m a n y versions to generate four types of e n t i t y for Timaeus
48e-53b where the text is adamant there are t h r e e . 1 also have worries
12

as to how w e l l the alternative view can account for the relation of the
entities of geometrical atomism to the receptacle. I shall r e t u r n to these
matters a l i t t l e later, b u t let us see w h a t we can make of the t r a d i t i o n a l
reading first.
A major concern over the t r a d i t i o n a l reading was the apparent discord
between the Timaeus and the Theaetetus. I t has been argued t h a t the
Theaetetus post-dates and corrects the Timaeus'View on the relation of flux
and language. I t is surely true, so Owen argued, t h a t i f everything is i n
13

radical flux, t h e n we cannot successfully refer to a n y t h i n g at a l l , and this

is a better position t h a n the l a m e plea' of Timaeus 49d ff. t h a t we can refer
to the four elements as 'the suchlike'. The alternative view was pioneered
14

by Cherniss, possibly w i t h the a i m of defusing the argument concerning

15

flux for d a t i n g the Timaeus before the Theaetetus. A t Theaetetus 182c ff.,
Socrates and Theodoras have the following exchange concerning the Her-
aclitean position and its relation to language:

Socrates: Let us ask them, are all things changing and in flux? ... Have
they both kinds of change that we distinguished, of place and of quality?
Theodorus: Of course; they must if they are to be completely in change.
So.: I f they only underwent change of place, and not of quality, then we
would be able to say what the flowing things that change position are. Could
we speak i n this manner? 16

Theo.: We could.
So.: But since not even this stays still, what flows flowing white, but
changing, so that there is a flux of this very thing also, the whiteness, and a
change of colour, lest i t be convicted of staying still, is it possible to name a
colour with the result that i t is correctly named?
Theo.: But how could one contrive that, Socrates? Or indeed for anything
else of this sort, i f i t always slips out from under us as we speak, being in
flux?

The flux described i n this passage is a very radical one indeed. N o t only is
there n o t h i n g w h i c h is free from qualitative change, b u t everything
changes i n every respect at every successive instant, l e s t i t be convicted of
standing still'. The passage t h a t is supposed to create problems for a late
dating of the Timaeus is the following at Theaetetus 183a:
 7. Geometrical Atomism - Flux and Language 191
So.: I f everything is changing, every answer to any question is equally
correct, both to say 'thus' and 'not thus', or if you wish, 'becomes', if we are
not to employ any expression that will bring them to a standstill.
Theo.: You speak truly.
So.: Except this, Theodorus, I said 'thus' and 'not thus'. I t is necessary,
though, not even to say 'thus'. For 'thus' would no longer be changing, and
nor would 'not thus'. Indeed, there is no motion in 'this', either. Those who
hold this theory must establish some other phrase with which to express
it, as by their own hypothesis they now have no words, unless 'nohow' is
allowed. That might be the most fitting expression for them, as i t is
indefinite.

Similarly at Cratylus 439d8-12 one cannot say that anything i n flux is either
ekeino or toiouton. So according to the Theaetetus we cannot refer at a l l to
items w h i c h are i n a radical flux. I t is i m p o r t a n t to note t h a t at Theaetetus
182c9 above, i f something only underwent phora and not alloiosis we
would be able to say w h a t i t was. W h a t t h e n allows us to refer to things i n
flux i n the Timaeus? The leading candidate here is the receptacle. There
is no question t h a t the receptacle can be called touto, and t h a t i t us
supposed to be free from change, and perhaps i t is this w h i c h gives a
sufficient basis for being able to call the elements to toiouton. A t 50b ff.
Timaeus introduces the gold example to explicate this:

I f someone were to mould all the shapes out of gold and without stopping
remodel each of these into the rest, then should he point one of them out and
ask what i t is, by far the safest answer in truth would be that i t is gold, but
as for the triangle and the other figures that occur, it would be wrong to
describe them as having being, as they change even as we state them, and
we should rest content if they willingly receive the description of the suchlike
with some safety.

So where we have flux, and here we would seem to have a radical flux, 17

we can refer to w h a t is changing by to toiouton. W h a t is i t about the

receptacle t h a t allows this, and does the Timaeus give us a coherent
account of the receptacle?

I I . Some paradoxes
I take the p r i m a r y task of the receptacle to be to provide a licence for to
toiouton expressions. I f so, i t would seem t h a t the receptacle cannot be
merely a theory of space. I f something is changing, and I a m a t t e m p t i n g
to make an identifying reference, i t does not seem to be p a r t i c u l a r l y
helpful to say t h a t this change is occurring i n space. Space may w e l l be
unchanging and so be a suitable referent of touto, b u t could h a r d l y be said
to constitute or lend any stability to the phenomena occurring i n i t . I f a
chameleon changes colour, I need to know something about the constitu­
t i o n of the chameleon, not about the tree t h a t i t is s i t t i n g i n .
192 Plato's Philosophy of Science
I n order to license to toiouton references to the things i n flux, then, the
receptacle m u s t be related i n some constitutive way to them. Certainly the
metaphors of the receptacle being like gold, or a moulding-stuff, or an
odourless base for scent would seem to invite some form of m a t e r i a l
i n t e r p r e t a t i o n of the receptacle. We m i g h t discuss here whether Plato
18

has one theory of a m a t e r i a l receptacle, and whether t h a t is a coherent

theory. I t is quite possible t h a t shapes relate to gold i n a different way from
how scents relate to bases or mothers to offspring etc., and t h a t at least
some o f these are incompatible w h e n we come to consider how t h e
receptacle relates to forms. A n a l t e r n a t i v e to keep i n m i n d t h o u g h is
t h a t Plato m a y give us several conceptions a n d wishes us to consider
theories of a m a t e r i a l receptacle i n general. As we saw w i t h pre-cosmic
chaos i n Chapter 1, he m a y provide something more flexible a n d more
powerful t h a n a s t r a i g h t f o r w a r d statement of cosmogony. W h a t e v e r
degree of pre-cosmic chaos we m i g h t suppose, a teleological o r d e r i n g is
r e q u i r e d . Here, however we conceive of a m a t e r i a l receptacle there m a y
be d i f f i c u l t i e s .
19

I n order for the receptacle to r e m a i n a proper referent of touto, i t must

r e m a i n changeless. A t Timaeus 50d4-51b2 great emphasis is l a i d on the
characterlessness of the receptacle. This is so i n order t h a t i t does not
distort w h a t is i n i t , b u t also the receptacle needs to be characterless i n
order t h a t i t be changeless. As Prior comments, I t s nature, one is tempted
to say, is t h a t i t has no nature.' Now we m i g h t frame a slightly different
20

problem. I n order to be entirely free from change the receptacle must be

characterless. I f i t is entirely characterless, i t is difficult to see how i t
relates to the things i n i t . G i l l argues t h a t i n the gold example, the gold is
not constitutive of the shapes as shapes 21
I would rather p u t this as a
dilemma. I f the gold is constitutive, t h e n the receptacle is i n grave danger
of changing as the shapes change, and so its status as the referent of touto
is endangered. I f i t is not constitutive, t h e n the receptacle cannot licence
to toiouton references.
These problems are perhaps best exemplified by the struggle Timaeus
has at 50bc to produce a coherent account of the receptacle and its relation
to the things i n i t . This passage follows directly from the gold example
where we have been told t h a t the triangles moulded out of gold can be
referred to as to toiouton. Timaeus 50b5-c2 t h e n tells us:

The same account applies concerning the nature of that which receives all
bodies. I t must always be called the same, for it never departs at all from its
own character (dunameos). For since i t always receives all things, i n no way
whatsoever does i t assume a shape (morphen) similar to any of the things
22

which enter i t . 23

T h a t seems quite reasonable for the receptacle i n order to r e t a i n its status

 7. Geometrical Atomism - Flux and Language 193
as a referent of touto. However, immediately following this, at Timaeus
50c2-5 we are told:

For i n nature i t is a moulding-stuff (ekmageion) for everything, changed and

shaped (kinoumenon te kai diaschematizomenon) by the things which enter
it, and because of this appears different at different times.

Timaeus cannot have t h i s b o t h ways and have a coherent account. T h i s

p r o b l e m I s h a l l refer to as the receptacle paradox. One m i g h t add t h a t
i f the receptacle only appears to be changed ( b u t is not changed), i t is
not r e a l l y c o n s t i t u t i v e , a n d i f i t appears to be changed (and is changed)
t h e n i t ceases to be a proper referent of touto. A t Timaeus 50c4-6 we are
told:

The things entering and leaving the receptacle are copies of those ever
existent, and are stamped (tupothenta) from them in some hard to explain
and marvellous manner (dusphraston kai thaumaston), which we will shall
follow up later (hon eisauthis metimen).

This Timaeus conspicuously fails to do i n 48e-53b. A l l we are told is t h a t

the receptacle partakes i n the intelligible i n a most difficult and h a r d to
grasp manner (aporotata ...kai dusalotaton) at 51a7, and at 51b and 52d
t h a t the receptacle is liquefied and ignified. Whether the geometrical
atomism introduced at 53c ff. w i l l help w i t h these matters we shall see i n
a moment.
There is a further problem w i t h considering the receptacle to be mate­
r i a l on the basis of w h a t we have seen of Timaeus 48e-53b. There w o u l d
seem to be very l i t t l e we can say about matter. Timaeus struggles to say
a n y t h i n g positive about the receptacle, and of course there w i l l be serious
epistemological problems w i t h a n y t h i n g so u t t e r l y characterless. A t 24

Timaeus 49a3 i t is s o m e t h i n g d i f f i c u l t a n d obscure (chalepon kai

amudron), at 52b i t is grasped w i t h o u t sensation by some bastard reason­
ing (logismoi tini nothoi) and is the subject of a dream. Something t h a t
25

cannot be explained i n terms of a n y t h i n g more basic and has no character

seems i n itself inexplicable. I f i t is not constitutive, i t is h a r d to see how i t
helps to explain the phenomena, and anyway i f i t is characterless there
w o u l d seem l i t t l e t h a t i t can explain. I t is difficult t h e n to see matter, i f we
equate i t w i t h the receptacle, as a n y t h i n g which is explanatory or is itself
subject to explanation. Perhaps Timaeus' dream, as McCabe comments, is
as problematic as Socrates' dream i n the Theaetetus ( 2 0 l e ff.), where the
basic elements t u r n out to be unknowable. 26

I f we treat the receptacle as being i n some way constitutive of the

phenomena, we r u n into two major problems. I t is not clear t h a t the
receptacle can play the role required of i t i n being so related to the
changing phenomena t h a t i t can licence to toiouton expressions, and i n
being independent of t h e m so as to preserve its own stability (the recepta-
194 Plato's Philosophy of Science
cle paradox). I f the receptacle is both inexplicable and non-explanatory, i t
seems a poor theory of matter. Can the i n t r o d u c t i o n of geometrical
atomism at Timaeus 53c help w i t h these difficulties?

I I I . Geometrical atomism
I n m y view i t is of the utmost importance to recognise t h a t there are
significant differences between the four-element theory p u t forward at
Timaeus 48e-53c and the geometrical atomism introduced at 53c. I n the
theory of 48e-53c, there is no mention of any sub-structure to the elements.
The only m e n t i o n of stoicheia prior to 53c ff. is at 48b9 w h i c h explicitly
denies these elements (earth, water, air, fire) are stoicheia. Each of the
elements of 48e-53c may transmute into any other element. I n the later
theory, w h a t we suppose we see here is contradicted by the ban on the
t r a n s m u t a t i o n of earth and the other elements.
O f considerable importance is the fact t h a t at 48e-53c, the entire
discussion of the elements is conducted at the perceptual level. This is
emphasised very strongly i n the opening passages. We see, so we believe
(hos dokoumen, 49b9 ...horomen, 4 9 c l ) the t r a n s m u t a t i o n of earth, water,
air and fire. They form a cycle 'so i t appears' (hos phainetai, 49c9, cf.
phantazomenon, 4 9 d l ) , and we see things such as fire constantly changing
(kathoromen, 49d4). This would seem to line up w i t h the t r i p a r t i t e onto­
logy of form, receptacle, and t h a t w h i c h is perceptible, w h i c h is stressed at
48e3 ff., 50c7 ff. and 51a4 ff., and 51e6 ff. W h a t is created by form and
receptacle is visible (horaton, 4 9 a l ) , sensible (aistheton, 52a5, cf. 52a7),
and at 51a the mother and receptacle of the visible and sensible (horatou
kai pantos aisthetou, 51a4) must not be called by any of the names of the
elements. A t 52b the receptacle is not accessible by perception (anaisthe-
sias, 52b 1). There is no discussion of a n y t h i n g w h i c h is below our
perceptual threshold rather t h a n i n principle imperceptible.
Timaeus 56bc however is very specific. The atoms of the elements
cannot be seen (ouden horomenen, 56cl-2) because of t h e i r smallness. Only
w h e n m a n y of t h e m are gathered together can a mass of t h e m be seen
(horasthai, 56c3). We can take i t t h a t the i n d i v i d u a l basic triangles and the
complexes they form as the faces of atoms are also too small to be seen.
These differences may have a significant bearing on Plato's intentions. He
may w e l l w i s h to show t h a t the phenomena are i n some sort of flux,
r e q u i r i n g a reform of our references to them, w i t h o u t being committed to
a theory where the u l t i m a t e constituents of the phenomena are i n flux. The
point being d r i v e n at by the first theory may t h e n be a conditional one. I f
we operate only at the level of perception, and we equate changing phe­
nomena, such as the four-elements, w i t h the u l t i m a t e constituents of the
w o r l d , t h e n we w i l l be unable to make any identifying references, or even
any to toiouton references. This may be a criticism of current four element
theories, and may be pushing a theoretical point about fundamental flux.
 7. Geometrical Atomism - Flux and Language 195
Empedocles may be a significant target here, though the point w o u l d apply
to anyone who believes t h a t earth, air, water and fire are basic. T h a t such
an attack is made on these elements for failing to be stable does not mean
t h a t the idea of a stable basis for phenomena is necessarily abandoned. I f
we suppose there to be some form of stable, amorphous substrate, such as
the receptacle t h e n we have the problems outlined i n the last section. The
later i n t r o d u c t i o n of atomic sub-structure may be an attempt to rectify
these difficulties by postulating stable physical entities at a level w e l l
below the threshold of h u m a n perception.
A t 53d Timaeus postulates two sorts of triangle as the archai of the
elements, b u t suggests t h a t there are some higher archai t h a n these. I t is
a reasonable assumption t h a t these archai (or as G i l l terms t h e m , sim­
ples), w i l l be stable. I doubt t h a t a n y t h i n g t h a t is not stable w o u l d qualify
as an arche for Plato. We are also quite justified i n using this assumption
as an hypothesis to see w h a t sort of account ensues. I would agree w i t h
G i l l that: T h e i m p o r t a n t point is not what t u r n s out to be basic b u t t h a t
there be u l t i m a t e simples.' I f there is something basic and simple, these
27

simples can be called to toiouton, and this gives a sufficient basis for the
elements to be called to toiouton as well. I f so, t h e n the problem of the
relation to the Theaetetus is resolved. The objection there was to a flux
where everything changed place and quality. I n the Timaeus we have the
foundation of some things w h i c h while they change place do not undergo
alloiosis. W h i l e G i l l is reticent about identifying the t w o fundamental
types of triangles, w h i c h I shall call the stoicheic triangles, as these basic
and simple entities, I am n o t . I take one of the key points of the
28

introduction of geometrical atomism to be precisely this, to introduce some

basic physical constituents, and the stoicheic triangles are clearly strong
n a t u r a l contenders here. As I shall argue i n the next few sections, there is
a very strong case for considering the stoicheic triangles both as basic and
as free from any change to t h e i r i n t r i n s i c order.
One m i g h t argue t h e n t h a t there is a case for the stoicheic triangles
being called touto, i f they are stable, the two standard objections being t h a t
they undergo phora or alloiosis. The very nature of the stoicheia (whether
they be the stoicheic triangles or something else) precludes alloiosis being
a problem. T h a t they are i n motion may not be a problem either, because
at Timaeus 52de, 57c and 88de we are told t h a t the receptacle moves, and
is shaken l i k e a winnowing-basket. I t is also interesting t h a t at Timaeus
29

49e7 ff. we are told t h a t only t h a t w h i c h the four elements are created from
or are dissolved into may be referred to as touto, and the elements
u l t i m a t e l y are created from and dissolved into the stoicheic triangles. I f
the basic problem here is identifying references, again i t would seem t h a t
the stoicheic triangles can quite safely be identified. To say t h a t this is the
position t h a t the Timaeus adopts would probably be to push matters too
hard, though i t may be something t h a t Plato is leading his readers to
consider. However, i t may be t h a t the fact t h a t the receptacle moves and
196 Plato's Philosophy of Science
can be called touto is an oversight on Plato's part, or one m i g h t argue t h a t
Plato gives us several images of the receptacle and we should not push any
of t h e m too hard. T h a t w h a t the elements can be dissolved into can be
called touto may be restricted to the elements and receptacle model, or
perhaps the dependence of the stoicheic triangles on the receptacle pre­
vents t h e m being called touto. G i l l argues t h a t the simples, whatever
30

they may be, w i l l be called to toiouton because they are distinct from t h e i r
relevant forms and move around. 31

IV. L e t t e r s , s y l l a b l e s a n d t r i a n g l e s
A r e the stoicheic triangles the u l t i m a t e constituents of the physical w o r l d
i n the Timaeus, and are they free from change? Let us begin w i t h a
distinction w h i c h the Theaetetus makes between two sorts of change. A t
Theaetetus 181c Socrates says:

What I want to enquire is this; do they say that there is only one kind of
change (kinesis), or, as i t seems to me, that there are two? ... Do you call i t
32

change when something exchanges one place for another, or spins around in
the same place? ... Let this then be one form of change. But when something
remains in the same place, but grows old, or becomes black instead of white
33

or hard instead of soft, or undergoes any other sort of qualitative change

(alloiosis), isn't i t right to say that this is another form of change? ... There
are then two forms of change, change of quality and change of position
(phoran)34

Consider this i n relation to the basic entities of the world, the stoicheia.
T h a t they move around is not i n question. W h a t is more i n t r i g u i n g is
whether they undergo any change i n themselves. F l u x theories here come
i n m a n y shapes and sizes. These changes m i g h t involve one, m a n y or a l l
of the properties of a basic e n t i t y and those changes m i g h t take place at
any rate up to the radical flux of change every instant. One m i g h t contrast
such flux theories w i t h the idea t h a t the basic entities are entirely stable.
To p u t this i n terms of a distinction I used i n Chapter 1 for Plato's stoicheia,
perhaps the stoicheia undergo no change i n t h e i r i n t r i n s i c order even i f
t h e i r extrinsic order changes. The ancient atomists held such a theory, and
we can compare an elementary modern physics view of the world. Accord­
i n g to t h i s there are three types of simple, uncompounded particles,
electrons, protons and neutrons, w h i c h are the exhaustive physical con­
stituents of more complex entities such as atoms and molecules. These
elementary particles change velocity and position, and enter into bonding
relations w i t h each other, b u t do not undergo any change i n themselves.
Complex entities w h i c h are ensembles of these basic particles on the other
hand, such as atoms or molecules or clouds on a w i n d y day may indeed be
said to be i n some form of flux, w i t h o u t compromising the stability of the
basic particles.
 7. Geometrical Atomism - Flux and Language 197
I n the above passage from the Theaetetus I translated alloiosis as
'qualitative change', as the Theaetetus is concerned here w i t h perceptible
qualities. One m i g h t t h o u g h , as i n the t w o examples above, have a
t h e o r y where the stoicheia are not perceptible so we w i l l have to t a l k of
t h e i r characteristics. We can t h e n d i s t i n g u i s h between t w o q u i t e differ­
ent theories of how t h i n g s change. I s h a l l t e r m these theories of
fundamental flux and of fundamental stability, depending upon
w h e t h e r t h e stoicheia undergo alloiosis. T h e p o s i t i o n ascribed to
H e r a c l i t u s i n t h e Theaetetus is c l e a r l y one of f u n d a m e n t a l ( a n d
i n d e e d r a d i c a l ) f l u x . A t Theaetetus 181d Socrates poses a h y p o t h e t i ­
3 5

cal q u e s t i o n for t h e H e r a c l i t e a n s :

So.: Do you hold that everything changes in both ways, changing both in
place and in quality (pheromenon te kai alloioumenon), or that some things
partake of both, and some only one?
Thes.: By Zeus, I don't know what to say; I suppose they would say both.
So.: I f they did not, my friend, then it would seem that things are both in
motion and standing still, and i t would be no more correct to say that all
things move than all things stand s t i l l . . . then since i t is necessary for them
to change, and since i t is not possible for something not to change i n every
way, all things are always suffering all kinds of change. 36

So too, one m i g h t argue, the flux described i n the Cratylus is a fundamen­

tal one. 1 have no disagreement w i t h this, b u t I shall argue t h a t i t is quite
37

possible t h a t the flux of the Timaeus as one of fundamental stability. A

second way of setting up the contrast here would be to employ a distinction
t h a t the Sophist uses between change relative to itself and change relative
to other things. This would b r i n g us close to Irwin's distinction between
aspect (a-) change and self (s-) change. The following arguments are
intended to establish t h a t i n the Timaeus there are some entities w h i c h
undergo no change of character, no change i n relation to themselves and
no s-change as defined by I r w i n . L e t me outline where I believe these
3 8

basic entities can be located. A t 48b the Timaeus says concerning earth,
air, fire and water:

No one has as yet revealed their generation, but we speak as if we know what
fire and each of them are, postulating them as the basic principles (archas)
of the physical universe, although i t is not fitting for them to be in any
likelihood compared to syllables, even by a man of little insight.

This m i g h t be read as a rejection of the letters and syllables analogy w h i c h

commonly occurs i n later P l a t o . However, i f the two most basic types of
39

triangle t h a t the Timaeus postulates are taken as the letters, and the
atomic faces they go to make up are the syllables, t h e n the atoms w i l l
indeed not be syllables, although the letters and syllables analogy may
s t i l l apply. The Timaeus refers to the two most basic types of triangle using
the substantive to stoicheion. This is i m p o r t a n t as stoicheion carries the
40
198 Plato's Philosophy of Science
general sense of simplest component part, and a more specific sense i n
Plato of letter as opposed to sullabe, 'syllable'. I shall refer to these two
basic types of triangles as the 'stoicheic triangles', and the compound
triangles t h a t are atomic faces as 'complex triangles' (see Figs 50 and 5 1 ) . 41

i i
Fig. 50. Stoicheic triangles - the half-equilateral and the half-square. The
Timaeus claims that earth, water, air and fire should not be called elements, not
even syllables. The basic unit of the physical world would seem to be these two
sorts of triangles, the half-equilateral and the isosceles, often referred to as to
stoicheion. That the proportions chosen include 2 and 3 may have considerable
significance for the history of cosmology and the question of how the world may be
structured mathematically, as I discuss later.

The terminology I use for geometrical atomism is as follows:

I use elements to refer to earth, water, air, and fire.
I use atoms to refer to the i n d i v i d u a l cubes, octahedra, icosahedra and
tetrahedra of the elements.
I use complexes or complex triangles to refer to the faces of the atoms.
I use stoicheic triangles to refer to the two basic triangles, the half-equi-
l a t e r a l and the half-square.

Geometrical atomism has a further refinement, i n t h a t there are different

types of each element. Friedlander, adding due caveats about the relation
to modern atomism, has termed these 'isotopes'. As Cornford has argued,
42

this does not mean t h a t there are different sizes of stoicheic t r i a n g l e s .

43

This w o u l d b o t h s t r a i n the Greek of Timaeus 57d and w o u l d be extremely

 7. Geometrical Atomism - Flux and Language 199

Fig. 51. Complex triangles and squares. I n the specific arrangements dis­
cussed in the Timaeus, six half-equilateral stoicheic triangles form one complex
triangle, and four isosceles stoicheic triangles form one square.

Fig. 52. F i r e atom and earth atom. To form atoms, the complex triangles and
the squares form into solid figures. Here we have fire, the tetrahedron and earth,
a cube. Icosahedra of water an octahedra of air are also made from the complex
triangles.

problematic for the theory of t r a n s m u t a t i o n , both between types of one

element and between elements. Indeed i t would be odd, given Plato's
44

solution to Meno's paradox and his clear choice of specific shapes of the
stoicheic triangles, i f he supposed there to be many sizes of stoicheics w h e n
he has a simple means at his disposal of generating different varieties of
the four elements. So the stoicheic triangles can form up i n different ways
to create different sizes of complex triangles and squares and so different
sizes of atom, corresponding to the different classes of each element, and
200 Plato's Philosophy of Science
we have the further factor t h a t there can be different mixtures of elements
and grades of each element i n order to explain the diversity of the phenom­
ena t h a t we encounter (see Figs 53 and 54).

Fig. 53. 'Isotopes' - more complex planes, squares.

Fig. 54. 'Isotopes' - more complex planes, triangles.

The stoicheic triangles t h e n undergo no change of character, no change i n

relation to themselves and no s-change, or to put i t another way, incorpo­
r a t i n g the single size thesis for stoicheics, a l l half-equilateral triangles are
congruent and r e m a i n so, and a l l half-square triangles are congruent and
r e m a i n so.

V. T h e e v i d e n c e of A r i s t o t l e
For the theory of change of Timaeus 53c ff. to work, no change i n the
stoicheic triangles is required and none is mentioned. Change is analysed
solely i n terms of the composition of stoicheic into complex triangles and
these i n t o atoms and aggregates of atoms, and the converse decomposition.
A t 59c, Timaeus is quite confident t h a t a l l phenomena can be fully ex­
plained i n this manner. I t is perhaps i m p o r t a n t t h a t while we are given an
account of the relative abilities of the atoms to cut up w h a t they encounter
(cf. Timaeus 56b, 56d), we are not given such an account of the stoicheic
 7. Geometrical Atomism - Flux and Language 201
triangles. Timaeus 54c ff. is i m p o r t a n t i n this context. We are t o l d t h a t
w h i l e air, fire and water may transmute into each other, they may not
t r a n s m u t e into earth, and vice versa, as the first three are constructed
from half-equilateral triangles while earth is made from the half-square. 45

I f the stoicheics come apart i n some way, further t r a n s m u t a t i o n would be

possible. W h i l e the Timaeus specifies how atoms and complex triangles
decompose, i t does not do so for the stoicheic triangles. We m i g h t take t h a t
as an indication t h a t they are not supposed to decompose into further
physical elements.
I f i t is the case t h a t the stoicheic triangles of the Timaeus are stable,
t h e n there are some interesting consequences. The theory of h u m a n
ageing tells us t h a t the bonds which hold the constituents of our bodies
together have the power to last only for a certain time, and t h a t as these
begin to fail, so we slip into decrepitude. The sort of flux to w h i c h the
46

atoms and complex triangles are subject is not t h e n the radical flux w h i c h
would have t h e m possessing a quality at t i m e t and not possessing i t at t
+ 1, b u t a rather gentler variant. This is i m p o r t a n t because i t allows the
physical w o r l d a large measure of stability, how m u c h depending on how
strong the 'power to last' is. I n relation to this, i f we examine w h a t the
Timaeus has to say on how h u m a n beings and the heavens are produced,
we find something very interesting indeed. Timaeus 42d ff. tells us t h a t
the task appointed to the demigods is to fashion the bodies of h u m a n
beings. A t Timaeus 42e we are t h e n told:

Taking the immortal element of mortal creatures, imitating their own fash­
ioner, and borrowing from the cosmos portions of fire, earth, water and air
on the condition that this loan could be repaid, they cemented together what
they had taken, though not with the indissoluble bonds (alutois ... desmois)
with which they themselves were held together, but with well constructed
bonds invisible because of their smallness, they fused them into one mass,
creating out of all these one body for each, thus clothing the revolutions of
the immortal soul in bodies subject to egress and ingress.

So while the constituents of our bodies are p u t together w i t h dissoluble

bonds, those of the demigods have indissoluble bonds. Our bodies w i l l be
subject to a gentle flux, while those of the demigods w i l l be entirely exempt
from flux i n terms of qualitative change. The embodiment of these demi­
gods though is i n the heavenly bodies. E a r l i e r i n his discourse, at 38e
Timaeus t o l d us of the generation of the planets, and stated:

When each of the bodies required to help in the production of time had been
brought into its proper motion, as generated living creatures having their
bodies bound with living bonds and having learnt their duties, they moved
according to the motion of the different.

Now one m i g h t attempt to draw a distinction here between the planets as

202 Plato's Philosophy of Science
l i v i n g creatures and the fixed stars as demigods, b u t i n fact the t w o terms
are used interchangeably of each. Timaeus 40b talks of the fixed stars as
l i v i n g creatures w h i l e 40cd talks of the planets as demigods. The stability
of the heavens is guaranteed by the qualitative stability of the stoicheic
triangles allied w i t h the 'indissoluble bonds' w i t h w h i c h they are bound
together. T h r o u g h the mechanism of the variable d u r a b i l i t y of the bonds
between the stoicheic triangles and between more complex entities, the 47

Timaeus theory can t h e n account for any degree of qualitative flux at the
perceptual level, up to and i n c l u d i n g the t o t a l stability of the heavens.
Aristotle's comments iriDe Caelo andDe Generatione et Corruptione are
interesting i n relation to the nature of the stoicheic triangles. A t De Caelo
306a9, A r i s t o t l e gives us an i m p o r t a n t general principle:

Sensible things require sensible principles, eternal things eternal, and per­
ishable things perishable, and generally something of the same kind
underlies.

T h a t of course is quite different from the atomism of Leucippus and

Democritus, and, as the context makes clear, quite different i n Aristotle's
view from Plato as well. T h a t geometrical atomism is Aristotle's target
here at 306a ff. cannot be doubted. No one else held a theory of analysis
into surfaces (306al-4), t h a t some b u t not a l l the elements could trans­
m u t e i n t o each other (306a4-6), and t h a t earth was the odd m a n out
(306al8-21). Those who 'resolve into surfaces', according to Aristotle have
incorrect first principles (306a8-9). Presumably those are t h a t sensible
48

things do not have sensible archai (though this is not stressed) and t h a t
perishable things have stable archai. These people stand on the t r u t h of
t h e i r archai and w i l l not admit t h a t t h e i r principles should be criticised i n
the l i g h t of t h e i r consequences.
H a v i n g discussed geometrical atomism i n the latter p a r t of De Caelo
I I I . 7 , A r i s t o t l e opens I I I . 8 by saying t h a t i n general i t is unsound to t r y to
give shapes to the simple bodies (ta hapla somata, 306b3 ff.). T h a t is a very
interesting phrase w h i c h w o u l d suggest t h a t 'those who resolve i n t o
surfaces' do posit specifically shaped simple bodies. A r i s t o t l e goes on to
raise the 'packing' problem, t h a t such shapes could not form a p l e n u m as,
of the solids, only the square and the p y r a m i d can do this and more shapes
t h a n these are required.
De Generatione et Corruptione is even more emphatic on these matters.
A t 315b25 ff. Aristotle poses w h a t he feels is the key question, do things
undergo change and g r o w t h (alloioutai kai auxantai) because of indivisible
magnitudes (megethos adiaireton)? Those advocates of indivisible magni­
tudes are Leucippus and Democritus who suppose t h e m to be bodies, and
the Timaeus w h i c h supposes t h e m to be surfaces. 316a2-4 is also i m p o r t a n t
here, as Aristotle denies t h a t p u t t i n g surfaces together can account for any
change of quality. Clearly he does not believe t h a t the Platonic surfaces
 7. Geometrical Atomism - Flux and Language 203
undergo any sort of alloiosis. De Generatione et Corruptione 325b25 ff., as
we saw i n Chapter 1 states t h a t Leucippus holds t h a t the indivisibles (ta
adiaireta) are solids and are u n l i m i t e d i n shape, and Plato holds t h a t they
are surfaces and definite i n number. The evidence of Aristotle t h e n is very
m u c h i n favour of the idea t h a t the stoicheic triangles do not undergo any
sort of alloiosis.

V I . T h e ageing process
One objection to the idea t h a t the stoicheic triangles undergo phora b u t
not alloiosis m i g h t be based on the theory of ageing of Timaeus 8 1 ff.
According to this, ageing and n a t u r a l death are to be explained by the
gradual deterioration and consequent loss of function of the particles
w h i c h constitute the body. One interpretation is t h a t i t is change i n (and
indeed the loss of) the intrinsic order of the stoicheic triangles t h a t is at
the root of this sort of deterioration. There is an alternative interpretation,
though, which is at least as plausible. This is t h a t while the bonds between
stoicheic triangles w h i c h form complexes may deteriorate, as may those
between the complexes w h i c h form atoms, as may the complexes them­
selves, there is no deterioration i n the i n t r i n s i c order of the stoicheic
triangles. A t 81b5-9 the Timaeus begins its discussion of ageing by saying:

When the organisation (sustasis) of the whole animal is new, the triangles
(trigona) which constitute the atoms being 'fresh off the stocks' (ek druo-
chon), they have strong joins between one another (pros allela).

I t is i m p o r t a n t t h a t the bonds t h a t the Timaeus describes here are pros

allela. These must be bonds which b i n d either stoicheic or complex t r i a n ­
gles to each other, rather t h e n any bonds i n t e r n a l to the stoicheic triangles.
T h a t the stoicheic triangles undergo change i n t h e i r bonding relations
does not entail t h a t they undergo any change of intrinsic order. A t Timaeus
81c6-el we are subsequently told:

When the sides of the triangles loosen (he hriza ton trigonon chala) due to
having contested many battles in a great amount of time ... finally the bonds
(desmoi) fitting the triangles of the marrow together no longer hold out
against toil, but separate.

A t t h i s point death occurs. I f these bonds are pros allela, t h e n the ageing
process can take place w i t h o u t any qualitative change of the stoicheic
triangles, he hriza ton trigonon chala is undoubtedly a rather strange
phrase. I would agree w i t h Cornford t h a t this 'curious metaphor' m u s t
contrast w i t h sunkleisin at Timaeus 81b7 and t h a t this strongly favours
Taylor's view t h a t hriza means side, i n the sense of the side of a complex
w h i c h w i l l form one face of an a t o m . 1 would agree w i t h Taylor against
49

M a r t i n and A r c h e r - H i n d t h a t chala means loosens (the sides) rather t h a n

204 Plato's Philosophy of Science
blunts (the angles). Again, there is no need to suppose any change i n the
50

intrinsic order of the stoicheic triangles.

R e t u r n i n g to 81b5-9, one m i g h t take the view t h a t as the triangles are
ek druochen 'fresh off the stocks' they have been newly created. However, 51

i f we b u i l d a wooden boat the wood does not undergo any change qua wood.
I t is merely reorganised from loose timber to a boat. The wood may t h e n
52

be 'fresh off the stocks' w i t h o u t having undergone any qualitative change

i n itself, and so too perhaps w i t h animals and stoicheic triangles. The verb
t h a t Plato uses i n t h i s context may support t h i s approach. This is
sunistemi, w h i c h has more of a sense of ordering t h a t w h i c h is already
existent t h a n of creation de novo. I t is notable t h a t this verb is used i n
53

the earlier discussion of atomism for the formation of atoms and complex
triangles out of complex and stoicheic triangles respectively, where no 54

change i n the stoicheic triangles is required or mentioned. So too at

Timaeus 57c9, where as we saw there was the question of whether there
were different sizes of stoicheics, the most reasonable i n t e r p r e t a t i o n w o u l d
have sunistemi referring to how stoicheics are arranged i n relation to each
other, rather t h a n t h e i r i n t e r n a l arrangement.
E v e n i f we were to take ek druochon as signifying some qualitative
change, i t is not clear t h a t i t is stoicheic triangles t h a t do the changing. I f
the referent of trigona at 81b6 is complex triangles, these can pass i n t o and
out of existence w i t h o u t prejudice to the stability of the stoicheics. The 55

correct referent of trigona is a problem w h i c h runs t h r o u g h the entire

discussion of ageing. Certainly there are some uses here t h a t can only be
references to complex triangles, e.g. the Timaeus talks of'the purest type
of triangles, smoothest and oiliest' (82d). As there are only two types of
stoicheics, and they could not be described as smoother or oilier t h a n one
another, this must be a reference to complex triangles. So too at 81d, 56

where we are told t h a t the process of n u t r i t i o n involves the triangles of the

body c u t t i n g and assimilating incoming triangles. Conversely, w h e n we
age and the bonds between our triangles become weak, the incoming
triangles cut up our o w n . 57

Now i t may be t h a t the reference of trigona slips between stoicheic and

complex triangles. However, i n the earlier discussion of how the complex
triangles are formed from the stoicheics at Timaeus 53c ff., Plato always
has Timaeus carefully m a r k the distinction between the two, opening each
discussion of the stoicheics w i t h an explicit reference to t h e i r n a t u r e . I n58

this context Timaeus 57d and 61a are important, as i n both of these
passages there is a distinction made between the stoicheic and the complex
triangles, the stoicheics being referred to i n each case as to stoicheion
triangles and the complex triangles simply as trigona. As some of the
references of trigona i n the discussion of ageing are clearly to complex
triangles, and there is at no point any distinction d r a w n between these
references a n d any other, one m i g h t conclude t h a t a l l the references of
trigona i n the discussion of ageing are to complex triangles. We now
 7. Geometrical Atomism - Flux and Language 205
have two tools w h i c h w o r k i n harmony for analysing statements such as
Timaeus 89c:

For plainly in the beginning the triangles (trigona) of each animal are
organised (sunistatai) with the power to last only up to a certain time.

I n a l l likelihood sunistemi refers to the organisation of existent particles,

and trigona to complex triangles. So i t is complexes of stoicheics w h i c h
undergo intrinsic change, w h i c h is consonant w i t h the idea t h a t i t is the
pros allela bonds t h a t are subject to temporal limitations. T h a t the t r i a n ­
gles can l a s t up to a certain time' (Timaeus 89d) m i g h t be a temporal
l i m i t a t i o n , or more likely i n view of Timaeus 81b ff., a 'wear and tear'
l i m i t a t i o n . I f we take the reference to be to complex triangles, neither
require change i n the stoicheic triangles. O n this interpretation, the
'biochemical' processes of ageing and n u t r i t i o n are merely specialised
cases of the more general 'chemistry' t h a t is discussed at Timaeus 53c ff.
They require no more t h a n the division and reconstitution of atoms and
complex triangles. I n this respect one m i g h t note the s i m i l a r i t y of the
b a t t l i n g imagery between Timaeus 53c ff. and 81b ff., as well as the similar
uses of sunistemi. Compare Timaeus 56e where we are told:

Whenever a small amount of fire is enclosed in a large amount of air, water

or earth, moving within their movements, battling and breaking up upon
being defeated, two units of fire are assembled to make one form of air. 59

I f we take the t w o images of b a t t l i n g to refer to the same sort of process,

t h e n i t is the atoms being broken into t h e i r faces t h a t constitutes old age,
rather t h a n any deterioration i n the stoicheic triangles. I f the theories of
ageing and n u t r i t i o n require something beyond the general theory of
chemistry, one w o u l d expect this to be m a r k e d and discussed. There is no
mention of change i n the stoicheic triangles and good reason i n the b a n on
the t r a n s m u t a t i o n of the scalene and isosceles triangles to believe t h a t no
such change occurs.

V I I . T h e gold e x a m p l e
Against t h i s general line of thought, one m i g h t argue t h a t the gold
example explicitly denies any sort of permanence to shapes, and to t r i a n ­
gles i n particular. There are several strategies which one m i g h t employ
against t h i s view, though. The first is to argue, as I have above, t h a t
geometrical atomism introduces a quite different theory of the nature of
the w o r l d and how we m i g h t solve the problem of referring to changing
phenomena to t h a t of Timaeus 48e-53b. I t may well be the case t h a t
everything is transient except the receptacle i n 48e-53b. T h a t seems to r u n
into difficulties though, and perhaps the determinate and stable shapes of
206 Plato's Philosophy of Science
the stoicheia are introduced specifically to circumvent such problems.
U l t i m a t e l y these are different and incompatible theories w h i c h the read­
ers may have to choose between, i f they w i s h to have a coherent theory,
and I shall say some more on the nature of t h a t choice and other possible
s i m i l a r choices i n the Timaeus i n Chapter 9.
I f one is seriously w o r r i e d about w h a t the gold example says about
shapes, t h e n i t is possible to abandon the idea of stable shapes w h i l e s t i l l
holding there to be some stable stoicheia, these being higher archai,
perhaps lines, perhaps points. I would rather emphasise the conditional
n a t u r e of the gold example though. So perhaps the key clause is at Timaeus
50b5-6:

I f (ei gar) someone were to mould all the shapes out of gold and without
stopping remodel each of these into the rest...

The conditional here may indicate t h a t this is a thought experiment rather

t h a n a direct statement of cosmology. I f someone were to do this remodel­
l i n g , t h e n we w o u l d appear to have something a k i n to a Heraclitean flux,
where the only permanence would be found i n the gold itself, and this is
supposed to explicate the flux of the elements and t h e i r relation to the
receptacle, and so the need for a reform of language. W i t h t h a t I have no
quarrel, t h o u g h one m i g h t wonder though w h a t w o u l d be the case i f this
hyperactive character was not perpetually engaged i n this remodelling.
One reason for introducing this i n conditional manner may be a k i n to the
various descriptions of chaos and of the m a t e r i a l nature of the receptacle
we are given. Plato may w i s h to make some general comments on how we
can t a l k about the w o r l d i f we suppose there to be a receptacle, whatever
degree of flux we suppose the w o r l d to be i n , up to and i n c l u d i n g t o t a l flux.
T h a t does not commit Plato to the view t h a t everything i n the w o r l d is i n
such a radical flux, though i f some things are we can s t i l l refer to t h e m i n
a certain manner. Even w i t h geometrical atomism i t is quite possible t h a t
some things (including the shapes i n the gold example, b u t not a l l things)
may be i n a r a p i d flux, b u t we are s t i l l able to refer to t h e m though i n a
slightly different manner to t h a t suggested i n Timaeus 48e-53b.
The gold example presents us w i t h a hypothetical manic remodeller of
gold. We are not t o l d w i t h i n 48e-53b w h a t n a t u r a l processes this remodel­
ler is an analogue for. Does he find a correlate i n geometrical atomism?
W i t h the four-element theory, i t is clear t h a t there is a remodelling effect
even i f the cause of this is not specified. W i t h geometrical atomism,
though, we can perhaps be rather more specific. First, there is the effect
w h i c h the atoms of each element have on each other, either as a general
function of t h e i r shape (fire cuts more because of its size and sharp angles,
61d ff.), or the effect a much greater amount of an element has w h e n i t
surrounds a smaller amount of another element (it assimilates i t to its o w n
k i n d , 57a-c). I f we ask w h y these changes should come about, Timaeus has
 7. Geometrical Atomism - Flux and Language 207
two answers. A t 57c the receptacle is s t i l l i n motion and s t i l l separating
out the elements by a like-to-like principle, so there is motion of the
elements and hence contact and change. A t 58a ff. we find t h a t the
revolution of the same has a compressing effect on the elements, such t h a t
atoms are pushed together and so again we have change. A l l this w i l l
produce a remodelling of the elements, b u t a l l this takes place w i t h i n the
standard theory of physical change, and there is no need here to suppose
t h a t a n y t h i n g remodels the stoicheic triangles.
L e t me emphasise t h a t I do not w i s h to deny t h a t Plato has a theory of
flux i n the Timaeus. The w o r l d we see is i n a constant flux w i t h earth,
water, air and fire a l l appearing to change into each other. The question is
how radical t h a t flux is at root. I f earth, water, air and fire are the roots
t h e n t h a t flux is indeed a radical one. Zeyl makes the interesting comment:

All commentators agree that Plato intends to show (a) that the constituents
of the physical world ('phenomena') are caught up in constant change (as is
forcefully argued in the case of the 'elements' at 49b7-c7) and (b) that this
fact necessitates a reform in the use of certain locutions as referring expres­
sions. 60

One m i g h t get the impression from Zeyl's remarks t h a t for Plato the
phenomena are a l l of the physical world, and t h a t the elements of 49b ff.
are the basic constituents of the phenomena. This may be so for the
four-element theory of 49b ff., b u t is simply not the case for 53c ff. where
the elements are decomposed to atoms, planes and stoicheic triangles, a l l
of w h i c h are far below our perceptual thresholds, such t h a t there is more
to the physical w o r l d t h a n the phenomena. Surely one of the great glories
61

of geometrical atomism is t h a t i t decouples the nature of w h a t we see from

the nature of the micro-world. Plato exploits this fact to great effect i n his
explanations of phenomena i n the latter part of the Timaeus, and not least
of course i n the very essence of geometrical atomism itself, t h a t the
micro-world is populated by these geometrical entities. The most basic of
these m a y w e l l be qualitatively stable. T h a t w h a t persists t h r o u g h change
should be different from t h a t w h i c h we see changing is beyond doubt, b u t
t h a t does not i m p l y w h a t persists must be indeterminate, merely t h a t i t
should be different, as geometrical atomism allows. I t is possible t h e n to
62

take a view of geometrical atomism, flux and the receptacle such t h a t there
is enough stability i n the w o r l d to u n d e r w r i t e the complete stability of the
heavens, i f the stoicheic triangles are qualitatively stable and the bodies
of the gods (i.e. the heavenly bodies, mostly fire) are p u t together w i t h
alutois desmois, indissoluble bonds, as at Timaeus 43a2, desmoi being
w h a t l i n k the stoicheic triangles into planes and the planes into solid
figures i n geometrical atomism.
208 Plato's Philosophy of Science

V I I I . Geometrical atomism and Socrates' dream

One objection t h a t has been raised against the idea t h a t the stoicheic
triangles are the basic physical entities for Plato is t h a t this would deprive
h i m of any analysis of matter. I n m y view this is not so. W h i l e the
63

stoicheic triangles may be the end of an analysis by physical decomposi­

tion, w h i c h m i g h t be t a k e n as one sense of giving an account of a physical
entity, there may be other sorts of account which we can have of the
stoicheic triangles. A l t h o u g h Timaeus talks of the stoicheic triangles as the
archai of the more complex triangles, he also goes on to t a l k of higher
archai. A t 48c3 he speaks of'the arche or archai of everything', and later
at Timaeus 53d, referring to the theory of stoicheic triangles he says:

These we hypothesise (hupotithemetha) as the principle (archen) of fire and

of the other bodies ... but the principles (archas) of these which are higher
(anothen) are known only to God and whoever is friendly to him.

These higher principles need not refer to some more basic physical or
mathematical entities, however. I t is notable t h a t Timaeus refers here to
higher archai, principles or s t a r t i n g points, rather t h a n to further stoic-
heia. A t the t e r m i n a t i o n of explanation i n terms of analysis by decomposi­
t i o n i n t o physical elements, one option for Plato is to s w i t c h to a
teleological analysis. These higher archai t h e n may be teleological expla­
nations of w h y the stoicheic triangles are as they are. One interesting
aspect of the passage at Timaeus 53d is t h a t the use of hupotithemi and
anothen is highly reminiscent of Phaedo 10 I d , the line allegory of the
Republic, and the defence of hypotheses by higher hypotheses. One m i g h t
compare here the uses of anothen at Phaedo 101d7, Republic 514b2 and
518b4, and the uses of hupotithemi at Phaedo l O l d ff. and Republic 510b
ff., and iheRepublic on line and cave, especially 511a5 where the soul m u s t
ton hupotheseon anotero ekbainen. One need h a r d l y stress t h a t the Phaedo
and the Republic are not looking for further physical hypotheses at this
point. The final clause concerning the m a n friendly to God m i g h t also
suggest some i n t e n t i o n to invoke explanations i n terms of the good. W h e n
we look at the sort of arguments t h a t Timaeus uses to justify his choice of
stoicheic triangles this is borne out. A t Timaeus 54a ff., we are told t h a t
w h i l e the isosceles triangle has b u t one form, the scalene has many and
we must postulate the fairest, and teleological language permeates this
entire section. I t is l i k e l y t h e n t h a t the stoicheic triangles are the most
elementary constituents of the physical w o r l d of the Timaeus, w i t h o u t
being the u l t i m a t e explanatory archai.
The geometrical atomism of the Timaeus is very interesting i n relation
to the passage of the Theaetetus at 20 I d ff. k n o w n as 'Socrates' dream'. As
w i t h Meno's paradox, the dream poses wide and deep philosophical ques­
tions, m u c h broader t h a n the themes for science t h a t I can tackle here, b u t
 7. Geometrical Atomism - Flux and Language 209
one issue for the study of the physical w o r l d posed by the dream is the
question of letters a n d syllables a n d the accounts we can have of the
simple and the complex. We can give an account of the complex i n t e r m s
of i t s component parts, b u t we cannot give a s i m i l a r account of t h e
simple.
According to Socrates' dream at Theaetetus 20 I d f f , the regress is finite
and terminates w i t h basic elements, the stoicheia, w h i c h lack accounts. I f
they lack accounts, t h e n they are perhaps unknowable. One problem for
science t h a t comes out of this is as follows. We can give an account of
complex physical entities, such as the elements earth, water, air and fire,
i n terms of t h e i r component parts. Doubtless the dream has a good deal to
say on the nature of t h a t account, v i a the question of Hesiod's wagon at
Theaetetus 207a, and whether a list of component parts is adequate, or an
account of the i n t e r r e l a t i o n of those parts, etc. There is a problem, though,
w i t h the physical stoicheia. However successful our physical analysis/
account of complex physical entities, we cannot give the same sort of
analysis/account of physical stoicheia. As far as the physical w o r l d is
64

concerned, i t is notable t h a t the Timaeus argues t h a t there are basic

elements, the stoicheic triangles, b u t at the t e r m i n a t i o n of analysis from
complex to simple these can be given teleological accounts. I t is interest­
65

ing, and surely no coincidence, t h a t the discussion of the Timaeus is

presented w i t h the same letters and syllables terminology. W h i l e letters
and syllables may be inappropriate i f we take the receptacle to be the key
basic i t e m , as argued i n section I of this chapter t h a t is not so i f the
stoicheic triangles are basic. They are the letters, the complexes of stoi-
cheics the syllables, and the elements are neither letters nor syllables i n
line w i t h Timaeus 48b. B u t the Timaeus would w a n t to deny t h a t the
66

elements of the physical w o r l d are perceptible, the contrary being claimed

at Theaetetus 202b. I t is i m p o r t a n t t h e n t h a t the Timaeus allows us some
account of these stoicheia as otherwise we would have no access to t h e m .
There is not t h e n a specific problem here w i t h the physical w o r l d ; the
physical 'atoms', here t a k i n g 'atom' i n its more l i t e r a l sense as indivisible
fundamental entity, can have accounts as w e l l as names, though of course
there may s t i l l be a problem w i t h logical atomism, w h i c h w o u l d beset any
claim to k n o w l e d g e . 1 do not w i s h to suggest t h a t this is Plato's solution
67

to the more general philosophical problems raised by the dream w h i c h r u n

deeper t h a n this, although i t may point us i n certain directions. W h a t I do
w i s h to b r i n g out is t h a t here we see a further epistemological reason for
teleology, to add to the cosmological and nomological reasons we have seen,
and of course the further epistemological reasons related to underdetermi­
n a t i o n and Meno's paradox. I n this sense, Plato's geometrical atomism is
68

a long way removed from t h a t of the Presocratics. So I w o u l d disagree w i t h

Vlastos w h e n he compares Plato to Empedocles, Anaxagoras and the
atomists, and says: 'So i n offering us his own stoicheia Plato shows t h a t he
has enrolled i n the same programme of physical elementarism.' A t one 69
210 Plato's Philosophy of Science
level t h a t is so, b u t Plato's atomism has a far richer structure b o t h
physically and philosophically.
The Timaeus, then, may have a teleological account of the stoicheic
triangles at the end of an account of physical entities i n terms of decompo­
sition. To look forward to the Philebus for a moment, another sort of
account of the stoicheic triangles may be available there. I n the Philebus,
we can t h i n k of points on a continuum as the imposition of l i m i t on the
u n l i m i t e d . I f we take two points on a continuum, we can t h e n define a
70

specific q u a n t i t y or distance between them. I f we do this i n t w o dimen­

sions, we can t h e n have an account of all the plane figures. So we arrive at
a point where the Timaeus could take over w i t h an account of how the
physical w o r l d is constituted from t w o specific types of plane figures, the
half-equilateral and the isosceles stoicheic triangles. I p u t this as generally
as possible as there is a question as to w h a t the triangles of the Timaeus
are triangles of and so too a question i n the Philebus concerning whether
per as and apeiron are constitutive of things or an analysis of things.
Whichever way we take these questions, the Philebus would be able to give
a further account of the stoicheic triangles. I t is w o r t h n o t i n g t h a t the
Philebus accepts four-element theory, and t h a t at Timaeus 53b we are t o l d
t h a t w h e n the demiurge orders the universe from its p r i m o r d i a l chaos,
where a l l is devoid of'reason and measure' (alogos kai ametros) (53a9), he
first 'created shapes by means of forms and numbers' (eidesi te kai arith-
mois) (53b4-5). 71

Geometrical atomism would seem to leave us i n a rather better position

t h a n the receptacle i n terms of explanation. The stoicheic triangles are
themselves explicable i n terms of a further teleological account, i n a way
i n w h i c h the characterless and basic receptacle was not. So too the stoi­
cheic triangles are explanatory of the phenomena, as the Timaeus from 53c
ff. goes on to show. They are constitutive of other things w i t h o u t i m p e r i l ­
l i n g t h e i r own stability i n a way t h a t the receptacle found difficult. We can
have an account of how they fit together to make more complex entities,
and how the particles of our bodies interact w i t h other particles to create
sensations.

IX. Geometrical atomism a n d language

One of the key consequences of there being some basic, stable entities i n
the Timaeus is t h a t the problem of a collapse of discourse concerning the
w o r l d around us is avoided. As G i l l puts i t , whatever t u r n s out to be stable
gives the w o r l d sufficient permanence for language to get a grip on i t . The 7 2

receptacle appears to be fatally flawed as a candidate for this role. I t

cannot be both constitutive of the phenomena and unchanging as i t is
required to be i n order to license to toiouton references to changing
phenomena. I f the stoicheic triangles are the basic, stable entities i t w o u l d
appear t h a t they can be properly constitutive of changing phenomena, and
 7. Geometrical Atomism - Flux and Language 211
this does not i m p e r i l t h e i r stability. Thus reference can be made to the
changing phenomena. T h a t w i l l be so for a n y t h i n g else w h i c h is qualita­
tively stable, w h i c h m i g h t w e l l include the heavenly bodies. So as Sophist
262e demands, our sentences w i l l have subjects, and I w i l l say something
more about false judgement i n the Timaeus i n Chapter 9. 73

This may underestimate the grip t h a t language can get on the w o r l d of

the Timaeus, though. I n several late works (Theaetetus 201d ff., Politicus
277d ff., Sophist 253a ff., Philebus 18b ff.) Plato employs the letters and
syllables analogy to i l l u s t r a t e some points about the nature of language,
and arguably i n some or a l l of these cases about the nature of the w o r l d . 74

Plato is often interested i n which letters do or do not combine to make

syllables, and is concerned w i t h the bonds (desmoi) between letters. 75

Consider t h e n w h a t sort of problem geometrical atomism is introduced

to solve, and how geometrical atomism is expressed. The original problem
is a puzzle about the correct use of language and the structure of the world,
and i t is introduced using the letters and syllables analogy. Geometrical
atomism is expressed i n terms of stoicheia - we are given an account of
w h i c h stoicheic triangles w i l l combine w i t h others, w h i c h w i l l not, and how
this combining occurs - and the bonds between the triangles are referred
to as desmoi. 16
Indeed the Timaeus throughout seems very interested i n
the question of mathematical bonds, the bonds w h i c h hold m e n and gods
together, the bonds w h i c h hold soul and body together and the bonds w h i c h
hold the physical stoicheia together. 77

The receptacle of 48e-53b m i g h t be thought to provide the ontological

counterpart to a subject-predicate analysis of the world. Qualities appear
i n the receptacle and the proper description of these is t h a t x is f-like,
where x is the receptacle, f-like is the quality and F the form t h a t this p a r t
of the receptacle is p a r t i c i p a t i n g i n . The receptacle always appears i n the
subject p o s i t i o n a n d has no qualities i n i t s e l f . I t h a l t s the regress of
78

Socrates' d r e a m because i t is u l t i m a t e l y basic ontologically, a n d is a

pure subject l i n g u i s t i c a l l y ( e v e r y t h i n g is predicated to the receptacle, i t
is predicated of n o t h i n g ) . The cost t h o u g h is h i g h , as we have seen. The
receptacle is inexplicable (we can say n o t h i n g positive about t h e recep­
tacle) a n d non-explanatory (its r e l a t i o n to w h a t is i n i t is h i g h l y
problematic).
W i t h geometrical atomism though we have a different model of the
w o r l d and perhaps a different conception of the way t h a t language relates
to the w o r l d . W h a t is not often recognised is how strong a reductive
79

account of the w o r l d the Timaeus gives us. Timaeus 61c-68d gives an

analysis of each of our senses i n terms of geometrical atomism. Thus for
instance w h i t e is the effect of the larger fire particles w h i c h dilate the
visual stream (67e), and hot is due to the sharpness and r a p i d motion of
the fire particles (61e). This t h e n is no longer a simple subject-predicate
arrangement as i t m i g h t be i f we were t a k i n g the receptacle to be the sole
basic e n t i t y (i.e. (this p a r t of) the receptacle is white, is hot). H o t and w h i t e
212 Plato's Philosophy of Science
are now related to the motion and arrangement of certain types of atom,
w h i c h i n t u r n can be given analyses i n terms of (the motion and arrange­
m e n t of) complexes of stoicheics and u l t i m a t e l y stoicheic t r i a n g l e s
themselves.
This brings us back to letters and syllables again. Instead of an isomor­
p h i s m between a subject/predicate analysis of language and a
substance/attribute ontology, we m i g h t look to an isomorphism between
the way t h a t the basic components of language can be arranged and
structured into something meaningful and the basic components of the
w o r l d can be arranged and structured into something beautiful and good.
Can language pick out the basic components and relations of the world?
T h a t w o u l d seem eminently possible. I n the same way t h a t we m i g h t say
t h a t 'Theaetetus flies', and be w r o n g or r i g h t about that, we m i g h t say t h a t
'(this atom of) fire cuts (this earth atom)' or 'this stoicheic triangle forms a
bond ( w i t h t h a t one)'.
I t is interesting to note t h a t relative to the discussions of language i n
the Theaetetus and Sophist w h i c h are keen to stress t h a t some words,
letters and syllables do, or do not, fit together properly, so the Timaeus is
keen to stress t h a t some of the components of the w o r l d and some com­
plexes of those components do, or do not, fit together properly. The
Philebus is interested i n the relationship between language and music at
17a ff. I n both w h a t is i m p o r t a n t i n producing a meaningful result is how
the components are l i n k e d and arranged together. A n interesting l i n k
between the Timaeus and this passage is t h a t Philebus lie warns us, after
the example of music b u t i n a quite general manner, t h a t an indeterminate
p l u r a l i t y leaves you w i t h an indeterminate account, while Timaeus 55c
makes the same sort of remarks about the number of worlds and the
number of elements. 80

The l i n k s to other late works are of course rather speculative. However,

i f we are to l i n k the various discussions of letters and syllables together,
and t h a t i n itself seems a reasonable t h i n g to do, t h e n I see no reason w h y
the Timaeus should not be i n c l u d e d i n t h a t project, especially as the
results are so i n t e r e s t i n g . E v e n i f we stay w i t h i n the Timaeus, geomet­
r i c a l a t o m i s m , i f we take the stoicheic t r i a n g l e s to be the basic entities,
can be seen to provide a different conception of the w o r l d a n d i t s
r e l a t i o n to language to the receptacle passages, i f we take the receptacle
to be the basic entity. As indicated, t h i s avoids the receptacle paradox
a n d the p r o b l e m of w h e t h e r the receptacle is either explicable or ex­
planatory, a n d language appears to be given a reasonable g r i p on the
w o r l d . I f we t a k e t h i s view, t h e n geometrical a t o m i s m e i t h e r replaces
r a d i c a l l y modifies the receptacle passages of 48e-53b. The f u r t h e r i m ­
plications a n d advantages of t h a t move, a n d w h e t h e r we can j u s t i f y i t
as a r e a d i n g of t h i s section of the Timaeus, w i l l be the subject of the n e x t
chapter. I f the Timaeus postulates the stoicheic t r i a n g l e s as stable,
basic entities t h o u g h , t h i s has considerable r a m i f i c a t i o n s for the n a t u r e
 7. Geometrical Atomism - Flux and Language 213
of f l u x envisaged ( i n r e l a t i o n to both the phenomenal and the micro
worlds), and for nature of the relationship between language and the
world.
 8

Geometrical Atomism - Matter

and Space

I n the previous chapter, we examined the relationship between geometri­

cal atomism, flux and language. I argued t h a t the receptacle, i n an
ontology of forms, receptacle and phenomena, is incapable of providing a
licence for to toiouton references to the phenomena. I f the receptacle is i n
some way constitutive of the phenomena, its own unchanging nature is
compromised and i t cannot license such references, nor can i t license such
references i f i t is not constitutive of the phenomena. N o r can theories
w h i c h postulate visible and changing u l t i m a t e constituents of the universe
succeed here. Plato's solution appears to be geometrical atomism, a r a d i ­
cally different theory w h i c h postulates stable micro-entities i n a new
ontology of forms, receptacle, micro-entities and phenomena. As these
micro-entities can combine and t h e n dissociate to create the phenomena
w i t h o u t i m p e r i l l i n g t h e i r own stability, the problem of to toiouton refer­
ences to changing phenomena seems to be solved.
Geometrical atomism, however, does not dispose of the receptacle, and
so we s t i l l have the problem of the nature of the receptacle and how i t
relates to w h a t is i n i t . This chapter t h e n w i l l deal w i t h the relation of
geometrical atomism, m a t t e r and space. As the stoicheic triangles do not
undergo any change, they do not need something more basic, such as the
receptacle, to constitute them. So perhaps the best conception of the
receptacle for geometrical atomism is a theory of the space i n w h i c h the
entities of geometrical atomism move, combine and dissociate, rather t h a n
some theory of a m a t e r i a l substrate. However, the evidence for considering
the receptacle to be i n some way m a t e r i a l is considerable. T h a t should be
no surprise, i f at some stage we are being i n v i t e d to consider the receptacle
as i n some way constitutive of the phenomena. I n this chapter, though, I
argue t h a t we are presented w i t h a different conception of the receptacle
after the i n t r o d u c t i o n of geometrical atomism. Before the i n t r o d u c t i o n of
geometrical atomism we are given descriptions w h i c h would suggest
either t h a t the receptacle is a m a t e r i a l substrate, or t h a t i t has both
m a t e r i a l and spatial features. After the i n t r o d u c t i o n of geometrical
atomism, we find spatial descriptions only, and no m a t e r i a l analogues.
Aristotle's evidence is interesting i n regard to this, as he appears to have
 8. Geometrical Atomism - Matter and Space 215
a change i n attitude between the Physics and De Caelo and De Generatione
et Corruptione.
A second theme of this chapter w i l l be the importance accorded to
geometrical atomism i n the Timaeus, something t h a t is often under-appre­
ciated. The Timaeus from 53c is given over to the nature of geometrical
atomism and a geometrical atomist analysis of the phenomena. We are
also given a detailed analysis of qualities and perception entirely i n terms
of the geometrical atomist constitution of our bodies and the external
w o r l d . T h a t seems to be of considerable significance, especially against any
theory t h a t qualities occur directly i n the receptacle. There are some
serious difficulties w i t h the qualities idea anyway r e l a t i n g to the localisa­
t i o n of qualities, the coherence of entities and the nature of causation.
Perception also raises the question of the relation of the Timaeus and the
Theaetetus again. I argue t h a t there are considerable differences between
these works on this matter, and t h a t the position offered by the Theaetetus
is u n l i k e l y to be Plato's own.

I. T h e receptacle and metaphors

There is t h e n a difficulty concerning whether the receptacle is supposed to
be space, or matter. Certainly at first glance, i t would seem t h a t we are not
given an entirely consistent account, as there appears to be a m i x t u r e of
strongly m a t e r i a l and strongly spatial descriptions of the receptacle. The 1

evidence for the receptacle as matter is t h a t i n the gold analogy the

receptacle is referred to as t h a t 'out of which' things are formed (specifi­
cally ek krusos, 'out of gold', 50a6). So too i t is referred to as a moulding-
stuff or wax (ekmageion, 50c2) or soft m a t e r i a l (malakon, 50e8), and as
mother (meter, 5062, 51a4) and nurse (tithene, 49a6, 52d6). I t is like an2

odourless base for perfumes (50e). The receptacle is also t h a t of w h i c h p a r t

is ignified and liquefied (to pepuromenon meros, 51b4, cf. 52d5) to produce
phenomenal fire and water, and so w o u l d seem to be a m a t e r i a l constitu­
ent. This w o u l d at least seem to invite us to consider the receptacle to be
3

m a t e r i a l , especially since, as emphasised i n the previous chapter, the

receptacle appears to need to be constitutive of phenomena i n some
manner i n order to license to toiouton references to changing phenomena.
A further question here is whether this collection of m a t e r i a l metaphors
can be made to yield a consistent account of a m a t e r i a l receptacle.
O n the other hand, the evidence for the receptacle as space is t h a t the
receptacle is specifically or i m p l i c i t l y referred to as space (chora, 52b4,
52d3, 53a6, cf. 58a7) and as a seat or place (hedra, 5 2 b l , 53a2) and as t h a t
' i n w h i c h ' things occur (en hoi, 49e7, 5 0 d l , en hautoi 50e5, en tini topoi
52a6, 52b4). I t is also referred to as a winnowing-basket (plokanon, 52e6,
4

cf. 57c. 88de), the motion of w h i c h sorts the particles i t contains. This is a
5

p a r t i c u l a r l y strong spatial metaphor, as i t would suggest t h a t the particles

are independent of b u t contained by the winnowing-basket.
216 Plato's Philosophy of Science
So perhaps Plato treats the receptacle as both space and matter? I
w o u l d agree w i t h A l g r a t h a t i t is u n w a r r a n t e d to suppose, w i t h o u t some
investigation and argument, t h a t Plato held essentialist assumptions
about space and m a t t e r (i.e. the essential nature of space precludes i t
being m a t t e r and/or the essential nature of m a t t e r precludes i t being
space). Especially so, one m i g h t feel, as Descartes held t h a t there could be
no body w i t h o u t extension and no extension w i t h o u t body, and so provides
a good example of a non-essentialist. Here t h e n is one strategy for
6

a t t e m p t i n g to give Plato a consistent account of the receptacle.

As we shall see, a good many of the metaphors Plato uses are open to a
wide variety of interpretation. The l i s t of m a t e r i a l and spatial metaphors
I gave above should be regarded as a p r e l i m i n a r y one based on first
impressions of the metaphors, as some metaphors can be t a k e n to be either
m a t e r i a l or spatial or m a t e r i a l and spatial. We m i g h t also be able to argue
t h a t different metaphors are meant to b r i n g out different aspects of the
receptacle and so are not inimicable to each other.
H a v i n g said this, I do not believe t h a t t r e a t i n g the receptacle as m a t t e r
is consistent w i t h some of the spatial metaphors w h i c h are employed. Here
I a m not overly concerned about whether i t is possible to reconcile the ' i n
w h i c h ' {en hoi) and the 'out of which' (ex hou) expressions. We m i g h t have
7

a theory where the receptacle is m a t t e r and is space and these expressions

answer different questions, say where is i t and w h a t is i t . I a m r a t h e r more
concerned t h a t the theory of geometrical atomism requires there to be
empty space and t h a t Plato seems quite aware of this requirement. The
winnowing-basket metaphor, w h i c h is critical to the functioning of the
cosmos, is not only one t h a t would seem to separate m a t t e r a n d space b u t
on a straightforward interpretation would require there to be empty space
between the particles so t h a t whatever is winnowed w i l l not pack to fit the
space i n the basket fully. I w i l l adopt a different strategy on the nature of
the receptacle, w h i c h is t h a t Plato presents us w i t h a m a t e r i a l or a
material/space theory of the receptacle prior to the introduction of geomet­
rical atomism, and a space theory after i t s introduction.
As commented i n the previous chapter, i t is of the utmost importance to
be able to relate 48e-53b to geometrical atomism. T h a t is so on the general
8

interpretational principle t h a t Plato wrote his works as wholes were we

need to be able to see the relation of each p a r t to the whole. So too there
appear to be several cues i n 48e-53b w h i c h prompt such a l i n k . There are
comments about shapes, shaping and triangles, there is the promise to
explain the receptacle/contents more fully (50c6) and there is the l i n k
between 48e-53b at 53b7-c3 w h i c h states t h a t now we w i l l look at the
disposition and origin of the elements.
 8. Geometrical Atomism - Matter and Space 217

I I . T h e Timaeus a n d empty space

I f Plato recognises the existence of empty space, this w i l l be i m p o r t a n t
evidence on the relation of space and m a t t e r and the nature of the
receptacle. O n this m a t t e r I a m largely i n agreement w i t h O ' B r i e n . A t
9

Timaeus 58a ff., Plato appears to say both t h a t there is no void, as the
revolution of the a l l compresses the elements, and 'allows no void space to
be left' (kenen choran oudemian eai leipesthai), and t h e n t h a t there are
voids (kenonta) and emptinesses (diakena) between the atoms of the
elements. This is far from being the only instance where we have such a
discrepancy. One can also look to Timaeus 59al-2, 60cl-2, 7 9 b l , 79b3,80c3
for a denial of a void, and Timaeus 60e5, 6 1 a l , 6 1 b l and 61b4 for the
existence of gaps between atoms. As O'Brien points out, the context here
is i m p o r t a n t . A t Timaeus 58a4 ff., Plato is seeking to explain w h y the
10

elements do not separate out completely, b u t are continually i n motion and

are changing into one another. There are two options for giving Plato a
consistent account. W i t h Cherniss, one can deny the existence of gaps
between the atoms. As O'Brien points out, though, this is mathematically
11

false, even i f one assumes many sizes of atoms, and Plato seems quite sure
there are such gaps, and t h a t breaking up particles w i l l result i n empty
12

spaces. Timaeus 58b is most easily read as recognition of a packing

problem as smaller atoms force t h e i r way into gaps left by larger ones. I t
is interesting t h a t at Timaeus 81b6 we are told:

The blood particles, being chopped up within us and being surrounded by the
living creature as by a heaven, are compelled to imitate the motion of the
universe. So when each of these divided entities moves towards its kin, i t fills
again the emptied space (kenothen).

I find i t h a r d to believe t h a t Plato was not aware t h a t his theory of atoms

and how they move and change i m p l i e d t h a t there must be gaps between
atoms and spaces w h e n they change. Certainly Aristotle is aware t h a t of
this sort of problem, and he asserts:

It is entirely irrational to attempt to assign shapes to the simple bodies,

primarily because when placed together they will not fill the whole. I t is
supposed that there are three plane figures which can fill space, the triangle,
the square and the hexagon, and among solids only two, the pyramid and the
cube. Of necessity they need more than these as they consider there to be
more elements. 13

Aristotle is of course aware of Plato's denial of a void, e.g. at De Generatione

et Corruptione 325b30 ff., though i t should be noted t h a t this is i n the
context of a comparison w i t h Leucippus.
There are two points I w o u l d add to O'Brien's critique of Cherniss, b o t h
based on the fact t h a t this is supposed to be a dynamic theory a t t e m p t i n g
218 Plato's Philosophy of Science
to account for motion and change. First, i t is one t h i n g to provide a n
arrangement whereby the atoms can be made to tessellate i n three dimen­
sions. I t is another to be able to account for the motion of the atoms i n t h a t
way, as the arrangement may w e l l be one t h a t locks r a t h e r t h a n allows
m o t i o n . Secondly, w h e n atoms decompose they do so from three-dimen­
14

sional entities into two-dimensional entities, and however b r i e f the t i m e

t a k e n to recompose, d u r i n g t h a t t i m e there m u s t be gaps. One m i g h t add
t h a t even i f the first arrangement of atoms tessellates i n three dimensions,
there is no guarantee t h a t after a t r a n s m u t a t i o n the new arrangement of
atoms w i l l continue to tessellate.
The alternative approach is to question Plato's denial of a void. A t 58a7
we ought to take kenen choran oudemian eai leipesthai to mean t h a t the
revolution of the whole tends to leave no empty space r a t h e r t h a n pre­
cludes the existence of empty space. W h y is t h i s so? Because the
15

revolution of the whole has a compressive effect (sphingei, 58a7) and so

brings the atoms together as far as possible, so there are no large-scale
voids and motion b u t there may be gaps between atoms. I agree w i t h
O'Brien against Taylor and Cornford, who argue t h a t as simple circular
m o t i o n does not produce such an effect, Plato cannot have countenanced
i t , t h a t Plato supposes j u s t such an effect. This is the n a t u r a l reading of
16

sphingei, 'bind t i g h t , press together' at 58a7. W h y should he suppose t h a t

to be so? We m i g h t look to Empedocles Fr. 38.4 where there is a possibly
s i m i l a r usage ofsphingon.17
The ancients t h o u g h were aware of the action
of a whirlpool i n moving heavy bodies to its centre. A t De Caelo 295a ff.
A r i s t o t l e states t h a t those who argue t h a t the e a r t h is i n the centre of the
cosmos because of the action of a vortex around i t argue from observations
of w h i r l p o o l s . Indeed, gravitational theories based on this idea were
18

common i n the seventeenth century, notably t h a t of Descartes. Plato, on

this basis, m i g h t well believe t h a t circular motion i n certain circumstances
does indeed create some form of compressive effect. However Plato envis­
ages the compressive effect working, the conclusion m u s t be t h a t w h i l e
there are no large-scale voids for Plato, there are gaps between atoms and
w h e n they decompose, and so for Plato there is empty space. T h a t there is
empty space for Plato goes very m u c h against the view of the receptacle as
matter, or space and matter.

III. Material metaphors

As Plato appears to have empty space as a consequence of geometrical
atomism, i t w o u l d seem t h a t the receptacle can be only space. T h a t t h o u g h
w o u l d seem to r u n up against a good deal of the evidence cited i n section
one w h i c h w o u l d seem to indicate t h a t the receptacle is material. L e t me
suggest t w o strategies for allowing Plato a consistent account. F i r s t , one
m i g h t t r y to argue t h a t the receptacle is at no stage intended to be
m a t e r i a l . As noted earlier, there is a good deal of flexibility i n how we
 8. Geometrical Atomism - Matter and Space 219
m i g h t i n t e r p r e t many of the material-looking descriptions, so perhaps i t is
possible to interpret a l l of these i n spatial rather t h a n m a t e r i a l terms.
Secondly, one m i g h t argue t h a t j u s t as geometrical atomism provides a
quite different solution to the problem of how we can licence references to
changing phenomena, so too we are given a different conception of the
receptacle w i t h geometrical atomism. So perhaps we are given m a t e r i a l or
spatial/material descriptions of. the receptacle up to the i n t r o d u c t i o n of
geometrical atomism, b u t only spatial metaphors after its introduction.
Let us begin by examining the metaphors w h i c h give rise to the idea of
the receptacle as matter, and see i f they are susceptible to some other
interpretation. The receptacle is described as the 'nurse of a l l becoming'.
Certainly one could take this as a m a t e r i a l metaphor, t a k i n g the nurse to
be a wet nurse giving m a t e r i a l sustenance to w h a t comes to be i n the
receptacle. There is though nothing w h i c h compels us to take this meta­
phor i n such a manner. Space as m u c h as m a t t e r m i g h t be described as a
nurse (everything has to be somewhere, 5 2 b l ) , and the occurrence of
tithene at 52d5 is sandwiched between and l i n k e d to the occurrence of
chora at 52d2 and the description of the receptacle as a plokanon at 52e6,
the strongest of the spatial metaphors. Another possibility is t h a t the
receptacle is a nurse i n the sense of rocking its charges. Tithene at 52d5
leads up to the description of the receptacle shaking the things t h a t are i n
i t , and 88de has the trophon kai tithene tou pantos, the 'rearer and nurse
of the universe' always moving and agitating the cosmos. This w i l l help
keep the cosmos i n good order, j u s t as proper exercise w i l l help to keep the
h u m a n body i n good health. We are by no means compelled to take the
description of the receptacle as a nurse to mean t h a t i t is i n some way
material.
T h a t the receptacle is described as a mother (meter), and at 50d there is
also a father and some offspring looks more of a difficulty. However, this
metaphor is introduced by Timaeus stating t h a t we must conceive of three
things, t h a t w h i c h becomes, t h a t i n w h i c h i t becomes (en hoi, 5 0 d l ) and
the model i n whose likeness t h a t w h i c h becomes is born. So the receptacle
as mother is supposed to explicate the en hoi r e l a t i o n . Cornford argues
19

t h a t a common view of conception i n the ancient w o r l d was t h a t the male

supplied the seed and the female supplied the place. Such a view is 20

mentioned by Aristotle, who debates the issue. A l g r a argues t h a t Corn­

21

ford's references to other views do not decide the issue. W h a t m u s t go a

22

long way to settling the issue though is the Timaeus' account of conception.
I f we look to Timaeus 91a ff., we find t h a t the male indeed supplies the
seed (like the f r u i t from a tree, apo dendron karpon, 91d) while the w o m a n
supplies the place i n which i t grows ( i t is sown as i n a t i l l e d field, arouran,
91d2), b u t does not supply any matter. The meter metaphor on this
i n t e r p r e t a t i o n t h e n looks more spatial t h a n material. I t may also be the
case t h a t neither the nurse or mother descriptions are meant to i m p l y t h a t
the receptacle is spatial or material, b u t are merely intended to b r i n g out
220 Plato's Philosophy of Science
the fact t h a t the phenomena are to be thought of as some form of u n i o n
between forms and the receptacle.
I f we take ekmageion (50c2) to be a moulding-stuff, out of w h i c h we
make things t h e n this looks very much a m a t e r i a l metaphor. So too one
m i g h t draw parallels w i t h the malakos, the soft m a t e r i a l of 50e8, though
i f we do draw this parallel i t is i m p o r t a n t to note t h a t the strongly spatial
en hautoi at 50e5 precedes malakos, as does en hoi at 50d6. Ekmageion is
not necessarily a moulding stuff out of w h i c h we model things b u t may also
be wax i n w h i c h one marks a seal. This is certainly the sense of ekmageion
i n the wax block model of the m i n d i n the Theaetetus. The verbs here are
23

i m p o r t a n t as well. The ekmageion is moved and shaped, and i n the next

sentence the verb tupo is used. Now tupo can have the sense of to stamp
24

or impress as well as to model. Tupo is again used i n the context of the soft
m a t e r i a l (50d4, 50d6), b u t so is apomatto at 50e8, w h i c h has more of a
sense of stamp or impress. So we m i g h t take this as a wax i n w h i c h we
stamp or impress figures, when the metaphor again looks rather more
spatial. One m i g h t again argue t h a t the point of these metaphors is
n e i t h e r s p a t i a l nor m a t e r i a l , b u t merely to emphasise the charac­
terlessness of the receptacle.
The gold example is seen by many to be the strongest m a t e r i a l meta­
phor, as here we have shapes being moulded and re-moulded out of gold
(ek chrusou, 50a6). M o h r has argued t h a t i t is the 'exigency of the meta­
p h o r ' t h a t produces the impression of a m a t e r i a l substrate, b u t I w o u l d
25

t e n d to agree w i t h A l g r a t h a t we can hardly take the gold analogy any

other way. One strategy we m i g h t employ is to argue t h a t some meta­
26

phors appear to be m a t e r i a l u n t i l the introduction of geometrical atomism

w h e n a non-material explanation becomes available. Consider the refer­
ences to the receptacle being ignified and liquefied (Timaeus 51b4, cf.
52d5). We are told t h a t 'that part w h i c h is made fiery each t i m e appears
as fire' (51b4-5). As A l g r a comments, i t is h a r d to make sense of the idea
t h a t phenomenal fire is an ignified part of the receptacle unless the
receptacle is some form of m a t e r i a l constituent. Indeed, i t is h a r d to see
27

how one m i g h t ignify or liquefy space, and so this looks strongly material.
W i t h the introduction of geometrical atomism, though, one m i g h t argue
t h a t w h a t this means is t h a t fire atoms have either moved into or been
formed i n this area of space.
However, w i t h the gold example there is no question of w a i t i n g for
geometrical atomism for a full explanation. Here we are t o l d t h a t on
account of this analogy we can have the safe answer of gold and the other
answer of to toiouton. This can only be the case i f the gold is constitutive
of the phenomena i n some manner, and so we are led to some form of
m a t e r i a l conception of the receptacle. E a r l i e r I argued t h a t i t is i m p o r t a n t
to recognise t h a t the gold example is prefaced by a conditional (ei gar,
50a5), and is a thought experiment rather t h a n a direct cosmological
statement. Thus the radical flux supposed here need not be supposed
 8. Geometrical Atomism - Matter and Space 221
elsewhere. B u t the gold example is supposed to explain something, and
t h a t surely is w h y we can refer to changing things as to toiouton, and the
answer to t h a t clearly enough is i f there is an unchanging m a t e r i a l
substrate, t h e n perhaps we can.
So too the idea of the receptacle as being like an odourless base for
perfumes seems to invite a m a t e r i a l interpretation. Whether t h a t is the
same r e l a t i o n of receptacle to form and phenomena as the gold example
could be debated at length. Here one m i g h t argue though, as w i t h the wax
and soft m a t e r i a l metaphors t h a t the point here is neither spatial nor
m a t e r i a l , b u t to emphasises the characterlessness of the receptacle as a
suitable receiver.
I t is h a r d to avoid impression reading Timaeus 48e-53b t h a t we are at
least being i n v i t e d to t h i n k about the possibility of the receptacle as a
m a t e r i a l entity. I f there is a t r a n s i t i o n to spatial metaphors, is i t possible
t h a t is relatively gentle? I t is noticeable t h a t m a t e r i a l metaphors predomi­
nate at the beginning of 48e-53b, and t h a t there is a t r a n s i t i o n to the
predominance of spatial metaphors near the end. We m i g h t t r y t h e n to
reformulate the earlier m a t e r i a l metaphors i n acceptable spatial terms. I
w o u l d argue t h a t the choice is rather more stark. I do not t h i n k we can
successfully reformulate a l l the m a t e r i a l metaphors. Rather, 48e-53b pre­
sents us w i t h a t r i p a r t i t e ontology of receptacle, forms and phenomena,
w i t h a theory of the relation between flux and language and an appropriate
conception of a material/spatial receptacle. 53c ff. presents us w i t h a new
ontology w i t h the addition of the entities of geometrical atomism, a new
theory of the relation of flux and language and, as I argue i n the next
section, an appropriate conception of the receptacle as space.

IV. S p a t i a l m e t a p h o r s
As a second strategy, we m i g h t accept t h a t there are metaphors w h i c h
indicate t h a t the receptacle is material, b u t offer the following argument.
I f we conduct the discussion of the constituents of the w o r l d solely at the
perceptual level, t h e n there are no physical stoicheia, as earth, water air
and fire appear to change into each other. I f so, t h e n i n order to make any
proper references to the phenomena, there must be a relationship between
the receptacle and the elements w h i c h supports such references, such t h a t
the receptacle must be i n some way constitutive of the elements, or i n other
words is matter, or is b o t h m a t t e r and space. So we get a m i x t u r e of
metaphors between Timaeus 48e and 53c.
Now consider the s i t u a t i o n w h e n geometrical atomism is introduced at
Timaeus 53c ff. We now have physical stoicheia w h i c h are non-perceptible,
the stoicheic triangles. I t is quite possible t h a t i t is t h e i r stability w h i c h
underwrites proper references to the visible and changing elements, and
the stoicheic triangles are certainly constitutive of the elements. After the
introduction of geometrical atomism then, there is no need for the recep-
222 Plato's Philosophy of Science
tacle to be constitutive of anything. I t can be j u s t space. The way i n w h i c h
geometrical atomism is set up, w i t h two-dimensional entities forming up
to create three-dimensional entities, and three-dimensional entities w h i c h
cannot be fitted together w i t h o u t gaps between, seems to demand such a
view. I t is also interesting to note t h a t after the introduction of geometrical
atomism, we find plenty of space-type descriptions, b u t no matter-type
descriptions at a l l . We can find a continuation of the plokanon metaphor
from 52e at 57c, w i t h the only mention of the receptacle after the introduc­
t i o n of geometrical atomism at 57c7 (tes dechomenes) and 88de where i t is
nurse and rocks the universe. As noted above, the winnowing-basket is
perhaps the strongest of the spatial metaphors as the particles i n i t are
independent of i t and i t is t h e i r container, and looks p a r t i c u l a r l y apposite
for geometrical atomism. So too we can find mention of chora i n m a n y
places after Timaeus 53c, and indeed hedra
28
We can also find m e n t i o n
29

of kenos at 58b3 asserting gaps, at 59al-2, 60cl-2, 7 9 b l , 79b3, 80c3

denying a void, and diakenos at 58b5, 60e5, 6 1 a l , 6 1 b l and 61b4 asserting
gaps. Meter is not used again although tithene is once at 88d6, where as
we have seen, the nurse agitates and keeps the universe i n motion for the
sake of its order. There is no need to take 'nurse' as at any t i m e i n d i c a t i n g
t h a t the receptacle is material. Ekmageion is used once at 72c for the
relation of spleen and liver, malakos is used several times either as 'soft'
i n contrast to h a r d , or i n relation to objects or as describing soft parts of
30

the h u m a n body. 31

I t is also possible to combine these two strategies. Perhaps this inter­

pretation of 53c ff. does not have to struggle against a m u l t i t u d e of
m a t e r i a l metaphors i n 48e-53c, i f we take mothering as spatial due to the
Timaeus' account of conception, n u r s i n g as rocking, and ekmageion/
malakos to be wax i n w h i c h we stamp impressions, or to be really about
characterlessness rather t h a n matter or space. I n w h i c h case one w o u l d be
left w i t h the gold example, and ignification and liquefaction. I f so, t h a t
m i g h t w e l l focus our attentions on w h a t I have argued are the key issue
here, w h i c h is whether the receptacle can be unchanging itself and yet be
constitutive of changing t h i n g s . 32

As emphasised earlier, i t is of critical importance to be aware of the

differences between Timaeus 48e-53b and 53c ff. The first theory operates
only at the level of perception for physical entities, w h i l e he second
postulates micro-structure. The second theory postulates more entities
t h a n is p e r m i t t e d by the t r i p a r t i t e ontology of the first. The second theory
corrects the first on t r a n s m u t a t i o n . I t is quite possible t h e n t h a t the second
theory involves a reformulation of the relation between m a t t e r and space.
To say t h a t this is the position of the Timaeus, or indeed Plato's position,
w o u l d perhaps be going too far, however, as this reformulation is not made
explicit, even i f some of the other differences between 48e-53b and 53c ff.
are evident and marked. B u t Plato may w i s h his readers to be active here
 8. Geometrical Atomism - Matter and Space 223
and consider t h e i r own attitudes to the theories presented, and the possi­
ble problems and solutions.
I t may be t h a t the reader is given (at least) three theories here, elements
on t h e i r own, elements and receptacle as matter/space, and geometrical
atomism. I say at least three, as i t may be the case t h a t Timaeus 49d6 is
genuinely ambiguous between whether we call fire t h i s or the suchlike, or
we call t h i s or the suchlike fire, and asks the reader to make the choice of
how to i n t e r p r e t i t . Clearly the elements on t h e i r own is an inadequate
theory, as i t does not license references to the phenomena. W h a t of the
receptacle as m a t t e r and space? Timaeus 50bc, as we have seen, seems
flatly contradictory. The receptacle does not depart from its own nature nor
does i t take on any shape, yet i n the next sentence i t is changed and shaped
by the things w h i c h enter i t , and this is 'hard to explain and marvellous'
(dusphraston kai ihaumaston, 50c6), and there is a promise to follow this
up later. Yet the second attempt at describing the receptacle ends w i t h the
conclusion t h a t the receptacle is invisible and shapeless (anoraton kai
amorphon, 51a8) and its relation w i t h the intelligible is 'most difficult and
h a r d to grasp' (aporotata ... kai dusalotaton, 5 1 b l ) . Timaeus 48e-53b t h e n
leaves us i n a radical aporia (impasse) concerning the receptacle. The
receptacle also appears to be inexplicable and non-explanatory. Perhaps 33

t h e n we are meant to at least question whether the receptacle cannot do

everything t h a t is asked of i t , and i f this is an adequate account of the
relation between matter, space and the phenomena. Timaeus 53c ff. t h e n
gives us the materials for a reformulation.
I f we are being generous to Plato, he may w e l l have seen some of the
problems inherent i n an undifferentiated theory of m a t t e r where m a t t e r
and space are identified, and may also have seen something along the lines
of Berkeley's criticisms of Locke on matter, t h a t i f w h a t is under discussion
has no characteristics i t is otiose, and i f i t has characteristics i t is not
matter. T h a t geometrical atomism can provide coherent answers to the
major problems w i t h the receptacle as m a t t e r theory seems more t h a n
coincidental.

V. T h e r e c e p t a c l e a n d q u a l i t i e s
There is a very i m p o r t a n t b u t often underappreciated fact about the
Timaeus. This is t h a t 53c t h r o u g h to 89d is given over to an analysis of how
geometrical atomism can account for a l l the phenomena w h i c h we per­
ceive. I t is not about how qualities occur i n the receptacle, b u t about how
qualities m i g h t be thought of i n terms of a geometrical atomist analysis of
our bodies and of external bodies. Given w h a t I have argued about the
relationship of the theories of Timaeus 48e-53b to geometrical atomism,
t h a t w o u l d seem to be highly significant. This is especially so as there are
certain problems w i t h the view t h a t qualities occur directly i n the recep­
tacle t h a t geometrical atomism m i g h t be thought to solve.
224 Plato's Philosophy of Science
There is a problem, w h i c h I shall call the localisation problem, i f
qualities occur directly i n the receptacle. The receptacle, i t is emphasised
is amorphos and although this is not explicitly stated, given its charac­
terlessness we m i g h t reasonably expect i t to be partless (ameros) as well.
Forms are not spatial entities, so they too are amorphos and we w o u l d
reasonably expect t h e m to be ameros as well. So w h e n the receptacle
partakes of the intelligible w h y is i t t h a t only a p a r t of the receptacle is
affected rather t h a n the whole receptacle? W h a t is i t t h a t spatially l i m i t s
and localises qualities w h i c h occur i n the receptacle? The only other
resource here is the demiurge, b u t t h a t looks highly problematic. I t is not
i n the least b i t clear how the demiurge m i g h t achieve localisation unless
we go to Timaeus 53b4-5 where the demiurge begins to generate shapes by
means of forms and numbers. T h a t though would seem to point us directly
at a geometrical atomism account of localisation. As the demiurge retires
after the ordering of the cosmos, i n order for localisation of phenomena to
continue w h a t he does at this stage must be permanent. Again, t h a t looks
problematic unless we go to Timaeus 53b.
O n a geometrical atomism analysis of the basis of qualities, w h i c h
appears to be w h a t we are given, the localisation of phenomena is straight­
forward. The demiurge creates spatially l i m i t e d entities, the stoicheic
triangles, w h i c h move around i n space and combine w i t h each other to
form more complex entities such as the four elements. So this area can be
described as fiery because i n this area the stoicheic triangles have formed
up i n t o tetrahedra, and the interaction of these tetrahedra w i t h our bodies
gives rise to the qualities w h i c h we associate w i t h fire.
One m i g h t question the relation of the stoicheic triangles to the recep­
tacle here. I f one were desperate to hang on to the receptacle as m a t t e r
theory, t h e n one m i g h t postulate the stoicheic triangles as triangles i n the
receptacle whose matter is provided by the receptacle. As they move, t h e i r
shape slips across this matter. Thus the stoicheic triangles w o u l d be
merely configurations of the receptacle. This would not solve the original
problem of the receptacle ( i f the receptacle constitutes the stoicheic t r i ­
angles and so the phenomena, i t changes as the stoicheics move/the
phenomena change, i f i t does not constitute them, i t cannot license stable
references), and there is a further serious objection. There is a problem i n
differentiating between 'proper' and 'phantom' shapes on the configura­
tions view. Here, i n two dimensions, we have some regular shapes w h i c h
have not tessellated. O n the geometrical entities and space view ('proper'
shapes shaded) this is unproblematic. We can easily differentiate between
the entities and the empty space, as i n Fig. 55 (left). So we have t w o types
of regular entities and four regular b u t small spaces and an i r r e g u l a r
central space. O n the r i g h t though, w i t h the shading removed, where we
have a configured receptacle, w h a t t h e n makes the regular shapes any
more (or less) configurations t h a n the small regular shapes or the central
i r r e g u l a r shape?
 8. Geometrical Atomism - Matter and Space 225

Fig. 55. A selection of two-dimensional shapes in the receptacle, on the entities and
space view, and on the configurations view.

There is no mention of any irregular shapes i n the receptacle once the

demiurge has done his work. Such shapes w o u l d have to be as causally
efficacious i n t h e i r own way as the 'proper' shapes yet we are given no
account of t h e m whatsoever.
A second problem along these lines is as follows. I n F i g . 56 on the left,
w o r k i n g on the geometrical entities and space model, we have, unproble-
matically, eight squares and some space. O n the r i g h t though, using the
configurations view, suddenly we have 9 squares, as the central square is
no more or less a square t h a n any of the others. The idea t h a t the entities
of geometrical atomism are no more t h a n configurations of the receptacle
is untenable.

Fig. 56. A counting problem on

the entities and space view and
the configurations view.

There is also a problem w i t h the coherence of entities and t h e i r move­

ment i n the receptacle, i f we are t h i n k i n g only i n terms of qualities. I f an
e n t i t y is only a collection of qualities, one m i g h t be w o r r i e d as to w h y this
collection of qualities stays grouped together as the e n t i t y moves. A g a i n ,
the demiurge is not going to be an adequate answer here, for w h i l e he may
set up the order of the cosmos, he does not m a i n t a i n i t . Another resource
available is the like-to-like principle, perhaps i n alliance w i t h the compres­
sive principle generated by the motion of the same. As these principles are
expressed quite generally, one m i g h t seek to apply t h e m to qualities,
t h o u g h one m u s t note t h a t this is not how these principles are applied i n
226 Plato's Philosophy of Science
the Timaeus and i t does look odd to have such a principle operate w i t h
qualities. I f the like-to-like principle does w o r k operate w i t h qualities, one
m i g h t w o r r y w h y one gets complex entities w i t h many qualities rather
t h a n simple patches of like qualities. One m i g h t also w o r r y about how a
tiger keeps its stripes, not only when i t is i n motion and we are concerned
as to w h y a l l these qualities move together, b u t also when i t is stationary.
Surely the black of the black stripes and the gold of the gold stripes w i l l
be affected by the like-to-like principle and the tiger w i l l dissociate i n t o
patches of l i k e qualities, i f there is n o t h i n g to prevent this happening? A t
the level of qualities, there does not appear to be a n y t h i n g w h i c h w i l l
prevent such a dissociation.
On a geometrical atomist account, the coherence and movement of
bodies is not a problem. The stoicheic triangles have coherence and
permanence w h i c h is not affected by t h e i r motion. The complex entities
w h i c h they form have varied permanence depending on the strength of the
bonds, as we saw i n the last chapter.
One m i g h t also consider the question of the nature of causal interac­
tions i f we have only qualities i n the receptacle. Perhaps one could p u t
together a theory of causation w i t h qualities, though one w o u l d have to be
careful t h a t this was not merely a regularity theory and t h a t i t w o u l d
satisfy the demand of the Republic's cave allegory t h a t there should be
something more t h a n the phenomena such t h a t we can escape the cave.
A g a i n , though, this is not the way the account of the Timaeus goes. We do
not have qualities interacting w i t h qualities, b u t causation appears to take
place between the various entities of geometrical atomism w h i c h give rise
to qualities.
C e r t a i n interpretations of the receptacle r u n i n t o trouble w i t h causa­
tion, and the problems seem most severe for Cornford's images i n a m i r r o r
view of the receptacle. Take two b i l l i a r d balls on a table and a m i r r o r image
of this. I r o l l one b a l l towards the other one, and i t bumps into and moves
i t . The b i l l i a r d balls have a causal relation w i t h one another, b u t do the
images i n the mirror? I f the images are images of forms, surely we don't
w a n t to say t h a t forms are causally related to one another? O n a geomet­
rical atomist account, causal interaction is not a problem. The entities of
geometrical atomism can quite happily interact w i t h one another (come
into contact, decompose each other), and provide a basis for changing
qualities.
A good deal of the m o t i v a t i o n for the view t h a t the receptacle contains
qualities, I suggest, comes from an attempt to align the Timaeus w i t h the
first p a r t of the Theaetetus. 34
The discussion of whether knowledge is
perception generates the Theaetetus/Protagoras/Heraclitus position of a
radical flux of qualities. Whether t h a t is Plato's own position is a m a t t e r
of debate. The Timaeus, as we have seen, differs from the Theaetetus i n
35

p e r m i t t i n g us to refer to changing entities as to toiouton. We m i g h t w i t h

Owen consider t h a t to be a 'lame plea' and so date the Timaeus earlier, or
 8. Geometrical Atomism - Matter and Space 227
investigate w h y the Timaeus m i g h t allow such references. The least
deviation from the Theaetetus is the idea of the receptacle containing
qualities, b u t on many grounds t h a t appears highly problematic. The
introduction of geometrical atomism looks much more promising b u t t h e n
the Timaeus looks m u c h less like the Theaetetus. There is more to come,
as there are radical and irreconcilable differences between the theories of
perception advanced i n the Timaeus and Theaetetus.

V I . P e r c e p t i o n i n the Timaeus and Theaetetus

There are t h e n problems w i t h the idea t h a t the receptacle contains quali­
ties, and indeed this is not the account of perception t h a t we are given at
Timaeus 61c ff. Perception i n the Timaeus is generated from the interac­
t i o n of geometrical atomist entities w i t h our bodies (which themselves are
constituted from geometrical atomist entities). This account is developed
i n some detail for each of the senses. Plato though also presents us w i t h a
theory of perception i n the Theaetetus. This raises two questions. Is the
Theaetetus view of perception the same as the Timaeus account? Is the
Theaetetus view of perception Plato's view? M y view is t h a t the account
given i n the Theaetetus is not Plato's view, and t h a t the account given i n
the Timaeus is radically different. Readers familiar w i t h Burnyeat's w o r k
on the Theaetetus w i l l recognise t h a t I am sympathetic to his 'B' reading of
the first p a r t of the Theaetetus i n w h a t follows (Plato does not accept the
Theaetetus/Protagoras/Heraclitus account of the world). I shall say more
about the relation of readings of the Theaetetus to how we m i g h t approach
i n t e r p r e t i n g the Timaeus i n the next chapter. There are of course alterna­
tive approaches. Taylor believes there are significant differences between
the Theaetetus and Timaeus b u t believes the theory to be Timaeus' a n d
not Plato's. Cornford believes the theory of perception i n the Theaetetus
36

to be Plato's own and not to have significant differences w i t h t h a t of the

Timaeus. 31
McDowell argues t h a t while there are differences between the
theories, these are not insuperable. 38

W h a t t h e n are the differences on perception between the Theaetetus and

the Timaeus? To begin w i t h , I agree w i t h Fine on some points of difference
between the two theories. I n the Theaetetus, there are no perceiver-inde-
39

pendent properties. A n object is w h i t e i f and only i f i t is perceived to be

40

white. I n the Timaeus though, the shapes and sizes of the entities postu­
lated by geometrical atomism are not perceiver-dependent. Indeed, the 41

very postulation of such entities w o u l d seem to r u n counter to the

Theaetetus. I f we p u t this i n terms of later terminology, the Timaeus
postulates p r i m a r y qualities where the Theaetetus does not.
W h a t of w h a t m i g h t be called secondary qualities, such as colours and
tastes? Here i t w o u l d seem t h a t there is some measure of agreement
between the Timaeus and the Theaetetus, at least on the idea t h a t percep­
t i o n is due to an i n t e r a c t i o n . I f Plato identifies colours as 'flames
42
228 Plato's Philosophy of Science
streaming off bodies h a v i n g particles so proportioned to the visual ray as
to yield perception', t h e n one m i g h t argue t h a t objects can (and do) do
43

this w i t h o u t being perceived, t h a t colours are thus perceiver-independent,

and so conclude w i t h Fine that:

The Timaeus allows, what the Theaetetus disallows, that something is white
when i t is not being perceived at a l l .
44

However, this is not a l l t h a t Plato has to say on the matter, and we are
subsequently t o l d t h a t particles of the same size as the visual ray are
transparent, while those t h a t are larger dilate i t and those t h a t are
smaller contract i t . Then at Timaeus 67e4-6 we are told:

Thus we should assign the name 'white' to that which dilates the visual ray,
and 'black' to the opposite. 45

This to some extent makes colours perceiver-relative again. Objects are 46

not w h i t e i n themselves, b u t emit fire particles of greater size t h a n those

of the h u m a n visual ray, and because of their effect on t h a t ray we perceive
w h i t e . One could easily imagine a creature whose visual ray consisted of
particles either slightly larger or slightly smaller t h a n those i n the v i s u a l
ray of humans. There w o u l d t h e n be some objects, e m i t t i n g particles close
i n size to those of these visual rays, which i t would see as w h i t e and we
w o u l d see as black, or vice versa.
This is not to say t h a t objects do not have perceiver-independent
properties. A n object may well have the property o f ' e m i t t i n g fire particles
w h i c h have the ability to dilate the h u m a n visual stream' or more simply
' e m i t t i n g fire particles of a specific size', whether observed by humans or
not. S i m i l a r l y hotness may only occur w h e n a sentient being comes into
contact w i t h fire atoms, b u t fire atoms have sharp angles and move swiftly
whether they are i n contact w i t h humans or not, as the Timaeus has t h e m
interacting w i t h other atoms unperceived by humans. So too objects w i l l
47

be sorted by the shaking of the receptacle and the compressive effect of the
same by a like-to-like principle whether they are perceived to be by
humans or not. Even i f we distinguish between weight as a perceptual
q u a l i t y generated by an interaction between ourselves and objects and
weight as an effect of physical processes, clearly objects (such as atoms)
w i l l have weight i n the second sense independent of perception. 48

There is though a more fundamental difference between the Theaetetus

and Timaeus on perception. Theaetetus 152c lays down two conditions for
knowledge. Knowledge must be unerring, and must be of w h a t is, such
t h a t perception, i f i t is to be knowledge, must both be u n e r r i n g and of w h a t
i s . One way of s u m m a r i s i n g the argument of the first p a r t of the
49

Theaetetus is as follows. I n order for perception to be u n e r r i n g , and for

50

the ' m a n is the measure' doctrine of Protagoras, perceptions need to be

51
 8. Geometrical Atomism - Matter and Space 229
relativised to persons and instances. A suitable ontology here w o u l d be one
of Heraclitean flux, where everything is changing i n every respect a l l the
t i m e . However, there is a problem w i t h such a flux, as we saw i n the
52

previous chapter. The Theaetetus is adamant t h a t we cannot refer to

entities i n t o t a l flux, and indeed at 157a ff. insists t h a t we cannot use the
verb 'to be', or any other name t h a t makes things stand s t i l l of these
changing entities. So i f perception is u n e r r i n g t h e n i t is not of w h a t is,
53

and so perception cannot be knowledge.

The Timaeus m i g h t w e l l agree on the prerequisites for knowledge, b u t
disagrees on w h y perception is not knowledge, and indeed reverses the
approach of the Theaetetus. The Timaeus is quite clear t h a t h u m a n percep­
t i o n ( i n contrast to t h a t of the w o r l d soul) is fallible. W h e n the revolutions
54

w h i c h constitute our minds are disturbed, t h e n we are prone to error. 55

Perception i n the Timaeus, though, is of w h a t is, at least to a greater extent

t h a n is allowed i n the Theaetetus. Unless we conclude t h a t both the
receptacle and geometrical atomism fail i n providing a licence for to
toiouton references to changing phenomena, t h e n the Timaeus allows
references to changing phenomena w h i c h the Theaetetus proscribes. I t is
interesting to note t h a t perception is not infallible i n other works. Cer­
t a i n l y the Republic does not treat i t so at 602d w i t h the example of things
w h i c h appear bent w h e n placed i n water, nor indeed does the Sophist i n
relation to m i r r o r images at 266c ff., nor the Philebus i n relation to seeing
at a distance at 38c.
The Theaetetus theory of perception t h e n is u n l i k e l y to be Plato's own.
I t seems more likely t h a t i t is a consequence of t r y i n g to make sense of the
idea t h a t knowledge is perception along w i t h the prerequisite t h a t k n o w l ­
edge must be infallible. The further consequence of t h a t is t h a t perception
cannot be of w h a t is and so the attempt to define knowledge as perception
fails. The theory of perception i n the first p a r t of the Theaetetus is radically
different to t h a t of the Timaeus.

V I I . T h e e v i d e n c e of A r i s t o t l e
The theory of the entities of geometrical atomism m o v i n g around i n space
seems to me to be an excellent one for someone i n fourth-century Greece
to hold. I t has the advantage of a t t r i b u t i n g to Plato a coherent theory of
the relation of flux, language, matter and space, and also suggests t h a t
Plato sees some of the problems w i t h the receptacle as m a t t e r theory. Here
I w i l l examine the evidence of Aristotle, w h i c h is commonly thought to
i n t e r p r e t the receptacle as a m a t e r i a l substrate. The evidence of Aristotle,
however, is not consistent, and may undergo a significant change between
the Physics and De Caelo and De Generatione et Corruptione. Certainly he
appears to give a classic statement of the receptacle as both m a t t e r and
space i n Physics IV.2, where he says:
230 Plato's Philosophy of Science
Plato identifies matter (hulen) and space (choran) in the Timaeus. For that
which participates (metaleptikon) and space are one and the same. His
account of that which participates differs in the unwritten opinions, but he
is consistent in identifying place (topon) and space. 56

W h i l e t h i s looks straightforward, one m i g h t raise two doubts here. Plato

does not use the w o r d hule at a l l i n the discussion of the Timaeus, so this
is an A r i s t o t e l i a n interpretation, subject to the usual worries about such
interpretations, rather t h a n a direct statement. The structure of this
passage is also interesting. One m i g h t read i t t h a t while Plato is consistent
between the Timaeus and the u n w r i t t e n opinions on the question of space
place he is not on the question of matter, place and space. Aristotle goes on
to make the key criticism against an identification of m a t t e r and place,
t h a t w h i l e place can be separated from an object its m a t t e r cannot be. 57

De Caelo I I . 7 presents a rather different picture. Aristotle is of course

w e l l aware of the gold example (see the passage from De Generatione et
Corruptione below). He discusses whether the t r a n s i t i o n from one element
to another m i g h t be effected by separation, and h a v i n g dismissed this
possibility, suggests two other theories:

Change of shape, just as a sphere or a cube might be generated out of the

same piece of wax (ek tou autou kerou), or the analysis into surfaces, as some
have said. 58

I t is clear from subsequent discussion of triangles t h a t he associates the

second theory, and not the first, w i t h Plato. T h a t is very i n t e r e s t i n g
because the first theory here looks very m u c h l i k e the gold example (note
the ek tou expression) and the subsequent wax and soft m a t e r i a l analogies.
Nor is there any question here of identifying m a t t e r and space. Especially
so, on m i g h t say, as Aristotle recognises t h a t the t r a n s i t i o n between
elements w i l l leave some i n d i v i d u a l triangles 'suspended'. Slightly later,
Aristotle does say that:

It is necessary for that which underlies (to hupokeimenon) to be devoid of

forms and shapes, so that i t will be more capable of modification, just like
the all receiving (pandeches), as is written in the Timaeus. 59

That, though, is i n the context of p u t t i n g forward his own theory of the

hupokeimenon, and t h a t Plato describes the receptacle as amorphos is
uncontentious. More i m p o r t a n t is w h a t use Plato may p u t the receptacle
to. De Generatione et Corruptione is interesting on these matters. Aristotle
tells us:

What is written i n the Timaeus is not well defined. I t is not at all clear if the
allreceiving (pandeches) is separate from the stoicheia (chorizetai ton stoi­
cheion) and he makes no use of this, after saying i t is a substratum
 8. Geometrical Atomism - Matter and Space 231
(hupokeimenon) prior to the so-called elements, just as gold things are made
out of gold. 60

T h a t is significant on several counts. Aristotle now seems to have serious

doubts as to whether Plato equates m a t t e r and space i f the elements m i g h t
be separate from the receptacle. Aristotle has also picked up the point t h a t
indeed Plato does not make any further use of the receptacle where one
m i g h t expect h i m t o . Aristotle goes on to say:
61

He produces an analysis of the elements which are solid as far as surfaces,

though i t is not possible for the nurse, the primary matter, to be the
surfaces. 62

So Aristotle recognises t h a t there is a problem w i t h the relation between

the receptacle considered to be m a t e r i a l and the stoicheic triangles, and
we m i g h t note i n this context the evidence we saw i n the previous chapter
on Aristotle's view t h a t the stoicheic triangles are basic, p r i m a r y and
undergo no alloiosis. 63

Aristotle's evidence then, from De Caelo and De Generatione et Corrup­

tione, at least recognises that:

a. There is a problem w i t h the relation of the receptacle and the

stoicheic triangles.
b. Plato's p r i m a r y analysis of the elements is i n t o planes and not into
m a t t e r like the gold example.
c. Plato describes b u t does not employ a m a t e r i a l receptacle.

One can see how the view Aristotle puts forward i n the Physics m i g h t have
arisen, especially given some of the similarities between the receptacle of
Timaeus 48e-53b and the analysis of Metaphysics V.8 and V I I . 3 , and the
Physics view seems to be a reasonable criticism of the receptacle as
matter/space theory. Identifying matter, space and place, or even j u s t
m a t t e r and space, is likely to lead to difficulties. T h a t assumption on its
own m i g h t not be so problematic, b u t i t seems to be coupled i n the
receptacle as matter/space theory w i t h the idea of a single, undifferenti­
ated m a t e r i a l substrate dealing w i t h change. I t is the combination of the
assumptions t h a t leads to the problem t h a t the receptacle cannot be
constitutive of phenomena w i t h o u t i m p e r i l l i n g its o w n status as a n u n ­
changing entity. Geometrical atomism of is very nature breaks w i t h these
assumptions, something Aristotle seems to come to realise.

V I I I . Plato, Democritus and Descartes

I m i g h t be accused of producing a 'Democritean' account of the receptacle
and its contents, something, w h i c h Cornford was vigorously opposed to
and, is held by some to have produced a definitive refutation. W h a t 64
232 Plato's Philosophy of Science
follows is an analysis of Cornford's arguments i n Plato's Cosmology pp.
198-210, h i g h l i g h t i n g where m y views differ from the Democritean ac­
count w h i c h Cornford attacks, and where they are similar, r e b u t t i n g his
objections. O n some points I shall be fairly brief as we have already dealt
w i t h the some of the relevant issues.
Contra Cornford (PC 199-200), as we have seen, shape is not a mistrans­
l a t i o n of morphe at Timaeus 50c 1. I n any case I am indifferent to w h a t is
i n the receptacle at this stage of the Timaeus. I do not read Democritus'
account into Plato on the basis of the plokanon, the w i n n o w i n g basket (PC
200-3). I agree w i t h Cornford t h a t there are considerable differences
between a plokanon w h i c h is shaken and other instruments w h i c h are
given a rotatory motion and so m i g h t be though to be like a Democritean
vortex. As argued i n Chapter 1, this may be significant as a move question­
ing the atomists r i g h t to assume a specific sort of motion. C e r t a i n l y I do
not t h i n k t h a t Plato accepts the Democritean account of a l l shapes and
sizes of atoms (PC 200).
Cornford puts forward his favourite analogy of the receptacle as a
m i r r o r (PC 200 f f . ) . As I have argued, there is no evidence for the
65

receptacle as a mirror, an analogy which Plato does not use for the
receptacle despite using many descriptions of the receptacle and despite
discussing m i r r o r s and m i r r o r images elsewhere i n the Timaeus. We have
66

also seen some of the problems Plato would have been saddled w i t h h a d
he thought the receptacle to be like a mirror. K u n g has argued t h a t the
receptacle is not a m i r r o r on three grounds: 67

1. For Plato, a m i r r o r is causally acted upon by the objects i t images,

w h i l e forms do not act and the receptacle is not acted upon.
2. A m i r r o r inherently distorts an image, m a k i n g left appear r i g h t , w h i l e
the receptacle ought not to.
3. For Plato, a m i r r o r neither contains nor constitutes its images.

A l g r a quite r i g h t l y argues t h a t i f we consider the receptacle to be l i k e a

mirror, we are not necessarily not committed to i t being l i k e a m i r r o r
(either as we understand mirrors, or as Plato did) i n a l l respects. This to 68

some extent disarms Strange's comment t h a t i t is odd t h a t Plato should

describe the relation between receptacle and forms as dusphrastos and
thaumastos w h e n he has already explained quite succinctly how m i r r o r s
w o r k . He m i g h t w i s h to deny t h a t the receptacle is like a m i r r o r i n this
69

respect, b u t t h e n one must ask i n w h a t respect is the receptacle l i k e a

mirror? The other major concern has to be t h a t i f the receptacle does not
constitute w h a t is imaged i n i t , as w i t h a mirror, t h e n the receptacle w i l l
not provide a basis for licensing to toiouton references to changing phe­
nomena, w h i l e i f the receptacle does constitute the phenomena i n some
manner t h e n i t would seem to be most u n l i k e a mirror.
According to Cornford (PC 204ff.), the Theaetetus gives us Plato's o w n
 8. Geometrical Atomism - Matter and Space 233
theory of sensible qualities and physical bodies, and we can use such
theories to help us interpret the Timaeus. This actually goes to the heart
of one of the key issues i n how we should look to interpret the Timaeus and
has some wide reaching implications, as I shall discuss i n the next chapter.
We have seen, however, t h a t there are considerable differences i n the
t r e a t m e n t of perception between the two works, and i t is by no means clear
t h a t the theories to w h i c h Cornford refers (Theaetetus 155d ff.) are Plato's
own.
Cornford has some very odd views on motion w i t h i n the receptacle (PC
205 ff.). I f I understand his view, i t is t h a t i f Plato postulated particles like
those of Democritus, they would be inanimate and so would not be capable
of motion, 'since no bodily change can occur w i t h o u t the self-motion of the
soul' (PC 205). This seems straightforwardly false. I t may be t r u e t h a t soul
is required to originate motion, and perhaps is required to m a i n t a i n
teleologically ordered motion (i.e. the heavens) b u t i t is not required to
m a i n t a i n motion itself nor is i t required for change to occur from such
motion. There is, on a l i t e r a l reading of the cosmogony, a problem about
the origin of motion prior to the w o r l d soul b u t not about the maintenance
of motion. I n relation to this, i t is interesting to note t h a t Cornford seeks
to deny t w o things. As we have seen, he denies t h a t the revolution of the
whole creates a compressive effect w h i c h forces the particles together (PC
243-6). He also wishes to deny t h a t the receptacle shakes or is shaken by
its contents. He says t h a t 'Plato cannot mean t h a t Space really shakes or
is shaken by the qualities' (PC 209), and so (wrongly) cuts off the t w o
70

critical sources of motion for inanimate particles.

Cornford also attempts to draw on the Politicus to support his view (PC
206 ff.). I t h i n k I have said enough i n Chapter 4 concerning the fundamen­
t a l cosmological differences between the Politicus and the Timaeus. One
simply cannot use the Politicus to help interpret the Timaeus w i t h o u t
recognition of these critical differences.
H a v i n g dissociated myself from Cornford's Democritean position, I do of
course recognise several similarities between Plato and the atomists,
notably on atomism, the void and perception. The critical differences
71

between Plato and the atomists w h i c h I would like to highlight, however,

are t h a t where the atomists postulate u n l i m i t e d shapes and sizes of
u l t i m a t e particles, Plato postulates a small number restricted by teleologi­
cal principles, postulates structured particles, and sees the need for the
teleological ordering of both the extrinsic and intrinsic organisation of the
u l t i m a t e particles.
M y account of Plato has some similarities, b u t also some i m p o r t a n t
differences w i t h the views of Descartes on the nature of the physical world.
B o t h Plato and Descartes treat m a t t e r as shapes w i t h specifiable geomet­
rical properties and n o t h i n g else. Descartes, however, identifies m a t t e r
and extension, such t h a t empty space is inconceivable. Plato's view is
72

rather of extended entities moving i n space allowing for some small-scale

234 Plato's Philosophy of Science
empty space, as we have seen. For Descartes the u l t i m a t e particles of
m a t t e r are three-dimensional, come i n an indefinite number of shapes and
sizes and are indefinitely divisible, a l l contrary to Plato's account. W h i l e
73

some hold t h a t Descartes geometrised space, we cannot say the same for
74

Plato. The receptacle is specifically shapeless (amorphon, 51b7).

Plato and Descartes appear to be i n strong agreement on the inherent
properties of matter. M a t t e r has geometrical properties and motion, and
no other properties. So for Descartes interaction between particles is by
contact action only, and t h i s serves to explain a l l attractive and repulsive
effects t h a t m i g h t otherwise be thought to operate at a distance, including
the h a r d cases of gravity, electricity and m a g n e t i s m . Plato is keen to
75

stress this as well, at Timaeus 80b8-c4:

And moreover the flowing of waters, the fall of thunderbolts and the wonder­
ful attraction effects of electricity and the magnet, all these are not due
76 77

to any power of attraction, but to the fact that there is no void and that the
particles push into each other.

He has Timaeus go on to say t h a t as particles disintegrate or combine they

w i l l move to t h e i r own regions, and by means of these complex processes
w h a t appeared marvellous ' w i l l become clear to someone investigating i n
the proper manner', a sentiment w i t h w h i c h Descartes would wholeheart­
edly agree.
So b o t h Descartes and Plato have a strongly reductive account of
physical phenomena. The only question here is whether Plato considers
the like-to-like principle to be an inherent property of bodies (as w i t h
N e w t o n i a n g r a v i t a t i o n operating at a distance), or an effect of certain
types of motion on bodies (as w i t h Cartesian gravitation w h i c h is an effect
on bodies generated by vortices). The latter seems more likely as we do not
hear of this principle i n the absence of motion, and electrical and magnetic
effects are explained along the same of lines. Plato's views have some
interesting similarities w i t h those of Leucippus and Democritus, and
indeed w i t h those of Descartes, b u t they have i m p o r t a n t differences as
well.
The comparison w i t h Descartes is worthwhile, not only to b r i n g out the
strongly reductive character of geometrical atomism b u t also to i l l u s t r a t e
the v i a b i l i t y of a strongly geometrical conception of the u l t i m a t e constitu­
ents of the world. We m i g h t ask whether the stoicheic triangles are planes
of m a t t e r or are more like coathangers w i t h m a t e r i a l boundaries enclosing
space. Coathangers w o u l d seem unlikely due to the possibility of stoicheia
or atoms passing t h r o u g h one another. Cornford objects to planes on the
grounds t h a t atoms w o u l d t h e n be hollow (PC 205), b u t as long as the
structure is r i g i d I do not see a problem here. A n alternative is t h a t these
are simply geometrical entities. One m i g h t agree w i t h Crombie that:
 8. Geometrical Atomism - Matter and Space 235
If we are tempted to ask: "Triangles of what?' we must resist the question.
They are simply triangles. 78

Crombie though has the triangles as triangles of space. So too Descartes'

particles are particles of extension/matter, there being no empty space. A 79

more radical view would have the triangles as geometrical entities, w h i c h

are extended i n space b u t not constituted of space. Thus they m i g h t move
i n space w i t h o u t changing space. One would t h e n have to t h i n k about the
analogues Plato uses for the interactions between atoms i n a more sophis­
ticated and mathematical manner (fire particles c u t t i n g because they have
sharp edges looks distinctly material), b u t t h a t is possible. The simplest
view, though, is t h a t the triangles are m a t t e r i n space.
Here, as elsewhere, Plato's proto-realism would seem to be an i m p o r t a n t
d r i v i n g force i n postulating real b u t unobservable micro-entities w h i c h
have specific inherent geometrical properties. T h a t i n t u r n forms the basis
for a realist account of the laws by w h i c h they interact. Such laws are not
empirical generalisations b u t are based on the inherent geometrical prop­
erties of, say, fire atoms, w h i c h of necessity have certain effects on other
atoms w h e n they meet t h e m i n certain ways. So too Plato seems to
postulate a relation between the entities of geometrical atomism and
language such t h a t we can get a good grip on these entities. He also seems
driven to postulate u l t i m a t e constituents of the n a t u r a l w o r l d w h i c h are
both explicable and explanatory.

EX. Teleology a n d b o n d i n g
Some further comments on the nature of geometrical atomism. I see no
reason to suppose there to be irregular figures, at the level of atoms,
complexes or the stoicheic triangles i n the cosmos. There is an issue
r e l a t i n g to the analysis of the sense of smell at Timaeus 66d ff. I f the
difficulty is that the olfactory veins are too small to admit individual atoms of
earth and water and too large to sense individual atoms of fire and air, and
none of the elements can be smelled, only the intermediate states where
water is t u r n i n g into air or air into water, then perhaps there are intermediate
atoms of some sort. However, Vlastos argues convincingly that the olfactory
80

veins cannot be so small, and that the problem is to do w i t h the viscosity of

the smoke or mist inhaled and not the sizes of individual particles. As we 81

can account for smell i n this manner, there is no need for irregular particles
which would not fit i n w i t h the scheme of beauty devised by the demiurge.
One question t h a t is frequently asked of any account of the receptacle
and its contents is w h a t is supposed to be i n the pre-cosmic chaos? A r e
there i r r e g u l a r figures here, especially as at Timaeus 53a the elements
only possess traces of t h e i r proper nature and at 69a n o t h i n g partakes of
proportion i n the pre-cosmos except by chance? Here I r e t u r n to some
points made i n Chapter 1. A r e we to take the ordering of the pre-cosmic
236 Plato's Philosophy of Science
chaos into a cosmos literally or metaphorically? There I argued t h a t i f we
take this literally, the point of the assorted passages on the pre-cosmos is
not to describe a specific state of chaos but rather to make the more general
point t h a t whatever less ordered state preceded the currently ordered
state, the t r a n s i t i o n between the two is not something t h a t could have
come about by necessity and chance alone. I f we take these passages as
metaphorical, or as I put i t , a piece of counter-factual analysis, t h e n the
emphasis is more epistemological on w h a t we need to assume i n terms of
the intrinsic and extrinsic order of the elements such t h a t we can have a
reasonable account of the current cosmos. I f Plato d i d believe i n a complete
chaos prior to the w o r k of the demiurge, then the stoicheic triangles m i g h t
lack intrinsic order, as indeed would everything else, though i t would be
h a r d to see w h y the elements would have traces of order. Even i f this were
so, I do not see w h y this would prejudice the stability of the stoicheia i n
the cosmos, although one would be able to draw a distinction between the
receptacle w h i c h seems to exist i n the same manner i n the pre-cosmos and
the stoicheia which do not.
W h y does Plato select the half-square and the half-equilateral as his
fairest triangles? Taylor cites the ratios of the angles, 1:1:2 (45:45:90) and
1:2:3 (30:60:90), while Cornford suggests t h a t such triangles allow for a
82

sequence of 'isotopes' which are relatively closely spaced i n terms of

triangles used, though as A r t m a n and Schafer argue i t is not clear w h y
83

t h a t should be an i m p o r t a n t criterion. Correctly, i n my view, they relate

84

this m a t t e r to t h a t of the fairest of bonds discussed at Timaeus 31c-32a,

and the question of doubling a square and t r i p l i n g a t r i a n g l e . The process
85

of doubling a square is something which is well k n o w n from the Meno 86

F r o m Timaeus 31c ff. the best bonds exhibit the ratio a:b = b:c. For the
square complex, where a is an a r m and b the hypotenuse of a half-square
stoicheic, and c the diagonal of the square, a:b = b:c, as 1:V2 = 2:V2, and c
= 2a (Fig. 57). For the complex triangle, where a is the short a r m and b the
long of a half-equilateral stoicheic, again a:b = b:c as 1:V3 = V3:3, and i n this
case c = 3a (Fig. 58). As A r t m a n and Schafer comment, we thus get the
88

fairest of bonds between 1, 2 and 3, and we also get 1, 2 and 3 i n the ratios

C=2A

Fig. 57. D o u b l i n g a
square. A:B = B:C,
87

w h e r e A is t h e side of t h e
s m a l l s q u a r e , B the
d i a g o n a l of the s m a l l
square a n d C the
d i a g o n a l of t h e l a r g e
s q u a r e , or
1:2 = 2:2, a n d C = 2A.
 8. Geometrical Atomism - Matter and Space 237
Fig. 58. Tripling a triangle.
A:B = B:C, where A is half the
base of the small triangle, B is
the height of the small triangles,
and C is the height of the large
triangle, or 1:3 = 3:3, and C = 3A.

of the angles as Taylor observes. This works very well for the constitution
of the standard atomic faces, though i t is less clear w h y on these grounds
the faces should j o i n to make the perfect solids. I f the like-to-like principle
of the Timaeus is quite general rather t h a n merely for types of atoms, and
Timaeus 81a2-3 would suggest i t is, then similar stoicheic triangles can be
brought together by this principle.
Aristotle does point out a major weakness i n geometrical atomism,
though he perhaps does not make the most of i t . A t De Caelo 299b25 ff. he
argues t h a t as lines can be placed next to each other, end to end, or on top
of one another, i t is too much to ask t h a t the surfaces only contact each
other side by side. Aristotle could considerably strengthen his case by
pointing out t h a t there are indefinitely many ways i n w h i c h surfaces
m i g h t be p u t together and we are being asked to believe this always
happens i n one way w i t h o u t any reason, and t h a t there are two layers of
this, the stoicheics j o i n i n g to make complex triangles and the complexes
j o i n i n g to make atoms.
One answer to this m i g h t be t h a t both do so i n order to instantiate the
best. Stoicheic triangles j o i n up i n the specified manner to produce the best
bonds, complexes j o i n up to produce the best three-dimensional figures.
One m i g h t accept t h a t i n the i n i t i a l ordering of the cosmos the demiurge
arranges everything j u s t so. The problem though is the theory of change.
Atoms decompose, but how do they reorganise into other atoms? I n Chap­
ter 1 1 suggested t h a t we m i g h t consider each atom to be a mini-cosmos. I
would stand by that, but there is a significant difference w i t h the cosmos
i n t h a t the cosmos is alive and shows the ongoing action of intelligence, as
do the heavenly bodies, and thus the heavens r e t a i n t h e i r order. There is
no suggestion t h a t stoicheics, complexes or atoms are intelligent entities.
A further difficulty here is t h a t of isotopes. I f there are different complexes
w h i c h m i g h t be formed, and different ways i n which complexes m i g h t j o i n
238 Plato's Philosophy of Science
together to create different sizes of atomic faces, t h e n how the stoicheics
and the complexes j o i n up seems underdetermined.

X . T h e s i g n i f i c a n c e of g e o m e t r i c a l a t o m i s m
I shall finish this chapter w i t h some more general comments on the
historical significance of Plato's geometrical atomism. I would agree w i t h
Vlastos t h a t one of the finest legacies left by the Greek atomist t r a d i t i o n
(and I include Plato and some of the physiologoi i n this) is the development
of two-level theories. The idea t h a t there is a micro-world beyond our
89

perception, w h i c h while i t underpins our sense perceptions can be radi­

cally different i n its nature from the macro-world, is an enormously
significant step forward i n the history of science. A l t h o u g h Vlastos puts up
some conjectural equations for physical processes involving analysis into
component parts, i t would be misleading to suggest t h a t Plato ever made
use of the possibilities for mathematisation opened up by geometrical
a t o m i s m . W h i l e Plato does make use of such possibilities i n astronomy
90

(or at least urges others to!), there is no evidence of any similar programme
for physics or chemistry, Vlastos makes the i m p o r t a n t point t h a t w h i l e i n
astronomy i m p o r t a n t 'hard facts' were being gathered to w h i c h theories
w o u l d have to adhere, no similar process was happening for t e r r e s t r i a l
disciplines, w h i c h h a d to rely on commonplace observations. Thus i t was 91

h a r d for the Greeks to choose between Plato and the atomists, and indeed
other m a t t e r theories. Indeed some issues here appear radically under-
determined, w h i c h is w h y we see Plato employing his teleological solution
to Meno's paradox and so hypothesising some ideal basic entities, w h i c h
i n some ways places h i m closer to modern atomism (all particles as perfect
spheres, atoms for the earlier atomists, sub-atomic particles for the later
ones).
I t is w o r t h considering some of the ways i n w h i c h Plato's geometrical
atomism is r a t h e r more l i k e modern atomism t h a n t h a t of Leucippus and
Democritus. Plato is strong on the idea t h a t there are a small number of
types of u l t i m a t e particles, w h i c h are m a t h e m a t i c a l l y w e l l defined,
against the indefinite number of shapes and sizes of the atomists. Plato's
discussion of geometrical atomism stresses t h a t m a t t e r has deep structure
i n the sense of u l t i m a t e particles forming structures w h i c h i n t u r n t h e m ­
selves form further structures, and so on. T h a t is an idea w h i c h is absent
from Leucippus and Democritus. A l t h o u g h Plato is not specific about how
bonding between particles occurs, and indeed this is the major theoretical
flaw i n the scheme, he is r i g h t t h a t this is not an accidental m a t t e r based
on mechanical interaction b u t happens i n a specific and well defined
manner. H i s theory of'isotopes' is also interesting, not as a precursor to
92

any modern ideas about isotopes b u t rather as another emphasis on the

intricacies of structure at the sub-atomic level. This of course is a some­
w h a t Whiggish approach to assessing Plato's geometrical atomism, b u t
 8. Geometrical Atomism - Matter and Space 239
given the praise t h a t the ancient atomists have received precisely along
those lines i t is w o r t h p o i n t i n g out w h a t we m i g h t be able to say about
Plato too. I f we are going to give the ancient atomists credit for i n this
manner, t h e n we must also give credit to Plato. Perhaps more importantly,
Plato's atomism, like his cosmology, is perfectly viable i n the context of
fourth-century Greece. I t is aimed at solving certain problems i n the
accounts of the atomists and physiologoi, and is no less plausible as an
account of the physical world.
As I argued i n Chapter 1, i t w o u l d be w r o n g to see the seventeenth
century as solely reviving the atomism of Leucippus and Democritus to the
exclusion of Plato's geometrical atomism. T h a t the seventeenth century
felt the need to cure atomism of atheism was not merely a religious
predilection. The philosophical problems w i t h Presocratic atomism, of w h y
atoms should have certain shapes and combine i n certain ways, w h i c h
Plato addressed w i t h teleology and the demiourgos, were now addresses
w i t h a C h r i s t i a n deity. So we find Boyle saying that:

The provident demiourgos wisely suited the fabric of the parts to the uses
that were to be made of them. 93

Similarly, we find N e w t o n saying that:

It seems probable to me that God in the beginning formed matter into solid,
massy, hard, impenetrable movable particles, of such sizes and figures and
with such other properties and in such proportion i n space, as most conduced
to the end for which he formed them. 94

So too there was a considerable debate between those who advocated a

p l e n u m of particles and those who favoured atoms and a void. I t is notable
t h a t those who favoured the plenum and rejected the idea of action at a
distance, such as Descartes, adopted a similar solution to Plato for the
a w k w a r d cases of gravity and magnetism. Timaeus 80c argues t h a t the
'attractive' powers of electricity and magnetism are not due to any action
at a distance, b u t can be explained by the fact t h a t there is no void and the
atoms jostle each other and move to t h e i r own region. Descartes, w h i l e he
used vortices to explain gravitational effects d i d so w i t h o u t a void, and
used screw-shaped particles moving among smaller particles to explain
magnetism. 95

A further i m p o r t a n t aspect of Plato's thought related to geometrical

atomism is the question of i r r a t i o n a l numbers and measurement. The
Pythagoreans treated geometry arithmetically, by a t t e m p t i n g to treat
geometrical problems as part of the theory of n a t u r a l numbers, t h a t is as
numbers composed of indivisible monads. Thus every geometrical length
96

ought to be expressible as the ratio of two n a t u r a l numbers. I f these

numbers represent a length, t h e n i f we ask how long something is, rather
t h a n measure the distance we count the number of monadic lengths
240 Plato's Philosophy of Science
involved. So too, according to Aristotle the Pythagoreans treated physical
entities as i n some way constituted out of number. The great problem for
97

these projects comes w i t h the discovery of the i r r a t i o n a l i t y of the square

root of two, for here we have a number/length t h a t cannot be expressed as
a ratio of two n a t u r a l numbers, or as a m u l t i p l e of a monadic l e n g t h . T h a t
98

Plato was aware of this for not only lengths b u t for areas and volumes too
is made clear by the Athenian's explicit comments at Laws 819d f f . 99

I t has been argued, most notably by Popper, t h a t i n response to the

100

difficulties of the Pythagorean programme Plato advocated geometrical

rather t h a n a r i t h m e t i c a l means for the description and explanation of the
w o r l d . The p a r t of the Timaeus i n this was of proposing a specifically
geometrical atomism. As i f to emphasise the overcoming of the difficulties
dogging the Pythagoreans, the stoicheic triangles have sides of root two
and root three. I t may also be the case t h a t Euclid's Elements were more
t h a n an exercise i n pure geometry, and were intended as an Organon of
the Platonic cosmos. Proclus for instance comments:
101

Some have thought that the subject matter of the various books [of Euclid]
pertains to the cosmos, and that they are intended to help us in our contem­
plation of, and theorizing about, the universe. 102

Popper comments t h a t one of Plato's m a i n contributions is t h a t

Ever since, but not before, geometry (rather than arithmetic) appears as the
fundamental instrument of all physical explanations and descriptions, in the
t h e o r y of m a t t e r a s well a s c o s m o l o g y .
103

Plato's geometrical atomism t h e n is of some importance. I t seeks to rectify

specific philosophical problems i n the ideas of the atomists and the physio­
logoi. I t provides i n the stoicheic triangles a stable basis both for the w o r l d
and for our references to the world. I t allows for diverse phenomena,
r a n g i n g from phenomenal flux of the elements to the t o t a l qualitative
stability of the heavens, and so underpins the r e g u l a r i t y and stability of
the motions of the heavens. I n the latter part of the Timaeus i t is employed
to provide outline explanations for a l l the major phenomena, including,
significantly, an account of how a l l our senses w o r k and how phenomena
can be analysed i n terms of the shape and motion of the atoms.
A final point on the seriousness of the Timaeus. O n m y view, the
Timaeus highlights problems w i t h atomism, w i t h the receptacle and w i t h
flux theories. I t advocates (or wishes its readers to consider seriously) the
merits of a teleological geometrical atomism. Clearly Plato objects to the
philosophy of the atomists and physiologoi. One point of the Timaeus may
be to show t h a t t h e i r positions on the physical w o r l d are inadequate too,
w h i l e views (newly) developed out of Plato's philosophy are adequate. I see
no reason on this view w h y Plato should not be considered a serious
cosmologist or the Timaeus be considered a serious w o r k of cosmology.
 9

Epistemology i n the Timaeus and

Philebus

I t would seem t h e n t h a t we can produce a coherent and viable set of

interpretations of i m p o r t a n t passages i n the Timaeus based on the ideas
t h a t the motions of the heavens are regular and stable and t h a t the basic
entities of geometrical atomism, the stoicheic triangles, do not undergo
any alloiosis. This m i g h t be thought to create some tension w i t h Timaeus
27c-29d. We m i g h t accept t h a t tension, either as unrecognised by Plato, or
as inherent i n any attempt to give a full account of the w o r l d of becoming,
w h i c h according to Timaeus we cannot give a fully consistent account of.
W h a t follows is a rather more radical way of a t t e m p t i n g to deal w i t h this
tension.
The Timaeus develops an i m p o r t a n t analogy between the w o r l d soul
and the h u m a n soul. This is done i n detail, and is critical to the conception
of humans and t h e i r place i n the cosmos i n the Timaeus. A model of m i n d
is developed, as is an account of the mechanics of false judgement. Yet the
epistemology t h a t is inherent i n this view does not sit w e l l w i t h 27c-29d,
and seems considerably more optimistic. So too there are other i n t i m a t i o n s
scattered t h r o u g h the Timaeus t h a t we can do rather better t h a n an i n i t i a l
reading of 27c-29d m i g h t lead us to believe, especially i n r e l a t i o n to
astronomy. The point of this is to show t h a t there are tensions i n the
Timaeus irrespective of the view we take on the heavens and the stoicheia.
So too there are some distinct oddities about the views we are given at
27c-29d, both inherent to those views and i n relation to views given i n
other works. Should we t h e n believe the views given at Timaeus 27c-29d
to be Plato's own w i t h o u t further consideration?
I argue t h a t there are good grounds for believing t h a t the Timaeus is
not a repository of Platonic doctrine. We have already seen t h a t the
astronomy of the Timaeus is best viewed as a prototype rather t h a n as a
completed and definitive view. So perhaps the Timaeus (at least i n part)
sets out to pose its readers some problems. One of t h e m may be this. W h a t
do we hold to be fundamental w h e n we investigate nature? Some version
of a two worlds (TW) theory? The stability of the heavens or the stoicheia?
Can these be held together w i t h o u t reformulation? I f we take the stability
of the heavens and the stoicheia as fundamental, t h e n we need to r e t h i n k
242 Plato's Philosophy of Science
T W somewhat, and there seem to be grounds for much greater epistemo­
logical o p t i m i s m t o w a r d the investigation of nature. T h a t is h i g h l y
significant for the sort of results Plato m i g h t expect from such an investi­
gation. One problem here of course is t h a t i f Plato is challenging his
readers to examine t h e i r own views, i t becomes a t r i c k y m a t t e r to discern
Plato's own views. I t h i n k there are some indications i n the Timaeus on
this m a t t e r though, and certainly i n other later works, notably the Phile­
bus, Laws and Epinomis, there appears to be greater o p t i m i s m t o w a r d the
idea t h a t the investigation of nature and the technai w i l l actually generate
some knowledge. Clearly this approach to the Timaeus is a conjectural one,
b u t i t does provide an interesting alternative to the view t h a t Plato
believes we simply cannot produce a consistent account of the physical
w o r l d and so should not be overly concerned about tensions and inconsis­
tencies i n the investigation of nature.

I . W o r l d soul a n d h u m a n souls
L e t us begin w i t h the question of w h a t the w o r l d soul can k n o w and w h a t
i t can form opinions about, and how i t acquires these states. H a v i n g
described the construction of the w o r l d soul, at 37a Timaeus goes on to say:

Therefore, as soul was blended out of its three constituent parts, sameness,
difference and ousias ... whenever she has contact with something that has
either dispersed or undivided ousian, she is set in motion throughout the
whole of herself and declares what this thing is the same as and what it is
different from, and how i t affects and is affected by each thing, in which
manner and how and when, both in relation to the things which come to be
and those which always are. When this gives rise to discourse which is
equally true whether concerning that which is different or that which is the
same, i t is carried within the self moved without speech or sound. Whenever
1

this occurs concerning the sensible (to aistheton) the circle of the different
running correctly announces this to the whole of the soul and opinions and
beliefs which are secure and true occur (doxai kai pisteis gignontai bebaioi
kai aletheis). On the other hand, where the rational (to logistikon) is 2

concerned and the smoothly running circle of the same declares this, intelli­
gence and knowledge of necessity result.

The epistemological implications of this passage have been underesti­

mated. I f the w o r l d soul can have opinions concerning the sensible w h i c h
are b o t h secure and true, t h e n there cannot be a n y t h i n g intractable i n the
n a t u r e of the sensible, or the nature of opinion, w h i c h prevents the
formation of such opinions. There may be difficulties for we mere mortals,
b u t t h a t is a different matter, and one t h a t locates the problems w i t h us
rather t h a n w i t h the sensible w o r l d or w i t h opinion. Timaeus goes on to
give us an account of the formation of the h u m a n soul and its capacities,
and a description of how i t is t h a t h u m a n beings are able to form false
judgements.
 Epistemology in the Timaeus and Philebus 243
A t 41c ff. h a v i n g created a race of demigods to be the souls of the
heavenly bodies, the demiurge t h e n gave t h e m the job of helping i n the
creation of mortals. He once more blends together sameness, difference
and ousia, though this t i m e not as u n i f o r m l y as before, and divides the
3

resulting m i x t u r e into the appropriate number of souls. Though not explic­

i t l y stated, i t w o u l d seem t h a t he also performs the same acts of d i v i d i n g
into strips and b i n d i n g into circles t h a t he does w i t h the w o r l d soul, such
t h a t our souls consist of the two revolutions. He then t u r n s these over to
the demigods, who proceed to fashion a body for each. H a v i n g been
i m p l a n t e d i n physical bodies subject to influx and efflux, m o r t a l souls are
t h e n subject to sensations, to desire mingled w i t h pleasure and pain, to
fear, anger and other emotions. The fundamental moral injunction is t h e n
4

to master rather t h a n be mastered by our emotions. The crucial difference

between our souls and t h a t of the w o r l d soul can be p u t like this. The
revolutions of the same and the different t h a t make up the w o r l d soul are
regular, stable and imperturbable. This makes the w o r l d soul perfectly
5

rational, and a l l of its judgements of sameness and difference are correct.

However, w h e n our souls are bound into our bodies, the revolutions are
disrupted, they partake i n a l l six motions, and the influx of sensations
6

further disorders t h e i r motions, such t h a t i n i t i a l l y the t w o revolutions

barely hold together, as at Timaeus 43d:

They produced all manner of twisted circlings, and all manner of fractures
and corruptions of the circles in as many places as was possible, with the
result that they barely held together, moving irrationally (alogos), being at
one time reversed, at another sideways, and at some times upside down.

The w o r l d soul immediately begins intelligent life, as its revolutions are

i n i t i a l l y perfect, u n l i k e those of newly born humans. A t Timaeus 43e we
7

are t o l d t h a t this is because of the imperfection of our mental orbits:

This is like when someone sets themselves upside down, with their head
against the ground, holding their feet aloft by pushing them against some­
thing. I n this case right and left both of the sufferer and of any onlookers will
appear to be reversed to the other party. Whenever they encounter some
external object, whether it be of the class of the same or the different, then
they label i t the same as or different from something contrary to the truth,
showing themselves to be incorrect and foolish.

O n this basis, we make some incorrect attributions of sameness and

difference. There are two i m p o r t a n t parallels w i t h other late works here.
A t Sophist 263d we are told:

When statements about you state as different what is the same and what is
not as what is, i t is very likely that the occurrence of such a combination of
verbs and nouns really and truly constitutes false statement.
244 Plato's Philosophy of Science
I t is interesting t h a t the Theaetetus alludes to the same sort of example as
the images i n its discussion of false judgement, though w i t h o u t d r a w i n g
the same conclusion. I n the discussion of the wax block i n the Theaetetus,
at 193c Socrates tells us that:

The possibility of false judgement remains in the following manner ... i t may
be that I am in error suffering as sight is affected by a mirror, left changing
into right. I n such a case other judging and false judging happen. 8

A t Timaeus 46a f f , we are given an analysis i n terms of fire particles for

why m i r r o r s produce this reversal of left and r i g h t . W h a t is remarkable
about this model of m i n d and false judgement is how optimistic Timaeus
is about the possibility of h u m a n beings bettering themselves. A t 44a ff.
Timaeus tells us t h a t at first we are hardly judge a n y t h i n g correctly, our
m e n t a l revolutions h a v i n g been disrupted at b i r t h , though w i t h t i m e and
effort as the orbits settle:

Those revolutions which are assailed by these external sensations are drawn
along by them together with whole vessel of the soul, and at such times they
appear to have the mastery but are i n fact mastered. On account of these
sufferings, now as in the beginning the soul is at first without intelligence,
as when i t was bound into a mortal body. But when the stream of growth and
nutrients flows less strongly, then the revolutions become calmer, pursuing
their own path and becoming more stable as time progresses, and eventually,
as each of the circles comes to travel its own pattern, the revolutions correct
themselves, and announce both the same and the different correctly, so
rendering their possessor intelligent. I f he now takes part in some correct
nurture of education, he becomes perfect and entirely sound in mind, having
escaped the greatest of maladies. 9

The Timaeus suggests both t h a t there is a n a t u r a l tendency for the orbits

to settle down (44b) and t h a t i t is our job to help t h e m do so as far as is
possible (44c), by mastering rather t h a n being mastered by our percep­
tions. Note how this relates to the moral injunction of Timaeus 42b. I f we
can master our emotions we can become j u s t , i f we can master our
perceptions, we can become rational.
So too Timaeus 90a-d is i m p o r t a n t here. I n order to become like god, 10

we need to rectify the revolutions i n our own heads by studying those of

the universe. This a l l looks very promising. I f we can t r a i n ourselves
properly, t h e n i t would seem t h a t we too can have some opinions concern­
ing the sensible w h i c h w i l l be l i k e those of the w o r l d soul. A l l the
difficulties here for humans are tractable ones, and given w h a t the w o r l d
soul is capable of, there are no intractable problems w i t h sensible entities
or the nature of opinion. We shall come back to the i m p o r t a n t role astro­
nomy has to play i n this i n a l i t t l e while. For the moment though I w i l l
merely note t h a t this a l l looks distinctly more optimistic t h a n Timaeus
27c-29d.
 Epistemology in the Timaeus and Philebus 245

I I . Souls a n d common concepts

One m i g h t gather from the account of the construction of the w o r l d soul,
i n the way t h a t the demiurge mixes together portions of sameness, differ­
ence and being (pusia) t h a t each of these are being treated as complete
predicates. T h a t is, i t would appear t h a t i t is legitimate to say of something
t h a t ' i t is same' perhaps i n v i r t u e of the sameness i t possesses, rather t h a n
saying ' i t is the same as x \ However, other passages seem to treat same­
ness and difference as incomplete predicates. A t Timaeus 37a9, w h e n the
w o r l d soul interacts w i t h anything, i t announces i n t e r n a l l y 'what i t is the
same as and different from, i n w h a t respect'. So too at Timaeus 44a2 where
humans judge things to be tauton toi kai thateron tou, 'the same as and
different from'. W h y should there be this apparent disparity? I t m i g h t be
t h a t Plato has not yet properly distinguished properties and relations, and
only apparently treats sameness and difference as relations. A n o t h e r 11

suggestion is this. As a l l of the complete uses are i n the passages concern­

i n g the construction of the w o r l d soul and h u m a n souls, perhaps t h i s
12

is a metaphorical means of a t t r i b u t i n g something to those souls. As 13

Timaeus goes on to t e l l us t h a t whenever they encounter anything, both

the w o r l d soul and humans make judgements of w h a t t h a t t h i n g is the
same as and w h a t i t is different from, perhaps w h a t is being a t t r i b u t e d to
the soul is an innate ability to make judgements of sameness and differ­
ence. W h a t though of the use of einai here? I f its treatment is s i m i l a r to
14

t h a t of sameness and difference, t h e n perhaps Plato reduces w h a t he

considers to be a two-place operator (the 'is' of i d e n t i t y and predication) to
a one-place operator (which t h e n looks like b u t is not the 'is' of existence)
i n order to a t t r i b u t e another innate ability, perhaps t h a t of forming
statements of i d e n t i t y and predication. This is even more fundamental, for
i n order to be able to form judgements of relations of sameness and
difference we require the ability to form such propositions. I f this is so, 15

i t w o u l d explain why, though souls are constituted from sameness differ­

ence and ousia t h a t Timaeus only speaks of m a k i n g judgements of same­
ness and difference, as i n doing so we are also employing the ability
a t t r i b u t e d by einai.
Plato may have had several reasons for introducing such innate abili­
ties. If, by this t i m e he has rejected anamnesis w i t h its recollected
knowledge, he may feel the need to postulate some fundamental innate
abilities i n order to be able to explain learning ab i n i t i o . The Timaeus
model avoids some of the notorious difficulties concerning w h a t the new­
born child can know or perceive according to the account of recollection at
Phaedo 73b-75c. According to the Timaeus, we are able to make judge­
ments about our sense perceptions straight away, although these w i l l be
rather poor due to the confused i n i t i a l state of our m e n t a l revolutions. One
advantage of the Timaeus account is t h a t i t allows for intellectual devel­
opment, i n the sense of an increase i n our cognitive powers associated w i t h
246 Plato's Philosophy of Science
the steadying of our m e n t a l revolutions as well as the mere gain of
knowledge allowed by recollection. Plato is perhaps not giving us an
exhaustive account of functioning of the m i n d , especially as i n later works
he allows t h a t we may t h i n k pictorially. Rather, he is allowing us to t h i n k
16

propositionally about a n y t h i n g t h a t is brought before the m i n d . One m i g h t

also note i n relation to m y discussion of the deuteros pious i n Chapter 3
t h a t Plato develops a very active and independent account of the m i n d .
Perhaps the clearest statement of this is at Timaeus 77bc, where we find
t h a t the t h i r d k i n d of soul seated i n the midriff, i n distinction to the
revolutions i n our heads:

Inasmuch as i t continues to be entirely passive, and does not revolve of itself

or repel external motion it has no capacity for reasoning about its own
experiences.

As we have seen, the two key injunctions i n the Timaeus are t h a t our minds
should have mastery over our perceptions so that we can become wise, and
over our emotions so that we can become good. We might also consider this
theory of innate abilities i n relation to the 'common ideas' of the Theaetetus. A t
185c Socrates and Theaetetus have the following interchange: 17

Socrates: Through what does the ability function which reveals to you
what is common (koinon) to all things, that which you name is or is not, and
others which we were asking about just now? What organs will you assign
for all of these, through what does the perceptive part of us perceive each of
these?
Theaetetus: You mean being and not being, likeness and unlikeness,
sameness and difference, and one and the others numbers as applied to them
... I don't know what to say, Socrates, except that i t seems that there isn't a
special organ for these as there is for others, but the soul is its own instru­
ment examining what is common in all things.

The theory of recollection would undoubtedly have a tale to t e l l here about

the acquisition of these common ideas. One of the interesting aspects of the
Timaeus is t h a t Plato presents us w i t h a fairly parsimonious theory of the
innate abilities of the m i n d . The Theaetetus mentions several common
ideas, such as u n i t y and p l u r a l i t y , likeness and unlikeness, beautiful
18 19

and ugly, good and b a d , as w e l l as sameness and difference. Now, a more

20

generous i n n a t i s t m i g h t make the ability to apply a l l of these innate

capabilities. The Timaeus though grants only the last two, w h i c h i n
conjunction w i t h the ability to t h i n k propositionally, m i g h t be thought to
be foundational. The Timaeus suggests t h a t we form some of our abilities
to make judgements from abstraction from w h a t is given i n perceptual
experience. For instance, at Timaeus 47a we are told that:

Sight of day and night, of months and the cycle of the years has procured for
us number and the concept of time, and has also led us to seek the nature of
 Epistemology in the Timaeus and Philebus 247
the universe. From these we have derived philosophy, the greatest gift of the
Gods to mortals. 21

I t is presumably by applying our judgements of sameness and difference

to our observations of the heavens t h a t we acquire the concepts of number
and t i m e t h a t we can subsequently apply to other perceptions. S i m i l a r l y
perhaps w i t h our knowledge of sameness and difference. I t is presumably
by applying our innate abilities to form judgements of sameness and
difference t h a t we come to some conscious knowledge of sameness and
difference.
A further i m p o r t a n t feature of the account of m i n d i n the Timaeus is the
way t h a t the structure and abilities of the m i n d relate to the structure of
the world. The w o r l d is made up of things t h a t are the same as and
different from other things. The w o r l d of the Timaeus is also structured i n
t i m e and space, and is organised by the demiurge according to number
(kaf arithmos). Our minds have a fallible ability to judge sameness and
difference, and are able to acquire the concepts of time, space and number.
So too the structure of the w o r l d can be though of i n terms of letters and
syllables as the arrangement of the stoicheic triangles giving an interest­
i n g isomorphism between language and the world. A l l this m i g h t indicate
some greater o p t i m i s m to the cognition of the w o r l d t h a n is suggested at
27c-29d. I t w o u l d seem t h a t Aristotle took the m a i n t h r u s t of Plato's
account of the soul to be along these lines too. I n his critique of Plato's view
of the soul i n De Anima, he states that:

In the same manner Plato in the Timaeus constructs the soul out of the
elements (ek ton stoicheion). For he believes that like is known by like, and
that things are formed from these elements (ek ton archon). 22

So too A r i s t o t l e at De Anima 406b26:

In the same way the Timaeus gives a physical account of how the soul moves
the body. The self-motion of the soul is responsible for the motion of the body
on account of its intimate relationship with it. For it had been created out of
the elements, and was divided according to harmonious ratios, in order that
it might have some innate (sumphuton) perception of this harmony and the
universe might move with harmonious motions, and the straight was bent
into a circle, and having divided one into two, meeting at two points, he
further divided one of these into seven circles, thus making the motions of
the heavens identical with the motions of the soul.

I I I . S o u l s a n d epistemology
Let us now compare Timaeus 37a ff. w i t h 29b ff. A t 29b6 ff. the i m p l i c a t i o n
was of a sungenes matching of secure (bebaios, 29b6), accounts w i t h secure
entities, and of l i k e l y ' accounts w i t h likenesses. A t 37a ff. though, there
exist beliefs and opinions (doxai kai pisteis) w h i c h are both secure and true
248 Plato's Philosophy of Science
(bebaios kai alethes) i n contradiction of both the letter and s p i r i t of
27c-29d. A t 37a ff. there is no sungenes stricture on explanation. Instead
23

of the idea t h a t any account of a likeness can be only a l i k e l y account

(29cl), we have an analysis of the psychological mechanics of error i n
terms of the revolutions of the h u m a n m i n d , and of false judgement i n
terms of sameness and difference. As the w o r l d soul can have secure and
t r u e opinions of the sensible (to aistheton) there cannot be a n y t h i n g
intractable w i t h the nature of the sensible w h i c h prevents the formation
of such opinions. So there is a contrast of intractable problems w i t h the
w o r l d of becoming to tractable problems w i t h h u m a n beings concerning
the sensible between 27c-29d and 37a ff. Given the analogy developed
between the w o r l d soul and h u m a n souls, and the analysis of error i n
terms of sameness and difference, i t is possible for us, i f we t r a i n ourselves
properly and have mastery over our perceptions, to have t h i s sort of
opinion concerning the sensible. T h a t too is contrary to the s p i r i t of the
closing p a r t of Timaeus' opening speech which tells us t h a t we who are
h u m a n i n nature should be ready to accept a likely account and seek no
further. H o w m u c h tension there is between 27c-29d and 37a ff. of course
depends on how strongly we interpret these passages, b u t there is no doubt
t h a t such tension exists. I t is i m p o r t a n t to stress here t h a t the passage at
Timaeus 37a ff. is not an isolated example which we can explain away as
a slip by the author. The h u m a n psychology developed i n analogy w i t h the
w o r l d soul is carefully worked out and developed over several pages and is
one of the central themes of the Timaeus. I t is i m p o r t a n t to note t h a t
Timaeus makes this a significant m o t i f w h e n he is concluding his speech
at 89d ff. A t 90d we should supply w h a t is f i t t i n g for the divine p a r t of us,
and w h a t is sungenes w i t h t h a t are the revolutions of the heavens. T h e n
at Timaeus 90d we are told t h a t we ought to be:

Correcting those revolutions in us that were disturbed at birth, by learning

thoroughly the revolutions and harmonies of the universe, by making our
intelligent part like to that intelligence according to its original nature.

Timaeus 37a ff. is also interesting i n relation to TW. Timaeus 28a contrasts
the intelligible w i t h sensible, b u t according to Timaeus 37bc, the domains
of opinion and knowledge for the w o r l d soul are to aistheton, 'the sensible'
and to logistikon, 'the accountable/calculable' respectively. W h i l e one24

m i g h t t r y arguing t h a t the domains of the sensible and the intelligible are

quite separate, there is no p r i m a facie reason w h y some things w h i c h are
sensible should not be amenable to calculations. I n order to be amenable
to calculations, these sensible entities m i g h t have to be stable, or t h e i r
motions m i g h t need to be describable i n a manner w h i c h is i n itself stable.
I f there were such things t h a t would go against the idea t h a t the w o r l d can
be simply divided into sensible things which are always changing and
intelligible things w h i c h are never changing. Prime candidates here of
 Epistemology in the Timaeus and Philebus 249
course are the heavenly bodies, w h i c h i f t h e i r fundamental particles are
qualitatively stable and are bound w i t h indissoluble bonds, and t h e i r
motions are entirely regular, are a set of stable and mathematically precise
entities. Timaeus has some very interesting things to say about how we
m i g h t use our observations here. Timaeus 47bc tells us:

God devised and gave to us vision in order that we might observe the rational
revolutions of the heavens and use them against the revolutions of thought
that are in us, which are like them, though those are clear and ours confused,
and by learning thoroughly and partaking in calculations correct according
to nature (logismon kata phusin orthotetos), by imitation of the entirely
unwandering revolutions of God we might stabilise the wandering revolu­
tions in ourselves (tas en hemin peplanememas). 25

The use of the p l u r a l throughout this passage strongly suggests t h a t both

types of our m e n t a l revolutions (same and different) are under discussion
here. I f so, and we can make calculation correct according to nature t h e n
the i m p l i c a t i o n would seem to be t h a t we can have some knowledge of the
heavenly bodies. T h a t w o u l d again be c o n t r a r y to the views p u t f o r w a r d
at 27c-29d. The importance l a i d on astronomy here as the subject w h i c h
w i l l b r i n g our m i n d s closer to the ideal of the w o r l d soul m a y go a l o n g
w a y to e x p l a i n i n g the c e n t r a l and otherwise somewhat disproportionate
role t h a t astronomy is given i n education i n the l a t e r w o r k s . One m i g h t
compare t h i s passage w i t h Timaeus 90cd, Laws 966e ff. a n d Epinomis
978b ff.
We need to exercise caution w i t h a l l of this m a t e r i a l . The language of
37a ff. is h i g h l y figurative, and the analogy between the w o r l d soul and the
h u m a n soul, w h i l e h i g h l y interesting, to some degree remains at the level
of m e t a p h o r a n d analogy. A great n u m b e r of questions r e m a i n concern­
i n g j u s t w h a t the w o r l d soul can k n o w and w h a t sort of s t a b i l i t y i t s
opinions m i g h t have. Plato does not spell out the f u l l consequences of
t h i s analogy for epistemology, nor does he make any direct comparison
w i t h Timaeus 27c-29d. We m u s t be careful t h e n not to press t h i s
m a t e r i a l too h a r d .
W h a t we can say, even i f we interpret 37a ff. i n a fairly m i n i m a l manner,
is t h a t there are significant tensions between the epistemology suggested
by the analogy of the w o r l d soul and the h u m a n soul and t h a t w h i c h
Timaeus outlines at 27c-29d, both i n detail and i n general tenor, t h a t of
37a ff. seeming to be considerably more optimistic. Timaeus 37a ff. also
provides a supplementary argument for the regularity and stability of the
heavens. I f the w o r l d soul is entirely rational, t h e n one w o u l d expect its
motions to be entirely regular. As its motions are those of the heavens,
26

the heavens must be regular and stable.

250 Plato's Philosophy of Science

IV. T i m a e u s 27c-29d a n d two w o r l d s (TW) t h e o r y

There are certain oddities about Timaeus' speech at 27c-29d, and there are
difficulties i n m a k i n g the speech sit w e l l w i t h several other passages and
comments i n the Timaeus. H a v i n g invoked the deities, at 27d Timaeus
begins by saying:

It is possible, in my opinion, firstly to make the following distinction. What

is i t that always is and never becomes, and what is it that always becomes
and never is? The one is apprehended by the intellect with reasoning, and is
always the same, the other in turn comes to be believed by opinion with
unreasoning sensation, always coming to be and passing away, but never
really existing.

This can be construed as a very strong version of TW. As Frede has argued,
coming to be here cannot mean coming into being i f w h a t becomes is
always becoming and never is. N o r is i t easy to take 'always' as meaning
'always changing i n some respect, b u t perhaps not air as t h a t w o u l d mean
t h a t w h a t becomes i n some respect could also be said to be i n some respect,
and t h a t appears contrary to the sharp bifurcation of w h a t always is and
w h a t never i s . This passage has an interesting s i m i l a r i t y to Theaetetus
27

152de, where Socrates says:

I t becomes all the things which we, incorrectly, say it to be. For i t never is
anything, but always becomes. 28

The Theaetetus passage occurs i n the description of a radical Heraclitean

flux. One issue we w i l l have to address soon is w h a t Plato's attitude to such
a flux is i n the Theaetetus and how 27c-29d relates to geometrical atomism
and the stability of the stoicheic triangles. Here I merely w a n t to note t h a t
Timaeus 27d is (or at least appears at first sight to be) an i n o r d i n a t e l y
strong statement of TW.
Clearly, i f we take Timaeus 27d at face value there are problems i n
a r g u i n g either t h a t the heavens are stable or t h a t the stoicheic triangles
are qualitatively stable. Let us leave t h a t on one side for a moment, and
look at whether this passage sits w e l l w i t h other aspects of the Timaeus.
F i r s t , there is going to be a problem w i t h status of the entities supposed
by geometrical atomism. Whether the stoicheic triangles are stable or not,
complex triangles and atoms undergo generation and destruction. None of
these are sensible as a l l are w e l l below h u m a n perceptual thresholds. Yet
the category of intelligible and undergoing change is not catered for at 27d.
So how can there be any account of these at all? These problems are
exacerbated by w h a t Timaeus says at 56b, t h a t at least part of w h a t he has
to say about geometrical atomism is kata ton orthon logon kai kata ton
eikota, i n accordance w i t h the correct account as w e l l as the likely. That,
although l i t t l e attention is p a i d to i t i n the literature, is a quite remarkable
 Epistemology in the Timaeus and Philebus 251
statement. I f we believe 27c-29d a l i k e l y account is the best we can hope
for, b u t here is Timaeus i n possession of the correct account, and saying
t h a t i n this case the likely account is i n agreement w i t h i t . One t h e n wants
to k n o w w h a t this correct account is, how Timaeus has access to i t and
w h a t the implications are for the epistemology w h i c h Timaeus espouses at
27c-29d. Even i f we take the subjunctive esto w h i c h introduces this phrase
to mean something like 'let this be i n accordance w i t h such t h a t
Timaeus is less certain t h a t this is the correct account, s t i l l he asserts t h a t
there is a correct account to be had. I would reiterate t h a t i n relation to
the heavens (whether they are stable or not) at Timaeus Ale we can make
logismon kata phusin orthotetos, calculations correct according to nature,
i n order to help stabilise our mental revolutions, and at 37b w h e n the
w o r l d soul encounters a n y t h i n g sensible doxai kai pisteis gignontai bebaioi
kai alethes, beliefs and opinions w h i c h are stable and true occur. There are
t h e n oddities both i n the nature of Timaeus 27d and i n its relation to other
parts of the Timaeus. A t 29b Timaeus states:

In every matter i t is of great importance to start at the natural beginning.

Thus concerning both the likeness and the paradigm, one must make the
following distinction, that accounts of them will be of the same family
(sungeneis) as the things they expound. Therefore accounts of that which is
steadfast, firm, and is clearly apprehended with the mind are steadfast and
unchanging, and so far as it is possible for them to be irrefutable and
undefeatable as befits such accounts, they must not fall short of this. On the
other hand, that which has been copied from the model, being a likeness, will
have accounts that possess likelihood, standing in an analogous relation to
the previous accounts. Whatever is related as becoming is to being, is also
related as opinion is to truth.

Here the sungenes ('of the same k i n d or family'), stricture seems somewhat
odd. One of the m a i n points of the theory of forms, one m i g h t t h i n k , is to
postulate some entities w h i c h are not sungenes w i t h w h a t they are sup­
posed to help explain. This is especially so i f one of the major motivations
behind the theory of forms is to postulate cognitively stable entities i n
order to resolve the compresence of opposites. There is a sense here i n
w h i c h the intelligible must be related to the physical, as the explanans
m u s t f i t the explanandum, b u t the forms could h a r d l y be thought to be
sungenes w i t h w h a t they are supposed to be explaining. Indeed, some of
the arguments of the first part of the Parmenides m i g h t be thought to play
precisely on that. We also have the word play of eikones and eikotas,
'likeness' and 'likely' at 29c2. Presumably t h a t is a consequence of the
sungenes stricture, b u t i t is a rather strange one. There is no suggestion
i n the Phaedo t h a t any account of w h y one t h i n g is larger t h a n another
w i l l be only a 'likely account' because i t deals w i t h the physical world. The
cases of two m e n and two horses are treated equally w i t h two numbers,
t e n and eight. W h i l e there are obvious echoes of the Republic here, the
29
252 Plato's Philosophy of Science
association of likeness and likely account is a new departure. Geometrical
atomism m i g h t be thought to breach this stricture, again whether the
stoicheic triangles are qualitatively stable or not. The entities postulated
by geometrical atomism are the archai of sensible things, w i t h o u t being
sensible themselves, and of course i f the stoicheic triangles are stable the
archai of changing things are stable. One of Aristotle's complaints, as we
saw, was precisely this, t h a t archai should be homogenes w i t h w h a t they
are archai of, b u t i n geometrical atomism they are n o t . 30

One m i g h t also ask who, prior to Plato, believed t h a t explanans and

explananda are sungenes, or more accurately, who Plato believed to hold
such views. Two p r i m e candidates here are Parmenides and Heraclitus.
There are indeed certain affinities between Parmenides' poem and Timaeus'
speech. B o t h separate the objects of reason and sensation, reckoning these
to be co-ordinate w i t h w h a t is knowable and w h a t is opinable, and both
require explanations to be sungenes w i t h w h a t they e x p l a i n . We m i g h t
31

also compare Timaeus' repeated use of eikos to describe the status of any
account of the physical w i t h Parmenides' s i m i l a r usage at Fr. 8/60. B o t h 32

Parmenides and Timaeus draw a distinction between the veracity of w h a t

they have j u s t said and the nature of the account they are about to g i v e , 33

and b o t h a i m to provide the best account of the physical w o r l d t h a t i t is

possible for m o r t a l s . A second figure who Plato believes t h i n k s t h a t
34

accounts are sungenes w i t h t h e i r explananda is H e r a c l i t u s . Whether this

35

is an accurate picture of the historical Heraclitus is another matter, b u t

the Theaetetus certainly takes this line about Heraclitean flux. So per­ 36

haps Timaeus' speech at 27c-29e sets up a Parmenidean account of reason

and the intelligible realm and a Heraclitean account of sensation and the
physical realm.
T h a t ought to make us pause though. I f the Timaeus comes after the
Parmenides (and i n particular after the final argument of the first p a r t of
the Parmenides, 133b-134e), w h a t is a Parmenidean account of the i n t e l ­
ligible doing here, especially w i t h the sungenes stricture i n play? W i t h
Heracliteanism, again there is a question of the relation to the Theaetetus.
Is a radical flux Plato's position there, and is i t Plato's position i n the
Timaeus? I f so, how does i t relate to geometrical atomism and the passages
cited above? So once more, Timaeus 29b is somewhat strange i n itself and
i n its relation to the rest of the Timaeus. A t the close of Timaeus' introduc­
tory speech at 29c, he says:

I f then, Socrates, on a great many matters concerning the Gods and the
generation of the universe we should not be able to give what is due
(apodounai) in terms of accounts which are in every way entirely in agree­
ment with themselves and perfectly exact, do not be surprised. But i f our
rendering is inferior to none in likelihood, this must please, remembering
that both I who speak and you the judges are human, such that concerning
these things we should gladly accept the likely story, as befits this, and not
seek any further.
 Epistemology in the Timaeus and Philebus 253
Socrates immediately acquiesces to Timaeus. Resonances w i t h Plato's
earlier works should also make us highly suspicious here. Should we
'gladly accept a l i k e l y account', or would Plato's Socrates of Meno 81d and
86bc castigate this as a l a z y ' approach, likely to stop m e n searching?
Would the Socrates of Gorgias 482b so glibly accept any account t h a t does
not agree w i t h itself? Wouldn't the Socrates of Phaedo 90b ff. reject
Timaeus as an antilogikos, one who does not put sufficient t r u s t i n the
power of argument? W h a t of the Socrates who has emphasised consistency
so strongly i n the Theaetetus? 37
A r e we meant to acquiesce so tamely w i t h
Timaeus on these matters?

V. T w o w o r l d s a n d s t a b i l i t y
So far i n this chapter I hope to have shown two things. F i r s t , the Timaeus
has an account of epistemology based on the analogy of the w o r l d soul and
the h u m a n soul w h i c h is worked out i n some detail, w i t h an account of
h u m a n intellectual development and of the mechanics of false judgement.
This epistemology not only seems rather more optimistic t h a n t h a t of
27c-29d, i n places i t also appears b l a t a n t l y to contradict i t . Secondly,
whether or not one considers the heavens and the stoicheia to be stable i n
the Timaeus, there are certain oddities i n t e r n a l to 27c-29d, and i n the way
i t relates to some other passages i n the Timaeus. There is I believe an
i m p o r t a n t conclusion for the interpretation of the Timaeus w h i c h comes
out of this. There are several key passages i n the Timaeus which, i f t a k e n
i n isolation, are indeterminate between different readings. A classic case
of this is how we construe Timaeus 49d ff., and w h i c h of fire, this and
suchlike are subjects and w h i c h are predicates. N a t u r a l l y we t h e n look
beyond the immediate context to see w h i c h i n t e r p r e t a t i o n w i l l harmonise
best w i t h the rest of w h a t the Timaeus has to say, and perhaps w i t h w h a t
other works have to say too. O n the grounds given above, though, we
cannot use the strong version of T W given at 27c-29d as an infallible guide
to the i n t e r p r e t a t i o n of such passages. P u t more colloquially, we cannot
use i t as a stick to beat the rest of the Timaeus into submission. O f course
this cuts b o t h ways. N o r can we use the stability of the heavens and the
stoicheia as infallible guides. Throughout I have chosen the options w h i c h
I believe harmonise best w i t h the stability of the heavens and stoicheia.
T h a t has been w o r t h doing to show the v i a b i l i t y of such an approach i n
dealing w i t h a wide range of passages, and to show how the interpreta­
tions this strategy generates can be l i n k e d together into a coherent pic­
ture. This remains a strategy for dealing w i t h indeterminate passages
though, and does not demonstrate how we m u s t interpret any of them, and
of course u l t i m a t e l y this reading comes up against the apparent b a r r i e r of
Timaeus 27c-29d.
WThat t h e n are we to do w i t h the Timaeus? Perhaps i t is incompetently
w r i t t e n and Plato was unaware of the tensions i n the account he has
254 Plato's Philosophy of Science
Timaeus give. Perhaps he is aware of at least some of the tensions, b u t
accepts t h a t we cannot have an entirely consistent account of these mat­
ters. T h a t though would mean t h a t he accepts 27c-29d, at least to the
extent of believing we cannot have a fully consistent account of the w o r l d
of becoming, i n w h i c h case one m i g h t expect several of the more indeter­
m i n a t e passages to be more clearly delineated i n t h e i r alignment w i t h
27c-29d, and there to be less evident contrasts w i t h the account of the
heavens and the epistemology of 37a ff. A more radical suggestion is t h a t
the choices being offered for the interpretation of the indeterminate pas­
sages are systematic, and t h a t we are presented w i t h the following
dilemma:

1. I f we accept the strong version of T W given at 27c-29d, and we apply

the principles involved there to other issues, t h e n there are problems w i t h
postulating stable heavens and stoicheia, and u l t i m a t e l y w i t h referring to
changing phenomena.
2. I f the heavens and stoicheia are stable, and we apply the principles
involved here to other issues, t h e n there are problems w i t h sticking to a
strong version of TW.

One of the difficulties w i t h (1), as I have sought to show i n the previous

two chapters, is t h a t i t m i g h t well leave us w i t h o u t the resources to deal
w i t h the problems t h r o w n up by the receptacle. Perhaps Plato wishes to
elicit some sort of a response from his readers. The idea of the tension
outlined above may be to provoke the readers into w o r k i n g t h e i r way
t h r o u g h the indeterminate passages and coming to t h e i r own conclusion
on the matter. They m u s t decide for themselves w h a t is fundamental i n
38

an account of the w o r l d about us and w h i c h consequences they can accept.

W h a t I shall do i n the next few sections is suggest some ways i n w h i c h we
m i g h t seek to undermine, at least to some extent, the idea t h a t the
Timaeus simply gives us Platonic doctrine. Here I shall attempt to counter
a common view, succinctly expressed by Sayre, that:

The major dialogues of the later period (with the exception perhaps of the
Timaeus and the Laws) were constructed to serve as dialectical instruments
rather than as repositories of doctrine. 39

This approach has certain advantages. I f we can read the Timaeus i n this
manner, i t narrows the gap i n style between the Timaeus and other
ipost-Republic and Phaedrus works. I n particular i t throws up some inter­
esting contrasts and parallels w i t h the Theaetetus, p a r t i c u l a r l y on the
question of whether Plato is committed to the idea of some form of
Heraclitean flux of the physical world. One difficulty w i t h this approach is
t h a t i t t h e n becomes a rather slippery m a t t e r to determine Plato's o w n
position. Perhaps t h a t is so anyway once we take the m a t t e r of Platonic
 Epistemology in the Timaeus and Philebus 255
anonymity into account. I n some ways t h a t is a difficulty for some of w h a t
I have argued i n this book, as I have sought to give a developmental
account of Plato's views on science. One can argue t h a t even i f Plato wishes
to provoke his readers to t h i n k for themselves, he s t i l l has some ideas of
his o w n , whether positive or negative, to get across. The account of
falsehood i n the Sophist a n d some refutations of positions, p a r t i c u l a r l y of
Protagoras i n the Theaetetus w o u l d be the prime candidates here. The
problem though is i n generating any h a r d and fast rules for distinguishing
between a position given to a character for some further purpose, and
Plato's own views.
I stand by m y developmental account of Plato's astronomy and cosmo­
logy, though i f we are to take this approach to Plato i t has to be p u t i n
somewhat more conjectural terms t h a n i f we were dealing w i t h someone
who simply wrote doctrine. Certainly I would argue t h a t the view I have
pursued here is quite a charitable one of Plato. I t sees h i m b u i l d on the
w o r k of the Presocratics and make some i m p o r t a n t advances i n both
astronomy and cosmology. I t is i m p o r t a n t t h a t a great number of those
views are repeated i n both the Laws and Epinomis, i n both cases i n ways
i n w h i c h w o u l d seem to indicate t h a t they are Plato's views at this stage.

VI. Writing and paidia

I take i t t h a t Plato's usual a i m i n w r i t i n g is not so m u c h to i m p a r t
knowledge, b u t rather to attempt to generate understanding, a n d so i n the
m a i n his purpose is dialectical rather t h a n dogmatic and t h a t he attempts
to create some sort of interaction between the reader and the t e x t . Merely 40

reading a n explanation w i l l not produce understanding i n the same w a y

t h a t reading a fact w i l l produce knowledge. So at Phaedrus 275d Socrates
41

tells us:

Writing has this strange property, in which i t is very like painting. For the
products of painting stand before us as i f alive, yet i f you ask them some­
thing, they are altogether solemnly silent. And so i t is with written words.
You might believe them to speak with some understanding, but if you should
question them wishing to learn their account, they always point out one and
the same thing. 42

Is w r i t i n g t h e n a singularly unhelpful m e d i u m for Plato? The fact t h a t he

wrote so much, a n d w i t h so m u c h care, suggests n o t . Perhaps t h e n there
43

is a v i a media for Plato between treatise w r i t i n g and oral dialectic, a style

t h a t attempts to create a form of dialectic between reader and text, m a k i n g
w r i t i n g a w o r t h w h i l e philosophical activity. Plato employs various liter­
ary/philosophic tools to achieve this objective, such as aporia, often pre­
s e n t i n g e x t r e m e a n d unacceptable a l t e r n a t i v e s as a n a p p a r e n t l y
exhaustive d i l e m m a , paradox, and complex irony. A common factor is
44 45 46

t h e i r ability to act as a challenge or a philosophical i r r i t a n t to the readers,

256 Plato's Philosophy of Science
forcing t h e m to do some t h i n k i n g for themselves, and to r e t u r n to read the
text at a different level of sophistication.
We also need to consider the contrast w h i c h Plato draws between spoude
(earnestness, serious) and paidia (child's play, sport, pastime). The key
m a t e r i a l here begins at Phaedrus 276b ff. w i t h a contrast between w h a t a
good gardener w o u l d do w h e n he is serious and w h a t he w o u l d do i n a more
playful mood. T h e n at Phaedrus 276d ff. Plato has Socrates say:

Fittingly, i t is for the sake of a pastime (paidias) that that he will sow seeds
and write in the garden of letters, when he writes, this being laid up as a
store of reminders, both against the onset of forgetful old age, and as a track
for all those who follow him, and he will be pleased to watch the growth of
these tender things ... but in my opinion, much better is the serious treat­
ment (spoude) of these matters, that is whenever someone employs the
method of dialectic, and taking a suitable soul, plants and sows words of
wisdom, which are able to defend themselves and their planter, and are not
barren but are pregnant with seed, whence other truths are brought forth in
other minds, rendering these seeds immortal, and making their possessor
happy, as far as is possible for humans.

One m i g h t begin to w o r r y here how seriously we are to take Plato's w r i t i n g

itself. T h i s is exacerbated at Phaedrus 277e-278a, where we are told:

The man who supposes that there is much that is playful (paidian) in the
written word concerning any subject, and that nothing (oudena) ever yet
written, whether in metre or in prose, is worthy of very serious (megales
spoudes) consideration ... Only (monois) in oral teaching for the sake of
learning are really written i n the soul things which are distinct, complete
and worthy of great trouble concerning justice, beauty and goodness.

So w h i l e Plato takes great care i n composing his works he states, i n

w r i t i n g , t h a t n o t h i n g w r i t t e n is w o r t h y of great pains, and is b u t a paidia.
Can t h i s paradox be resolved? Consider the way i n w h i c h Plato's theory of
w r i t i n g philosophy ought to be presented i n w r i t i n g . I t s principles should
apply reflexively, and i n line w i t h the dilemma strategy, Plato overdraws
the contrast between oral dialectic and w r i t i n g to the point where paradox
occurs, leaving the reader i n aporia, and t h a t this contrast is thoroughly
steeped i n the paidia and spoude terminology. The r e s u l t i n g book paradox
is self-referential and heavily ironic. I n order to resolve i t we need to shift
levels, and this should provoke us into considering the nature of Plato's
w r i t i n g . The challenge to the reader is to sort out the relationship between
the w r i t t e n and the spoken w o r d such t h a t the w r i t t e n w o r d can have some
m e r i t i n the r i g h t circumstances, and so avoid paradox, while not being the
equal of oral dialectic.
So the theory w h i c h says not to t r u s t the w r i t t e n w o r d b u t to examine
the t h o u g h t t h a t lies behind i t is itself not to be t r u s t e d as a w r i t t e n theory
but only as the thought behind the resolution of the w r i t t e n paradox t h a t
 Epistemology in the Timaeus and Philebus 257
expresses i t . We m u s t also consider t h e p a i d i a description and the paidia
4 7

and spoude contrast. Once w r i t i n g escapes the book paradox, i t is no longer

a worthless amusement. Certainly a w o r k should be w r i t t e n playfully i n
order to provoke us to investigate the thought t h a t lies behind i t , and
w r i t i n g is s t i l l inferior to oral dialectic, b u t i t has real value and serious
i n t e n t . Thus the paidia description is one t h a t carries a complex irony,
guaranteeing both the playfulness and the seriousness of a piece of w r i t ­
ing, h i g h l i g h t i n g b o t h its merits and its drawbacks as w r i t i n g .
This is of some importance, for while Plato does not have a fixed
philosophical terminology, some terms b u i l d up resonances, and the con­
t r a s t between spoude and paidia and the paidia description are t w o of
these. I m p o r t a n t examples of the paidia and spoude contrast are at
Republic 545e, 602b, 608a, Phaedrus 276b, 276e, 277e, 278a, Sophist 237b,
Politicus 288c, Philebus 30e, and Laws 688b. I m p o r t a n t examples of the
48

paidia description i n relation to the status of a piece of w r i t i n g are at

Republic 536c, Parmenides 137b, Theaetetus 161e, Politicus 268d, Timaeus
59cd and Philebus 28c. This widespread and consistent usage of paidia
and spoude w o u l d suggest a coherent strategy on Plato's p a r t of r e m i n d i n g
his readers of his views on w r i t i n g .
One perennial w o r r y about Plato's philosophy of science is whether he
believed t h a t the investigation of nature could yield any w o r t h w h i l e
results. O n some accounts, the Timaeus states t h a t such w o r k is a mere
childish amusement, t h a t here we can only have a l i k e l y account, and can
aim no higher t h a n to have unstable opinions concerning the physical,
sensible and changeable things t h a t make up the w o r l d about us. This is
sometimes argued on the grounds t h a t at 59d Timaeus tells us t h a t
providing accounts of the phenomena w i l l be pleasant, and t h a t anyone
doing so ' w i l l be creating i n his life a pastime (paidian) b o t h suitable and
prudent'. However the use of paidia and its cognates has certain reso­
nances i n later Plato, and i f this is so here, t h e n this is a very positive t h i n g
for Plato to say. Rather t h a n relegating the Timaeus to a mere harmless
diversion, the paidia description elevates i t to a serious w o r k where we
perhaps ought to be looking beyond the playful facade to the arguments
and debates r u n n i n g beneath the surface. So perhaps we should not take
the Timaeus as a direct statement of cosmology, b u t as an interesting and
playful series of hypotheses t h a t w i l l lead us to the serious thought t h a t
lies behind them. A n i m p o r t a n t comparison here for the Timaeus is w i t h
the Philebus. One w o u l d h a r d l y w i s h to say t h a t the Philebus is not a
philosophically serious w o r k on the grounds t h a t Philebus 28c and 30e7
use paidia, i n both cases Socrates commenting on the nature of the
discussion. I see no reason, at the outset, to suppose t h a t the Timaeus is
not a philosophically serious w o r k or t h a t Plato's general theory of w r i t i n g
does not operate i n i t . I f the Timaeus is a joke, or some other form of
amusement, i t is no more and no less so t h a n other late period w o r k s . 49
258 Plato's Philosophy of Science

V I I . Plato and myth

Let me t r y to undermine the idea t h a t we should take w h a t Timaeus has
to say at face value i n another way. The account of the Timaeus is
presented as a m y t h . H o w should we deal w i t h that? For a good deal of
Plato's later cosmology, this is not a major concern as m u c h of the critical
m a t e r i a l is presented i n non-mythical form i n other works (the Philebus,
and especially the Laws and Epinomis). However, the attitude we take to
Plato's use of m y t h w i l l affect the status of he account of the Timaeus.
There is a problem w i t h Plato's use of m y t h i n the Phaedrus, i n t h a t he is
apparently critical of i t b u t s t i l l employs i t . Socrates is not interested i n
discussing the m y t h t h a t Boreas abducted Orethuia, because this w i l l not
help h i m w i t h the Delphic injunction to know himself. N o r is he inter­
50

ested i n the tales of monsters t h a t are part of this m y t h . A t Phaedrus 230a

he says:

I investigate myself rather than these things, to see if I am a more complex

and fearsome beast than Typhon, or a gentler and more single-minded
creature, who has by nature a less puffed up fate appointed by God. 51

Given t h a t Plato overdraws the contrast here to w o r r y his readers about

the status of m y t h , how m i g h t some myths be philosophically interesting
and able to help w i t h the Delphic injunction, and others not? One m i g h t
draw a distinction between those myths w h i c h are i n a sense purely
historical, where the only judgements we m i g h t make are whether to
accept the general form and particular details of t h a t m y t h , or how to 52

rationalise t h a t m y t h , and those t h a t call upon the reader to make some

53

philosophical judgements, perhaps even offering various criteria on w h i c h

they m i g h t be judged. The problem w i t h the m y t h of Boreas and O r e t h u i a
is t h a t i t falls i n the first category here. This is i n sharp contrast to the
m y t h t h a t Socrates offers at 246a ff., w h i c h deals w i t h the h u m a n soul and
commentators have found to be a mine of information, and w i l l help the
readers w i t h the Delphic injunction, as i n coming to a general under­
standing of the h u m a n soul they w i l l know themselves better. 54

W i t h t h a t i n m i n d , consider the treatment of the m y t h of Solon. Socrates

wishes to hear an account of the state of the Republic i n action. A t 19b ff. 55

Socrates compares his description of the ideal state w i t h a p a i n t i n g

depicting animals. I n both cases, he says one wants to see the subject i n
action. This seems to be an allusion to Phaedrus 275d and the discussion
of the nature of w r i t i n g . The function of this, I suggest, is to t h r o w doubt
on b o t h m y t h and w r i t i n g as vehicles for philosophy. Timaeus 19b ff.
56

suggests t h a t some sorts of m y t h are useful b u t others are not. The

problem here is exacerbated by the fact t h a t at the beginning and the end
of i t , Solon's m y t h is described as wholly t r u e . W h a t t h o u g h is the nature
57

of the w a r r a n t for its t r u t h ? Several criteria are suggested, a l l of w h i c h are

 Epistemology in the Timaeus and Philebus 259
external to the readers, a l l calling upon t h e m to accept the m y t h on
someone or something else's authority. A l l the criteria alluded to here are
ones t h a t Plato has given us grounds to be suspicious of i n earlier works.
A t Timaeus 20d, we are asked to accept the m y t h because Solon is wise
and old. Yet surely we have seen enough of the fate of the 'experts' i n the
early dialogues not to p u t our t r u s t i n the testimony of an eminent man,
whatever his credentials. I n the Phaedrus, Socrates upbraids Phaedrus for
being interested i n the pedigree of a m y t h rather t h a n its content. A t 2 I d
58

Solon's sources are questioned, and at 23a we find t h a t the source is the
ancient w r i t i n g s of the Egyptians. The Phaedrus, however, has w a r n e d us
not simply to accept the w r i t t e n word, whatever its provenance, and the
reference here to Egypt may be a further allusion to the Phaedrus i n
addition to Timaeus 19b. A t 26a ff. a different tack is taken. Perhaps the
59

story is t r u s t w o r t h y because i t is well remembered. However, at 19a

Socrates finishes his synopsis of the Republic, on w h i c h he should be the
expert, h a v i n g discoursed the previous day, by saying t h a t this completes
the contents, and asks i f Timaeus agrees t h a t this is a complete l i s t of the
contents, w h i c h he does. This synopsis though is far from complete.
Memory, then, is not to be trusted. The overall effect of Timaeus 19b-27a
is t h e n to make us suspicious of all external authorities i n j u d g i n g a m y t h
and to make us rely on our own i n t e r n a l criteria. I n contrast to the m y t h
of Solon, which deals w i t h historical matters, Timaeus' m a i n account not
only deals w i t h the h u m a n soul and its place i n the universe, b u t also
discusses the criteria by w h i c h we should be j u d g i n g the contents of the
m y t h , one t h a t invites philosophical rather t h a n mere historical judge­
ment. T h a t i n t u r n may affect debates about whether we should take the
tale of creation literally, or more allegorically, setting up a framework of
postulates about the order of the universe w i t h i n w h i c h to discuss cosmo­
gony and cosmology Our suspicions ought also to be aroused by the fact
t h a t Timaeus is described as an expert i n astronomy, remembering the 60

fate of other experts i n earlier works.

V I I I . I n t e r p r e t i n g the Timaeus
Two further thoughts on u n d e r m i n i n g the view t h a t the Timaeus presents
us w i t h dogma. I n Chapter 5 we saw t h a t the best way to view the model
of the heavens t h a t Plato has Timaeus p u t forward is not as a l i t e r a l piece
of cosmology, b u t rather as something w h i c h embodies i m p o r t a n t cosmo­
logical principles while having flaws i n its astronomy. I n short, i t is a
prototype of this sort of cosmology and astronomy for others to develop. I
have also argued t h a t i f we take Plato be presenting us w i t h a straightfor­
w a r d cosmology and cosmogony, his views on chaos, flux and the m a t e r i a l
nature of the receptacle appear confused i f not contradictory. A more
generous way of viewing these issues is t h a t Plato presents us w i t h a range
of options and a general way of dealing w i t h them. So whatever chaos is
260 Plato's Philosophy of Science
supposed a teleological ordering is required to produce the current cosmos,
and no version of a m a t e r i a l receptacle can do w h a t is being asked of i t .
Thus Plato's treatment of these issues is rather more flexible and more
powerful, and one m i g h t note t h a t although he does not give a full, direct
cosmology there are principles w h i c h he wishes to get across.
I w i l l now p u t forward two suggestions for how we m i g h t go about
reading the Timaeus. B o t h are tentative, b u t both lead to some interesting
perspectives on w h a t the Timaeus may be t r y i n g to do. The Timaeus, i f one
reads i t carefully, can be a very difficult w o r k to p i n down. Sometimes one
m i g h t feel t h a t reading the Timaeus is rather (frustratingly!) l i k e looking
at a kaleidoscope. Each t i m e you look at i t anew, t a k i n g a fresh s t a r t i n g
point, a different p a t t e r n emerges. U n l i k e a kaleidoscope, though, certain
patterns keep emerging. Depending on w h a t you hold to be fundamental
i n the Timaeus, and how one interprets certain key passages i n order to
accord w i t h those fundamentals, different, non-compatible b u t r e c u r r i n g
patterns emerge. T h a t is somewhat disorienting, and m i g h t cause one to
wonder w h a t precisely is going on i n the Timaeus. Perhaps, though, t h a t
is a deliberate effect t h a t Plato wishes the Timaeus to have on i t s readers.
W h y m i g h t t h a t be so? I n some works, the nature of the discourse appears
to be related to the nature of the subject matter. Thus one m i g h t argue t h a t
i n the Philebus the key ideas of l i m i t and u n l i m i t e d and t h e i r relation to
philosophy i n p a r t shape the nature of the discourse between Socrates and
his interlocutors. W h a t I w a n t to suggest here is t h a t the problems the
reader is presented w i t h i n dealing w i t h the account t h a t Timaeus gives
are like those t h a t Plato believes are critical i n beginning to give a proper
account of the physical world. Quite simply, w h a t do we take as fundamen­
tal? I f we make certain assumptions about the physical world, t h e n how
w e l l do they accord w i t h our other beliefs and how coherent and philo­
sophically satisfactory is the broader p i c t u r e account t h a t we can
generate? Perhaps this is also w h y the Timaeus is i n the m a i n a mono­
logue. We cannot have dialectic and elenchus directly w i t h the physical
w o r l d , so we need to make assumptions and see how these pan out, to see
whether these account for w h a t we see and whether they are philosophi­
cally acceptable. So too w i t h the monologue t h a t we are given by Timaeus.
I f I a m r i g h t about Plato's strategy here, t h e n he is asking his readers to
do quite a lot i n terms of t h i n k i n g for themselves and interacting w i t h the
various problems t h a t the Timaeus m i g h t set. T h a t though w o u l d make
61

the Timaeus stylistically rather closer to other later period works (and here
I mean philosophical and l i t e r a r y style rather t h a n stylometry). This
brings us back to the dilemma posed between T W and the stability of the
heavens and stoicheia w h i c h I suggested earlier. The choice between the
two is critical to the understanding of the w o r l d about us, and critical to
the account w h i c h Timaeus gives.
W h a t relation does the Timaeus hold to other late works? Here I shall
introduce m y second suggestion for reading the Timaeus. Burnyeat, i n his
 Epistemology in the Timaeus and Philebus 261
book on the Theaetetus, proposes two readings, A and B, for the first section
on the question of whether knowledge is perception. O n reading A, Plato
62

h i m s e l f accepts the views of Protagoras and Heraclitus provided t h a t they

are restricted to perception and the sensible world. Sensible things are
t h e n i n a 'perpetual flux of becoming', and each person is the judge of
63

w h a t sensible qualities they perceive i n t h i s flux. So according to

Burnyeat:

Thus the argument, taken as a whole, supports the view set forth at length
in earlier dialogues like the Phaedo and the Republic, the celebrated Pla­
tonic doctrine that true reality is a nonsensible realm of changeless being,
the Forms, and i t is in these alone that knowledge can find its objects. 64

O n reading B , however, Plato does not accept the views of Protagoras and
Heraclitus. Rather, i f one wants to hold t h a t knowledge is perception
(Theaetetus' thesis) t h e n one must also hold a Protagorean epistemology
and a Heraclitean account of the world. I f these are necessary and suffi­
cient conditions for Theaetetus' view, t h e n a l l three stand, and u l t i m a t e l y
fall, together. 65

T h a t one can generate something a k i n to the position of the Theaetetus

reading A from the Timaeus is w e l l k n o w n . We simply accept the statement
of T W at Timaeus 27c-29e as Plato's own view, the discussion of the
elements and the receptacle gives us a flux, w h i c h meshes w i t h the p a r t i a l
imposition of reason on necessity where necessity can be t a k e n as chaos.
So Plato holds a Heraclitean view of one realm and a Parmenidean view
of the other. One m i g h t take the Theaetetus and Timaeus to be i n rough
agreement here, i f one goes for the A reading of the Theaetetus, or argue
t h a t the Timaeus is early (should be t a k e n w i t h the Phaedo and Republic,
and prior to the Parmenides and Theaetetus) i f one goes for the B reading
of the Theaetetus. Some commentators, notably Cornford, use w h a t is
effectively an A reading of the Theaetetus as a reference point i n attempt­
i n g to establish an A reading of the Timaeus. 66

W h a t I w a n t to suggest is that, as i n the Theaetetus, we have an

alternative to the A reading i n the Timaeus 61
This is based around the
regularity and stability of the heavens, the stable stoicheia and the
alternative epistemology of the analogy w i t h the w o r l d soul. We need not
take i t t h a t Plato is committed to a Heraclitean account of the physical
w o r l d i n either work. The Theaetetus is aporetic, and consciously so. W h i l e
there is no successful definition of knowledge, Theaetetus is better for
realising he does not know w h a t he though he knew. One task of the
68

Timaeus may be to offer us the TW/Heraclitean account again, this t i m e

t a l k i n g i n greater detail about the nature of the physical world, b u t
offering something more constructive as an alternative account of t h a t
w o r l d . There are some interesting parallels between the Theaetetus and
69

the Timaeus. The Timaeus gives us a model of the m i n d , t i e d to a discus-

262 Plato's Philosophy of Science
sion of the psychology of false judgement (using s i m i l a r examples to the
Theaetetus), a discussion of flux, of perception, of the relation of language
and the w o r l d and perhaps most importantly, a discussion of letters and
syllables. I f I a m r i g h t t h a t at least for the physical world, at the end of
70

analysis by decomposition we can switch to a teleological account of the

physical stoicheia, t h e n the Timaeus provides at least a p a r t i a l answer to
one of the key aporia of the Theaetetus. The implications of t h a t need
investigation on a wide stage, and that, arguably, is w h a t the Timaeus
provides.

I X , T h e Philebus a n d d i v i n e cosmology
The Timaeus, then, presents us w i t h some alternatives, w i t h the onus
placed on us to make a decision about our own beliefs. I f we w i s h to hang
on to TW, t h e n we need to adopt the commensurate views on the more
indeterminate passages. I f we believe t h a t the stability of the heavens and
stoicheia is fundamental, t h e n we may need to reconsider how strong a
version of T W to accept, i f any. The l a t t e r may give us grounds for greater
epistemological optimism. The w o r l d soul can have opinions w h i c h are
t r u e and stable, and so can we, i f we w o r k h a r d enough, especially at
astronomy. Is this greater epistemological o p t i m i s m t o w a r d the investiga­
t i o n of nature reflected i n other late works? I believe this to be so for the
Philebus, Laws and Epinomis. I n the Philebus, the discussion of k n o w l ­
edge, techne and measurement from 55d on is highly optimistic and w o u l d
seem to allow there to be different types of knowledge, arranged i n a
hierarchy from the most accurate and abstract down to those w h i c h rely
very m u c h on empirical knowledge w i t h o u t a great deal of measurement.
A t Philebus 55c, Socrates considers whether some types of cheirotechni-
kais, m a n u a l crafts, are closer to knowledge t h a n others. He goes on to
71

say t h a t i f we abstracted arithmetic, measuring and weighing from a l l

branches of the technai, a l l we would be left w i t h w o u l d be guesswork
learned by empeiria kai tini tribei, an empirical knack and g r i n d i n g away. 72

Philebus 56a ff. praises those technai w h i c h employ number and measure­
ment against those w h i c h employ guesswork, and considers t h e m to be
closer to the t r u t h . E a r l i e r i n the Gorgias Plato had reckoned t h a t i t was
the possession of a 'rational account' t h a t distinguished a techne from mere
empeiria.13
Here we see something w h i c h is typical of Plato's later devel­
opment, w h i c h is a greater mathematical orientation i n explanation. This
is not to suggest t h a t the earlier account excluded mathematics or t h a t the
later excludes more discursive accounts, the subject m a t t e r helping to
determine w h a t is appropriate, b u t rather there is a significant change of
emphasis.
According to Philebus 56a, music, at least i n terms of flute p l a y i n g and
lyre playing, uses very l i t t l e of arithmetic, measuring and w e i g h i n g any­
way, and the actual playing of an i n s t r u m e n t requires a good deal of
 Epistemology in the Timaeus and Philebus 263
practice i n finding the r i g h t notes. This is also the case w i t h medicine,
farming, piloting and generalship. O n the other hand, there are technai
w h i c h use instruments, such as a compass, and a rule, typically the
b u i l d i n g crafts such as ship b u i l d i n g and house building. H a v i n g distin­
guished two sorts of technai, those using measurement being the more
exact, Socrates t h e n proceeds at 56cd to distinguish two sorts of arithme­
tic, t h a t of the many and t h a t of the philosophers. Some people reckon w i t h
unequal units, such as two armies, two bulls, or p a i r i n g the smallest and
greatest of things, while others demand t h a t a l l the units are equal. So one
can contrast too calculation and measurement used i n b u i l d i n g and trade
w i t h philosophical geometry and diligent calculations and distinguish two
sorts of arithmetic. The conclusion t h e n at Philebus 57ce is t h a t there are
different types of knowledge (episteme) w h i c h can be arranged i n order of
the clarity and accuracy they possess. Throughout a l l this, Socrates has
been quite happy to t a l k of knowledge i n conjunction w i t h the technai.
There seems to be no question here t h a t to be i n possession of a technai is
to possess knowledge of some sort.
I t is also interesting to note the order of the technai here, w h i c h i n some
ways is quite surprising. The various technai of b u i l d i n g are rated higher
t h a n those of medicine and generalship on this account. Perhaps though,
coming after the demiourgos, the craftsman god of the Timaeus, t h a t is not
such a surprise, though i t does r u n against the famous comment from
Xenophon, t h a t 'what are called the mechanical arts carry a social stigma,
and are r i g h t l y dishonoured i n our cities'. 74

Following the discussion of various types of pleasures, at Philebus 61d

Socrates returns to t a l k i n g about knowledge again, and we are told:

So knowledge (episteme) differs from knowledge, on the one hand concentrat­

ing on the things that undergo generation and destruction, on the other on
things that do not suffer generation and destruction and are always the same
and similar.

Furthermore, these lower grades of knowledge are accorded some consid­

erable importance for practical decision m a k i n g , as the following exchange
between Socrates and Protarchus at Philebus 62a ff. demonstrates:

Socrates: Let us suppose then that there is a man who understands what
justice itself is, and has an account in accordance with his understanding,
and moreover that he has the same sort of thoughts concerning all other
things.
Protarchus: Let that be supposed.
Socrates: Now will this man have sufficient knowledge, having account of
the divine circle and sphere, but being ignorant of the human circle and
sphere, even if he employs these circles and other rules in house building? 75

Protarchus: We ought to call it a ridiculous state, Socrates, for someone to

have divine knowledge only.
264 Plato's Philosophy of Science
Socrates: What is this you say? Are we to throw the unstable and impure
skill of the false and circle and rule into our mixture?
Protarchus: Necessarily, if we are to be able to find the road home on each
occasion.
Socrates: And music, which we just previously said was full of guesswork
and imitation and was not pure?
Protarchus: That seems necessary to me, if our life is to be any sort of life
at a l l .
76

The last comment is highly significant here as one of the m a i n themes of

the Philebus has been the nature of the life we m i g h t have i n relation to
the debate between intelligence, good and pleasure. There is the sugges­
t i o n t h a t i f we lived a life solely of pleasure, and so entirely l a c k i n g i n
intelligence, our life would i n fact be like t h a t of a shellfish. I n order to
have a reasonable life, i t seems we need the knowledge of the practical
technai as w e l l as t h a t of the philosophers. W h a t is not i n doubt again here
is t h a t to be i n possession of a techne is to be i n possession of knowledge of
some form, and of useful knowledge. The Philebus i n this respect seems to
have an even greater epistemological o p t i m i s m to the investigation of
nature t h a n the Timaeus. The whole discussion of the technai here is
notable for its lack of any strong T W assumptions. For the Philebus i t
seems a straightforward m a t t e r t h a t we can have some k i n d of useful
knowledge of the w o r l d about us. I t is also perhaps significant t h a t the t w o
leading examples of techne i n the Philebus' discussion of letters and
syllables, and later i n the discussion of the techne and types of knowledge,
are music and grammar, precisely those examples at Theaetetus 207b,
where the expert's discipline differs from t h a t of the child.
 10

Conclusion

A t the outset I h i g h l i g h t e d several key issues for Plato's philosophy of

science. I w i l l now look at how the arguments of this book affect those
issues and how we should assess Plato's conception of nature and the
investigation of nature.

I . The role of empirical work

The anti-empirical charges against Plato's conception of science are i l l
founded. We have looked i n detail at the two passages most often cited by
Plato's detractors, Republic V I I and Timaeus 68b-d. I n the Republic pas­
sage, Plato draws a distinction between teaching astronomy as p a r t of the
education of the guardians and doing astronomy as an investigation of the
heavens. W h i l e we may w e l l leave the things i n the heavens alone' i n the
former, the contrast implies t h a t we do not i n the latter. This is entirely
consonant w i t h the conception of investigation advanced i n the epistemo­
logical allegories. Even i f he held T W at this point, there is no demarcation
for investigation at the border of the two realms.
A t Timaeus 68b-d, Plato appears to be concerned w i t h the proportions
of different sized fire particles w h i c h constitute a colour rather t h a n w i t h
m i x i n g pigments. Certainly he believes t h a t this is something t h a t cannot
be found out by practical tests, b u t t h a t w o u l d appear to be due to
something quite specific to this problem, and may reveal a relatively
sophisticated doubt about w h a t can be found out by such tests. I t is
significant t h a t Plato's consistent conception of the technai allows l e a r n i n g
of a l i m i t e d sort by empeiria kai tribe, by t r i a l and error and g r i n d i n g away,
and so is consonant w i t h his views i n the Republic and Timaeus passages.
This is not to say t h a t there are not weaknesses i n relation to empirical
w o r k i n Plato's conception of science. Plato is somewhat condescending
about the empirical parts of the technai, and i n general underrates both
the importance and the extent of the empirical bases for most disciplines.
Against t h a t i t m u s t be recognised t h a t he holds t h a t there are experts i n
these technai, m e n who have mastered the empirical basics and have gone
on to a proper sytematised knowledge, p a r t i c u l a r l y i n medicine, and t h a t
he is fulsome i n his praise of the accuracy of the E g y p t i a n astronomical
observations. One weakness shown up by the nature of the Timaeus is t h a t
266 Plato's Philosophy of Science
Plato does not envisage a dialectic between investigators and nature. A t
one level, of philosophical dialectic, t h a t is so, b u t modern philosophers of
science w o u l d stress very strongly the interaction between theory and
experiment w h i c h is so fruitful for modern science. Plato does not ignore
empirical evidence, b u t there is no sense of the dynamic i n t e r p l a y w i t h
nature w h i c h allows modern science to make r a p i d progress.
That, however, could be said of v i r t u a l l y a l l Presocratic science, w i t h
very few exceptions, the Hippocratics perhaps being the likeliest candi­
dates here. Plato may have weaknesses i n this area b u t they are not so
great as to pick h i m out as notably weak against the background of
Presocratic science. A criticism t h a t m i g h t be levelled at m u c h of Greek
science is t h a t m a n y of the contributors presented t h e i r w o r k as the final
w o r d on the m a t t e r and i n a finished system. I t is very rare to find anyone
leaving possibilities open or indicating t h a t further research along certain
lines w o u l d be fruitful. I n this sense Greek science i n general tended to be
dogmatic. Plato, to some degree at least, can be absolved of this, especially
i f we consider the Timaeus to be more about how to do cosmology t h a n
specific cosmology itself. One can perhaps see this i n the presentation of
geometrical atomism at 54a, where anyone who could give an argument
for better basic entities t h a n the two types of triangle w o u l d be welcomed
as a friend rather t h a n a foe. Plato was w e l l aware of m a n y of the
deficiencies of the very simple astronomical model he puts forward i n the
Timaeus, and so could not have meant this model to be t a k e n l i t e r a l l y or
dogmatically.
Observation and experiment are key issues not only because they are
i m p o r t a n t for science, but also because negative assessments of Plato's
views have been used i n a number of scenarios to attack his conception of
science and to downplay its significance i n the ancient world.

I I . P l a t o as a scientific (proto-)realist
Plato can best be seen as someone who is struggling w i t h the foundations
of scientific realism. I f we adopt this strategy for looking at Plato's concep­
t i o n of science, i t allows us to explain w h y he takes certain stances t h a t
m i g h t otherwise be thought to be rather odd. Regularity is something
w h i c h needs to be explained for Plato, i n particular the r e g u l a r i t y of the
heavens. Lacking conceptions of gravity and force, Plato makes the cosmos
and a l l the heavenly bodies animate, intelligent deities w h i c h obey good
laws because they themselves are good. I n short, lacking the idea of
physical l a w Plato uses an analogue from civil law. Plato is not alone i n
using biological and political metaphors to explain the i n t e g r i t y and
behaviour of the cosmos. Plato does not revert here to supernatural or
animistic explanations. He strips these l i v i n g beings of a l l b u t the neces­
sary attributes of self-motion and intelligence. These are not capricious
entities, b u t being gods w i l l always choose the good thus guaranteeing and
 10. Conclusion 267
explaining the regularity of the heavens. Plato makes use of the h i g h l y
l i m i t e d tools he has available to make sense of the regularity of the
heavens. A n i m p o r t a n t part of Plato's reaction to the atomists is t h a t t h e i r
m u l t i p l e cosmoi w i l l not support a n y t h i n g more t h a n a de facto r e g u l a r i t y
theory.
We also need to understand Plato's reaction to Meno's paradox/under-
determination i n relation to his proto-realism. I n order to be a realist, one
m u s t be able to justify a choice from an indefinite field of possibilities
where others deny the v a l i d i t y of any further criteria. This can be seen
most clearly i n choice of shapes for the u l t i m a t e particles, b u t the effects
r u n deep throughout Plato's approach to problems i n science and the
choice of regular circular motion for the heavens is also i m p o r t a n t here.
Plato is clearly not an instrumentalist. There is no m o t i v a t i o n towards
i n s t r u m e n t a l i s m i n his views, as he has a means of dealing w i t h underde­
t e r m i n a t i o n and is u n l i k e l y to be worried by other anti-realist arguments.
Hypotheses w h i c h save the phenomena have a rather different meaning
for Plato, especially as he asks for regular, circular and ordered motion.
Someone who believes we can escape from the cave, and has a solution to
Meno's paradox has no need to be an instrumentalist. However, neither
can we cannot assimilate Plato to modern scientific realism. He is strug­
gling w i t h the foundations of this approach i n the philosophy of science,
he lacks i m p o r t a n t scientific and philosophical tools, and the ancient
debates differ significantly from the modern ones. W h a t we can do though
is appreciate w h a t Plato was t r y i n g to do, and narrow the gap w i t h modern
philosophy of science. W h i l e there may appear to be a y a w n i n g chasm
between Plato and modern positivist and i n s t r u m e n t a l i s t accounts of
science, not to mention relativist and sociological analyses, there is a
considerable affinity w i t h modern scientific realism. T h a t Plato was a
proto-realist is also of considerable importance i n understanding his mo­
t i v a t i o n for a t t e m p t i n g to solve some difficulties i n the philosophy of
science by the use of teleology.

I I I . R e a s o n s for teleology
Plato has interesting epistemological, cosmological and nomological rea­
sons for teleology, i n addition to any other reasons he may have. These
reasons are a l l related to his proto-realism. Doubtless Plato has other
reasons for teleology as well, b u t i t is i m p o r t a n t to recognise the role t h a t
teleology plays i n allowing Plato to answer certain problems i n a realist
manner. Plato's answer to underdetermination/Meno's paradox is to give
p r i o r i t y to hypotheses w h i c h cite the good. So too w i t h the epistemological
problems of Socrates' dream. A t least for physical entities, at the end of an
analysis by decomposition into component parts, we have a teleological
account of the stoicheia.
Plato also has cosmological reasons for advocating teleology. A key
268 Plato's Philosophy of Science
question for cosmogony and cosmology is how the cosmos acquired its order
and how i t maintains i t , given t h a t cannot be explained i n terms of m a t t e r
and forces alone. A survey of the history of science, r i g h t down to the
present day, indicates two recurring strategies for answering this ques­
tion. One is to suppose t h a t this cosmos is part of an array of many, perhaps
indefinitely many differing cosmoi, so i t is no great surprise t h a t one of
these has order. The other is to suppose further, typically teleological or
theological principles to explain this unique cosmos, especially i f one holds
t h a t m u l t i p l e cosmoi are non-explanatory and ontologically profligate.
So too Plato has nomological reasons for employing teleology. To meet
the demands of his proto-realism i n the absence of theories of gravitation,
force and physical law Plato employs a teleological account based on
biological and political analogies. I t is i m p o r t a n t here t h a t because the
'souls' of the stars are good, they always act i n the best possible manner
and so i n a perfectly regular manner.
Plato's employment of teleology is not based on a metaphysical fairy
tale. I t is an attempt to answer real and perennial problems i n the
philosophy of science i n a realist manner, w i t h extremely l i m i t e d re­
sources. Plato's answers have strong affinities w i t h modern realist
positions, the more so as his conception of the good becomes more techni­
cally and mathematically oriented i n later works, and so his solution to
underdetermination looks more modern. A t the strategic level, the project
of using further explanatory principles for a unique universe is s t i l l alive
a n d w e l l . I t is less o v e r t l y teleological, b u t Plato's is the f i r s t a t t e m p t at
a teleological cosmos, and we can forgive h i m for not being possessed of
t w o m i l l e n n i a of h i n d s i g h t . The question of w h a t underpins r e g u l a r i t i e s
is helped by the m o d e r n conception of physical law, t h o u g h i t m i g h t w e l l
be argued t h i s only tells us how entities obey laws r a t h e r t h a n w h y t h e y
do so.
Teleology is an i m p o r t a n t issue as i t is often thought to be the Achilles'
heel of Plato's conception of science, along w i t h observation and experi­
ment. I f we can understand w h y Plato felt the need for teleology i n science,
and w h y t h a t m i g h t have been a reasonable move given the context of
fourth-century Greek science and his realist goals, this w i l l go a long way
towards a more positive assessment of his views. Plato's teleology is a p a r t
of an attempt to answer some serious and interesting problems i n the
philosophy of science i n an original and systematic manner. This teleology
is a l l the more remarkable for the lack of significant predecessors. We
m i g h t to some extent look to Anaxagoras, and to the perhaps Socratic
conception of the craft analogy for knowledge leading u l t i m a t e l y to the
demiurge, or a snippet or two from Presocratic literature. Plato t h o u g h
gives us the first comprehensive and considered principles of a teleological
cosmogony, cosmology and science.
 10. Conclusion 269

IV. T h e a i m s of the i n v e s t i g a t i o n of n a t u r e
Plato believes the investigation of nature w i l l produce w o r t h w h i l e results.
This needs to be discussed i n two parts. First, w h a t sort of knowledge w i l l
be generated by the investigation of nature? The T W view of the middle
period m i g h t give us cause for concern here, though as I have argued,
investigation i n the Republic is not w i t h o u t empirical content nor does i t
recognise a demarcation between the two realms. B y the t i m e of the
Timaeus and the Philebus however, a more optimistic attitude appears to
be t a k i n g over. I n the Timaeus, Laws and Epinomis i t would seem t h a t
very good results are envisaged, at least for astronomy. This m i g h t be
argued to be a special case, though: the one instance where we have
physical entities w h i c h are both qualitatively entirely stable and amen­
able to precise mathematical description. The analogy w i t h the w o r l d soul,
however, m i g h t suggest t h a t as our false opinions are correctable, and the
w o r l d soul can have t r u e and stable opinions, there may also be more
optimism towards other aspects of the investigation of nature.
The technai quite generally are seen as producing useful results. Plato
never doubts, for instance, t h a t there can be experts i n medicine as w e l l
as the more craft oriented technai, and t h a t they w i l l be able to apply the
results gained by the technai fruitfully, whether they are gained by t r i a l
and error or by a more sophisticated analysis of the nature of the topic.
The Philebus gives us a very positive view of the technai, w i t h a hierarchy
beginning w i t h those t h a t rely most on empeiria kai tribe, to those t h a t
employ arithmetic, measuring and weighing to those w h i c h deal w i t h
theoretical matters. The Philebus is quite happy to see this as a hierarchy
of differing types of knowledge, a l l of w h i c h are i m p o r t a n t i f we are going
to be able to live properly
The second p a r t of the discussion of this issue is w h a t effect the
successful study of nature w i l l have for humans. The Timaeus is very clear
on the benefits of the study of astronomy. B y studying the visible revolu­
tions of god, i.e. the heavens, we w i l l help to stabilise the revolutions i n
our o w n heads and so become r a t i o n a l and knowledgeable. T h a t w i l l of
course help w i t h the standard Platonic project of becoming like god. So too
a standard m o t i f i n Plato is t h a t studying theoretical sciences is i m ­
mensely i m p o r t a n t for the education of the young, and quite generally as
a road to the cognisance of the good.
H a v i n g said t h a t , one great weakness of Plato's conception of science
from a modern point of view is the lack of practical benefits and technology.
Since t h a t is arguably the case w i t h v i r t u a l l y every conception of science
u n t i l Francis Bacon at the beginning of the seventeenth century, I h a r d l y
t h i n k we can single Plato out for criticism here. I t is w o r t h noting,
however, t h a t for Plato a l l the technai w i l l do good. T h a t is, a proper
practitioner of a techne w i l l b r i n g about order and structure i n his subject
matter, be he a doctor, a builder, an astronomer or a cobbler.
270 Plato's Philosophy of Science
W h i l e Plato employs m y t h , this is no mere story-telling, something
w h i c h is strongly criticised, b u t a tool w i t h specific philosophical ends.
Plato is not r e v e r t i n g to p r e - M i l e s i a n m y t h t e l l i n g about t h e n a t u r e of
t h e cosmos, b u t is often challenging his readers to t h i n k c r i t i c a l l y about
the mythological account. M u c h of the astronomy and cosmology of the
Timaeus is anyway repeated, w i t h o u t mythical setting or l i k e l y account
qualifiers i n the Laws and Epinomis. Again, while there are weaknesses
and oddities to the results Plato envisages from an investigation of nature,
they do not pick h i m out for special criticism against the background of the
Presocratics.

V. Plato's d e v e l o p i n g p h i l o s o p h y of s c i e n c e
Plato's views on the nature of science, and on astronomy and cosmology,
develop and do so i n interesting ways. Plato's major contributions to
astronomy are the first model of the heavens w i t h an offset ecliptic/zodiac,
and the principle t h a t the motions of a l l the heavenly bodies can be
resolved into combinations of regular circular motions, both of w h i c h are
significant advances on the middle period views. The model of the Timaeus
is best seen as a prototype embodying the new principles rather t h a n as a
finished article. I find i t highly unlikely t h a t Plato was not aware of at least
some of the severe drawbacks of his model, and u t t e r l y improbable t h a t he
hoped to deal w i t h these w i t h the 'contrary power', w h i c h may simply
m a r k out t h a t there is something odd about M e r c u r y and Venus. Doubtless
Plato's relation to Eudoxus w i l l r e m a i n controversial, not least for lack of
h a r d evidence, b u t some form of close alliance between the two seems
reasonable to suppose. Certainly the standard objections to a significant
relation between t h e m , t h a t Plato was anti-empirical, t h a t he d i d not
believe i n regular circular motion, t h a t there was a radical difference i n
philosophies of science between t h e m fail. The Platonic challenge, i f Plato
d i d issue i t , and there are no philosophical or scientific reasons w h y he
should not have done, was of enormous significance to astronomy r i g h t
down to Copernicus and Kepler.
I n terms of cosmogony and cosmology, Plato gives us the first thorough­
going teleological account. The m i n u t i a e of the development of Plato's
cosmological models are also of some significance. The model of the Repub­
lic is supported on a pivot, requires braces, needs an external source of
motion, gradually degenerates and is not alive. B y the Timaeus, the
cosmos is seen as exhaustive of a l l there is, requires no pivot or support,
no bracing, moves itself and is an i m m o r t a l e n t i t y not subject to any
degeneration. J u s t as the Presocratics developed theories about the na­
ture, position and stability of the earth, so Plato develops ideas about the
cosmos. The move to regular circular motion, or combinations therof, and
to a stable cosmos is a significant one against a background of philosophy,
science and l i t e r a t u r e w h i c h assumes a degenerating cosmos. So too is the
 10. Conclusion 271
move from the Pythagorean arithmetical conception of the relation of
mathematics to the w o r l d to a geometrical model.
Of great importance is the meeting of the traditions of observational
astronomy and cosmological speculation t h a t I have argued occurs i n later
Plato. W h i l e the model of the Republic is h a r d l y any advance on t h a t of
Philolaus, t h a t of the Timaeus, introducing the offset ecliptic, is the first
serious attempt to match the detail of w h a t can be observed to a model for
the motions of the heavens. The cosmological sophistication is greater
t h a n the astronomical accuracy, b u t t h a t is only to be expected i n a
prototype w h i c h indicates how cosmology and astronomy can be brought
together i n a fruitful manner.
I t is also possible to discern changes i n Plato's though at a more general
methodological level as well. While there is no ban on observation i n the
Republic, there is a change of emphasis on the importance of observation
by the Timaeus, and the Philebus uprates the importance of the more
practical technai. The Timaeus is also more optimistic about the results we
w i l l get i n astronomy, and quite possibly other disciplines as well, and the
Philebus allows t h a t a l l the technai can be considered to be forms of
knowledge, w i t h some being more mathematised and so more accurate.
The conception of the good i n the Philebus also seems to be more mathe­
matically and technically oriented. Plato's thought t h e n develops. Some of
t h a t development has considerable significance for the history of Greek
astronomy, cosmology and philosophy of science. Some of the criticisms
t h a t m i g h t be aimed at a middle period w o r k like the Republic are wide of
the m a r k w h e n i t comes to the later works.

V I . C r i t i c i s m s of the atomists a n d physiologoi

Plato has a series of i m p o r t a n t and cogent criticisms of the atomists and
the other physiologoi. H i s position on cosmogony at one level is quite
simple. I f we suppose some sort of preceding chaos, we cannot account for
the order of the current cosmos by necessity or chance alone, b u t m u s t
postulate some form of teleological ordering. That, given the paucity of
mechanisms for the accretion of order available to the Presocratics, seems
a perfectly reasonable position for fourth-century Greece, especially i f we
consider zoogony as well as cosmogony here. Plato's conception of cosmo­
gony and chaos bears closer analysis. His predecessors presume an e x t r i n ­
sic disorder of the stoicheia, b u t an i n t r i n s i c order. Is there any argument
to support that? Take away the intrinsic order and t h e i r account looks even
more implausible. Some assume t h a t a vortex, an orderly motion, w i l l be
formed. W h y t h a t motion, rather t h a n a n y t h i n g else? A chaos of motion
w i l l again make such an account implausible. Meno's paradox and under­
determination loom large here. W h a t reasons do the atomists and physio­
logoi have to suppose order rather t h a n chaos?
Plato's o w n cosmogony can be viewed i n two ways. I f he d i d believe t h a t
272 Plato's Philosophy of Science
the cosmos was ordered from chaos, t h e n he wants to make a general point,
t h a t whatever the preceding chaos a teleological ordering is required,
rather t h a n commit himself to any specific picture of chaos. I f we take the
cosmogony as metaphorical, as a counter-factual analysis of the cosmos
now, the emphasis appears more epistemological. I n order to account for
w h a t we see, w h a t sort of order do we need to suppose? W h a t w o u l d the
cosmos be like w i t h o u t t h a t sort of order? However we take Plato on
cosmogony (and I t h i n k we can take this i n both ways), this is an i m p o r t a n t
piece of t h i n k i n g about the origins of the cosmos w h i c h brings to l i g h t
several difficulties i n the atomist account.
We have seen some of the cosmological and philosophical objections
Plato has to the idea of explaining the order of our cosmos i n terms of
m u l t i p l e entities, be they cosmoi, shapes of the stoicheia or species. Also
i m p o r t a n t for Plato is an adequate explanation of the continuing order of
our cosmos. T h a t forms part of the drive towards laws w h i c h explain the
continuing r e g u l a r i t y of our cosmos rather t h a n considering this to be a
m a t t e r of luck or r e q u i r i n g a further m u l t i p l i c a t i o n of cosmoi.
Plato's o w n geometrical atomism, I have argued, sets out to rectify m a n y
of the problems inherent i n the Presocratics. I t postulates t w o basic shapes
for the u l t i m a t e constituents of the cosmos, rather t h a n an u n l i m i t e d
number. These stoicheic triangles are not subject to any qualitative
changes. This is critical for a coherent theory of change. I f the basic
entities, such as the elements of Empedocles are subject to change, the
there are serious problems w i t h identifying references. We also need to be
able to account for aspects of phenomenal flux and phenomena stability.
The key to geometrical atomism is postulating stability at a level far below
t h a t of h u m a n perception, and t h e n explaining flux and stability i n terms
of the bonds t h a t are formed between the stoicheia, w h i c h can be insub­
s t a n t i a l leading to a phenomenal flux or permanent leading to stability
(the heavenly bodies) or any degree between. I t is also of considerable
importance t h a t geometrical atomism provides a conception of the struc­
t u r e of the w o r l d on w h i c h language has a grip. Indeed, the letters and
syllables analogy t h a t geometrical atomism is expressed i n w o u l d suggest
an i m p o r t a n t isomorphism between language and the w o r l d , w h i c h is
significant for the sort of results we m i g h t expect from the investigation of
nature.
Plato's reaction to the atomists and physiologoi is not simply t h a t of a
theologian and a teleologist objecting to materialism. There are serious
philosophical difficulties w i t h the Presocratic accounts of cosmogony, cos­
mology and the stoicheia w h i c h Plato addresses. Certainly his account is
teleological, b u t as we have seen t h a t is at least i n part d r i v e n by his
attempt to solve some serious philosophical problems i n a realist manner.
The r e s u l t i n g accounts of these topics are at least as plausible and viable
i n the context of fourth-century Greece as those of the atomists and
physiologoi.
 10. Conclusion 273

V I I . P l a t o r e l a t i v e to the atomists a n d physiologoi

Plato's views on science and atomism are undervalued relative to those of
the atomists and physiologoi. Geometrical atomism is often left out of the
history of science, and a l l to often even omitted from histories of atomism
where the focus is on Leucippus, Democritus, Epicurus and Lucretius.
This is a considerable injustice as geometrical atomism has a great deal to
offer. I t has significant criticisms of the shortcomings of its predecessors,
and proposes interesting solutions to these problems. I f we judge this
m a t t e r on affinities to modern atomism, i t is by no means clear t h a t the
atomists w i l l have the advantage. Geometrical atomism postulates a s m a l l
number of w e l l defined u l t i m a t e particles, allows microstructure, gives
prominence to mathematics and has a structured system of bonding. I f we
look at this i n the context of fourth-century Greece, i t is probably j u s t as
good at dealing w i t h the phenomena as the atomists, and has significant
philosophical advantages. I t is at least as plausible and viable as the views
of Leucippus and Democritus. I t is also i m p o r t a n t to recognise t h a t w h e n
atomism is revived as a significant p a r t of the scientific revolution of the
seventeenth century, this is done i n a C h r i s t i a n context where the role of
the demiurge i n solving some of the problems of early atomism is t a k e n
over by a C h r i s t i a n god.
Preference for the views of the atomists is often based on affinities to
modern science. The atomists advocate a reductionist, mechanical, atomis­
tic, anti-teleological conception of nature. Whether t h e i r conception of
nature is entirely mechanical is not entirely clear cut, as they s t i l l employ
biological analogies for physical processes, and among the physiologoi
Anaxagoras uses nous and Empedocles love and strife. I f we look for
affinities to modern science w i t h Plato, t h e n most strongly of a l l there is
the use of mathematics wherever possible. W h a t is less often recognised
is t h a t the Timaeus offers a significant reductionist programme, based on
the geometrical properties of the atoms. W h i l e this is undoubtedly crude,
i t is more sophisticated t h a n the atomists. Plato's alleged anti-empirical
attitude and pessimism about the results of the investigation of nature are
often cited as significant differences w i t h modern science, b u t we have
seen t h a t these allegations are groundless.
There is also an i m p o r t a n t philosophy of science dimension here. Plato
is going to have greater affinities w i t h realist conceptions of modern
science. We have seen t h a t his alleged a n i m i s m and supernaturalism i n
relation to the heavens is actually an attempt i n the absence of key
resources to give a realist explanation of regularities. Plato has consider­
able affinities, both philosophical and scientific, to a conception of science
where real laws and a solution to underdetermination are i m p o r t a n t . I t
may u l t i m a t e l y be the case t h a t there are greater affinities between
atomism and modern science, b u t the matter is not clear cut and they are
not as great as is often supposed, and a great deal hangs on the question
274 Plato's Philosophy of Science
of whether realism is the proper way to approach modern philosophy of
science.
The other major question here is how plausible and viable i n the context
of fourth-century Greece are the views of the atomists and physiologoi
relative to Plato? O n a great number of issues, notably cosmogony, cosmo­
logy and zoogony, they are struggling. L i k e Plato, they too lack key
resources t h a t are developed later. The brute fact of the m a t t e r is t h o u g h
t h a t Plato's views are at least as plausible and viable as any Presocratic
conception of science i n the context of fourth-century Greece. Plato's
atomism, science and indeed philosophy of science stand up w e l l i n com­
parison to the views of the atomists and other physiologoi.

V I I I . Plato's p h i l o s o p h y of s c i e n c e
There is a set of standard criticisms of Plato's views on science w h i c h I hope
to have rebutted. There are weaknesses i n relation to observation and
experiment, b u t no more so t h a n w i t h the Presocratics. He d i d not substi­
t u t e mathematics for physics. Certainly he employs teleology, b u t there are
interesting epistemological, cosmological and nomological reasons for this,
often related to his struggle w i t h the foundations of scientific realism. He
does believe w o r t h w h i l e results w i l l come from the investigation of nature.
Plato's views also develop i n interesting ways, and he makes significant
contributions to astronomy, theory of matter, cosmogony and cosmology.
Some of the strategies he employs i n these areas, notably i n the debate
about universe or multiverse, are s t i l l relevant to modern debates. So too
there are some interesting affinities between Plato's proto-realism and
modern scientific realist positions. There is a series of cogent and impor­
t a n t criticisms of the atomists and physiologoi. The teleological alternative
he offers is designed to circumvent the serious philosophical problems he
finds w i t h his predecessors. Plato played a significant p a r t i n the develop­
m e n t of ancient science and philosophy of science. The conception of nature
and the investigation of nature w h i c h he offers was at least as viable and
plausible i n the context of fourth-century Greece as a n y t h i n g offered
previously, and is one t h a t has h a d considerable influence on the develop­
m e n t of science.
 Notes

Introduction
1. Though perhaps that is a personal prejudice as I do not work in this manner!
2. See TheslefT (1982) p. 203, Taran (1975).
3. See De Vogel (1955) for discussion here.
4. For the moment I avoid calling the technai 'sciences' here so as not to
prejudge later debates, though as will become evident I believe 'science' is a
perfectly reasonable translation of techne in some contexts. Rhetoric is also an
important and often discussed techne for Plato.
5. Simplicius in De Caelo 492.3, cf. 488.23.
6. Vlastos (1975) p. xiii.
7. Vlastos (1975) p. 29.
8. Vlastos (1975) p. 65.
9. Graham (1990) p. 8.
10. Zeyl (2000) xv.
11. See e.g. Vlastos (1975) p. 63.
12. Bhaskar, in Bynum, Browne and Porter (1981) p. 362.
13. Musgrave (1991) p. 244. Musgrave also makes an important distinction
between dogmatic realists who believe we can know for sure which theories are
true and critical realists who deny this. I t is important to recognise the distinction
between critical realists and instrumentalists.
14. Leplin (2000) p. 393 comments that The debate between realism and
instrumentalism is at an impasse'.
15. Hesse, in Edwards (1972) vol. 4 p. 404.
16. Duhem (1906a), (1908).
17. Duhem (1908) p. 28.
18. Duhem (1906a), (1908). Lloyd (1987) p. 312 describes the view that Duhem
sees as typical of ancient astronomers as 'a lack of concern with the physics of the
problems in favour of a preoccupation with the mathematics, the construction of
models that are purely calculating devices with nothing to do with any underlying
physical realities'.
19. Vlastos (1969) p. 106.
20. Knorr(1993)p. 399.
21. Hetherington (1993a) p. 85.
22. Musgrave (1991) p. 250.
23. Throughout I shall translate aitia as 'reason' or 'explanation' rather than
'cause'. On this see Vlastos (1969) p. 76 ff. and cf. Sedley (1998), the literature on
cause/because and the relation of aitia to to dia ti in Aristotle, e.g. Hocutt (1974),
Moravscik (1974), Sorabji (1980), Annas (1982a), and note that Plato too uses aitia
and to dia ti interchangeably, e.g. Phaedo 96a T was very interested in the kind of
wisdom known as natural science; i t seemed to me to be excellent to know the
276 Notes to pages 17-29
reasons (aitias) for each thing, why (dia ti) each should come to be, why (dia ti) i t
should cease to be, and why {dia ti) i t exists', so also Phaedo 98c, 99a, 100c and
101b. See too Frede (1980) p. 129 ff. on the use of aitia generally.

1. Cosmology, cosmogony and teleology

1. See generally Rowan-Robinson (1996), Silk (1989), Barrow and Tipler
(1986), Smoot (1993), Weinberg (1993), Gribbin (1986), (1993), Liddle (1999).
2. Cf. Timaeus 81a.
3. Aplokanon is anything woven, a liknon is specifically a winnowing-basket.
See here Cornford (1937) p. 201 for discussion and a (marvellous!) illustration.
4. Cf. Aristotle and the idea that the four elements would settle out into four
concentric spheres if i t were not for the action of the sun.
5. See Chapter 8 on whether there are small voids due to the problems of
packing the geometrical atoms together and three-dimensional atoms dissociating
into two-dimensional components.
6. Rowan-Robinson (1996) pp. 62 ff., Silk (1989) pp. 89 ff.
7. Barrow and Tipler (1986) pp. 367 ff., Davies (1982) pp. 60 ff., Gribbin and
Rees (1990).
8. One might also point out that once we have inhomogeneities modern cosmo­
logy can give a good account of the development of the cosmos in terms of gravity.
But what of the source of the homogeneities? I f the answer to that is quantum
mechanical, as seems most likely, and that answer is a probabilistic one, then there
are many possible arrangements for a cosmos which has our values of the funda­
mental constants.
9. Cf. Philebus 64e.
10. So Cornford (1937) p. 200.
11. See Zeyl (2000) xx ff., Sorabji (1983) pp. 276 ff., Vlastos (1964) p. 401 ff., cf.
McCabe(1994)p. 180.
12. Democritus Fr. 164, Sextus Empiricus Against the Mathematicians, V I I
116-18, Barnes' translation and italics, Barnes (1987).
13. See Diogenes Laertius IX, 31.
14. Diogenes Laertius IX, 45. On the nature of necessity, 'Democritus means
by i t the resistance and motion and blows of matter.'Aetius I , 26, 2.
15. Hippolytus Refutatio I , 13, 2, cf. Aristotle Physics II.4, 196a24 ff.
16. Diogenes Laertius IX, 31.
17. Anaxagoras Fr. 12, Simplicius Physics 164, 24 and 156, 13.
18. See Phaedo 98b.
19. Anaxagoras Fr. 13, Simplicius Physics 300, 31.
20. One might compare and contrast Plato's demiurge here.
21. See here Alexander (1956) pp. xvi ff., and the Leibniz-Clarke correspon­
dence generally.
22. This is a standard difficulty for all static cosmoi with attractive forces.
23. See astronomical appendix for an explanation of the problem here. I n the
eighteenth century Laplace demonstrated the solar system to be stable within
certain limits, and on being asked where was the place for god in his system, is
reputed to have replied 'Sir, I have no need of that hypothesis'.
24. See e.g. Apology 26cd, and cf. Hippolytus Refutatio I , 8, 3-10 and Diogenes
Laertius I I , 8-9.
25. Cf. the view of Kirk, Raven and Schofield (1983) p. 380.
26. Anaxagoras Fr. 4, Simplicius Physics 35, 3.
27. Anaxagoras Fr. 2, Simplicius Physics 155, 31.
 Notes to pages 29-39 277
28. Simplicius Physics 178, 25.
29. Descartes Principia Philosophia III.47: Cottingham, Stoothoff and Mur­
doch (1985) p. 257.
30. Descartes Principia Philosophia III.45: Cottingham, Stoothoff and Mur­
doch (1985) p. 256.
31. While it would be wrong to consider Newton himself to be a mechanical
philosopher, in the light of his views on religion, cosmology, and in his alchemical
writings the 'active' qualities of matter, later Newtonians certainly did hold a
mechanical view.
32. Barrow and Tipler (1986) p. 16.
33. Barrow and Tipler (1986) p. 21.
34. See here Davies (1982) pp. I l l ff., Barrow and Tipler (1986), Leslie (1989)
pp. 127 ff., Gribbin and Rees (1990).
35. Barrow and Tipler (1986) p. 11.
36. Barrow and Tipler (1986) p. 22.
37. See Davies (1992) pp. 70-1.
38. Barrow and Tipler (1986) p. 1.
39. I f extra-terrestrial life is not found, i t would be the first instance in a long
series of historic scientific debates in which the anthropocentric hypothesis had
prove even partly valid. Barrow and Tipler (1986) p. 601.
40. See here Borelli (1680) and Hankin (1985) pp. 114 ff.
41. See here Hankin (1985) pp. 115 ff., Brown (1974), Hall (1969).
42. So too in the debate concerning preformation or epigenesis - the mechani­
cal philosophers tended to favour preformation, as then the organisation of the
embryo was already given and all they had to explain was its growth. See Bodemar
(1964).
43. The infamous "Trembly's polyp'. See Hankin (1985) pp. 131 ff., Vartanian
(1957).
44. Theophrastus Hist. Plant. 3, 1,4.
45. Living creatures come from moisture, heat and earth - see Guthrie (1978)
vol. 2 p. 315.
46. Empedocles Fr. 98, Simplicius Physics 32, 6, KRS translation.
47. Empedocles Fr. 98, Simplicius Physics 300, 21, KRS translation.
48. Aetius v, 19, 5.
49. Empedocles Fr. 57, Aristotle De Caelo III.2 and Simplicius De Caelo 587, 1.
50. Empedocles Fr. 61, Aelian Nat. Anim. XVI, 29.
51. Aristotle, Physics II.8, 198b29. There may be some Aristotelian gloss here,
but the multiplication is still evident.
52. Empedocles Fr. 59, Simplicius De Caelo 587, 20.
53. Reading anotato (highest) here. Wicksteed and Cornford (1929) pp. 144-5
argue for ano e kato as a plausible conjecture (though Simplicius has anotato) to
bring this passage into line with De Generatione et Corruptione 334a ff.
54. Aristotle Physics 196a ff., cf. De Generatione et Corruptione 334a ff.
55. One might also argue that in line with later evolutionists (cf. the nine­
teenth-century debate about the age of the earth), Empedocles feels the need for
there to have been a great amount of time for the process of the formation of
species: When the human race was first generated from out of the earth, the day
was of the length that ten months are now, owing to the slowness of the sun's
advance. As time went on, it became the length of seven months.' Aetius 5, 18, 1.
56. Simplicius Physics 28, 8. Cf. Aristotle Physics III.4 203b25, I f that which
lies outside is infinite, then i t would seem so are bodies and cosmoi. For why should
something be here rather than there in a void?
278 Notes to pages 39-50
57. Dionysius ap. Eusebium Preparatio Evangelica XIV, 23, 3.
58. Cf. Aristotle De Caelo III.4, 303a3 ff.
59. Aristotle Physics 1.4, 187a23 ff., cf. Simplicius De Caelo 242, 18.
60. See Aristotle De Generatione et Corruptione 314a8 ff., cf. De Caelo III.4,
302bl0 ff.
61. See Phaedo 10ld9-el.
62. Against Leucippus and Democritus in particular, there is no evidence for
very large atoms.
63. Simplicius Physics 28, 8, see above.
64. Cf. also Philebus 64e and Theaetetus 183b.
65. See e.g. Cornford (1937) p. 220 ff.
66. See e.g. Timaeus 53a7.
67. Timaeus 32c considers there to be only four elements, though the Epinomis
allows aether to be the fifth element.
68. Although it is not an exactly parallel case, such explanations always
remind me of an old Goon Show joke: 'Eccles - what are you doing here?' 'Every­
body got to be somewhere.'
69. See Keyt (1971), and cf. Parry (1979), (1991) and Patterson (1981).
70. Aristotle at least uses taxis in relation to such theories: There are some
who make to automaton the explanation of the heavens and of all the cosmoi. For
from to automaton arose the vortex and motion which by separating and settling
brought the universe into order (taxin)' Aristotle, Physics II.4 196a24 ff.
71. See Aristotle On Democritus sap. Simplicium in De Caelo 295, 11 and
Simplicius in De Caelo 242, 21.
72. Vlastos (1969) p. 82.
73. Boyle and Newton are important thinkers here, even leaving aside Leibniz
and the idea that this is the best of all possible worlds.

2. Astronomy, observation and experiment

1. See the papers by Vlastos, Mourelatos, Turnbull and Mueller, as well as
Anton's own introduction in Anton (1981), Mourelatos (1981a), Bulmer-Thomas
(1984) and Lloyd (1991) pp. 333-51 for the most recent contributions.
2. Most influentially Heath (1913), and Neugebauer (1957). See Lloyd (1968)
p. 78 notes 1-9 for a review of the earlier literature.
3. See Republic 510d ff. on the line allegory. Following G. Fine (1990), pp.
85-115. There is varied opinion on what these real objects of astronomy might be;
see Mourelatos (1981a) p. 63 notes 2-10 for some possibilities here.
4. Cf. G. Fine (1990) note 3 p. 3, and Stokes (1992).
5. Heath (1913) note 2 p. 135, cf. e.g. Mueller (1981) p. 104 who comments that
Tlato assumes every science to be the study of some kind of real but non-sensible
objects', so too Mittlestrass (1962) p. 117, and Neugebauer (1957) p. 152; cf. Shorey
(1935) vol. 2 p. 187 note c and Lloyd (1968) p. 78 notes 1-9 for further examples of
this view.
6. H.P.D. Lee (1955), p. 311. This position might cite Timaeus 27c-29d4 in its
support.
7. See Cornford (1937) p. 29, cf. De Santillana (1961) p. 195, 'As Plato insists
towards the end of his life there can be no "true physics" ', and p. 197 (referring to
Republic 529ab) I t shows vividly how Plato's conception of science has nothing to
do with anything happening in space and time.'
8. One link with the Republic here is epanabasmois at Symposium 211c3,
 Notes to pages 50-55 279
matching the allusion of ascent; cf. Republic 511b5-6 where hypotheses are hoion
epibaseis te kai hormas.
9. Cf. Epistle VII 342d.
10. When astronomy is discussed later, its methodology is compared with
geometry; cf. Republic 511b, 529de.
11. See Vlastos (1981) pp. 1-31, p. 11 ff. here.
12. See G. Fine (1990) p. I l l ff.
13. See e.g. Republic 529c, 530a in relation to the heavens; the construction is
the best that can be achieved. Fine (1990) p. 97 ff. argues that the form of the good
is the teleological structure of the world; 'Forms are its parts and individual
sensibles instantiate it'.
14. The point about the teleological structure of the heavens and its relation
to explanation can be made independently of Fine's model of knowledge (or the
form of the good) here.
15. Compare the Theaetetus on how things are brought before the mind for
judgement.
16. See G. Fine (1990) p. 85, Stokes (1992) p. 132.
17. Thus we can first opine intelligibles and then know them. Cf. my horizon­
tal/vertical and dynamic/static discussion of the line.
18. G. Fine (1990) p. 85 draws this consequence if we attribute her TW view to
the Republic.
19. G. Fine (1990) p. 86 takes Republic 520c to contradict her construal of TW.
20. See especially Republic 510ab, where the analogue links L2 and L3. I
believe there to be two issues here.
21. See Republic 529d8-530b4, especially paralattein at 530b3.
22. One might argue that for Plato there are no just acts to be known, all
actions being both just and unjust, so this is the most we can expect from the
guardians. See Republic 479a ff., and cf. G. Fine (1990) p. 87, who is worried that
if the guardians only have opinions of sensibles like everyone else, then they are
no more fit to rule.
23. See Vlastos (1981) p. 10 ff.
24. See Vlastos (1965b) p. 68 ff. on necessary and contingent and (1969) p. 99
ff. on a priori and a posteriori.
25. See Jordan (1984) pp. 14-16, 24-5, 70-2.
26. See Mueller (1981) p. 103, pp. 110 ff.
27. I f we think back to the epistemological allegories, geometry is used as an
example of how investigation in general works.
28. See also Republic 526de where a division is made in geometry along the
intelligible/sensible distinction for the purpose of determining which part will be
useful for education, suggesting that there is a sensible part of geometry, however
small i n relation to other disciplines. I n a much later work Philebus 55e ff. is clear
on this; when we remove the theoretical/mathematical element of any techne we
are left with the empeiria, however small that may be.
29. See e.g. Republic 511b.
30. See Republic 523a ff. especially 524e, where any contradictory perception
will force the mind to consider unity, plurality, individuation and counting.
31. This is possibly implied by the order of the subjects in the curriculum. Note
the fuss Socrates makes to get this right at Republic 527d.
32. So too with mathematics at Republic 523a ff., cf. Timaeus 47ab.
33. See Crombie (1963) vol. 2 p. 191.
34. See Crombie (1963) vol. 2 p. 37.
280 Notes to pages 55-62
35. Or how we might train astronomers. This is a point anticipated by Lloyd
(1968) pp. 79-80, (1970) pp. 68-9 and (1991) pp. 333-4.
36. This passage affirms that a subject straddles the two worlds, as a contrast
is drawn between its more and less advanced sections along this distinction, but
both are parts of that discipline.
37. Cf. e.g. H.P.D. Lee (1955).
38. Both an eminent physicist and a secondary school class 'study' physics, but
one researches and the other learns.
39. Republic 529c7-d5.
40. Republic 529d7-530b2. See Republic 510b-511c for parallels with the line
and the treatment of geometry; cf. Mourelatos (1981) p. 37 ff.
41. See Republic 531c3 for clarification of the nature of these problems, where
we ascend to consider 'which numbers are concordant, which are not and in each
case why'.
42. I t picks up Republic 521c ff. which sets the educational purpose of the
whole section, as well as 526de and 529c.
43. A point originally made by Dicks (1970) p. 234 note 39; cf. Vlastos (1981) p.
3. Some other instances ofchairein ean as meaning to renounce are at Protagoras
348a, Phaedo 63e, 10led (see Brandwood (1990) for a full list). Even this stronger
phrase is not always totally dismissive.
44. Heath (1913) p. 139, cf. Shorey (1935) p. 187, Lloyd (1968) pp. 80-1, Vlastos
(1981) p. 12.
45. Cf. Timaeus 47a ff. on the benefits of eyesight and hearing.
46. Shorey (1935) translates 'Have a part in the true science of astronomy', and
Vlastos (1981) p. 3, 'By doing real astronomy', Cavagnaro (1997) p. 357 'have a
share in real astronomy'.
47. See Vlastos (1981) p. 10.
48. See Mourelatos (1981) p. 34. Mourelatos does not specifically translate
ontos astronomias metalambanontes as 'real astronomy', but instead cites toi onti
de astronomikon at Republic 530a3 and takes i t that the whole of 529a-530c deals
with 'real astronomy'.
49. See Mourelatos (1981a) p. 34 ff.
50. See Mourelatos (1981a) p. 34.
51. Cf. Shorey (1935) pp. 180-9, Dicks (1970) pp. 103-6.
52. See Vlastos (1981) p. 10.
53. See Vlastos (1981) p. 11.
54. See Vlastos (1981) p. 13 ff.
55. Vlastos (1981) p.16.
56. Cf. Hetherington (1996) pp. 276 ff.
57. Dicks (1970) p. 109.
58. See Bulmer-Thomas (1984) p. 108 ff.
59. See Bulmer-Thomas (1984) p. 109 ff.
60. i.e. the three fingers at Republic 523a ff.
61. Cf. Timaeus 39c ff. where this linkage is explicit.
62. Vlastos (1975) p. 50. Vlastos is referring specifically to Timaeus 40c4-d3.
63. Though I would agree with Lloyd (1996) that Aristotle was an 'amateur
astronomer', cf. Gregory (2000).
64. See e.g. Aristotle De Caelo 270bl3, 292a8.
65. Compare the Epinomis passage with De Caelo 292a8.
66. See Timaeus 22d ff. As much of the subsequent discourse concerns astro­
nomy, this may be highly significant. This suggestion illuminates the Timaeus
passage whether the Epinomis is genuine or not.
 Notes to pages 63-74 281
67. Shorey (1935) p. 186, cf. his (1927) pp. 172-3, and Lloyd (1968) p. 80. For
Aristotle's view see De Caelo 270bll ff.
68. So too the Politicus: see Chapter 4 for further discussion.
69. Taking a moderately realist position here.
70. See Farrington (1936) p. 127 ff.
71. See e.g. Vlastos (1981) p. 15 ff.
72. Cf. Shorey (1935) p. 184 note c and Adam (1902) on Republic 530b.
73. LSJ gives 'to change or alter a little, esp. for the worse ... to pass aside,
turn from the path, to deviate, vary'.
74. Agreeing with Vlastos (1991) p. 22 note 17 against Cornford (1941) and
Duhem (1906a) that auton must mean 'about them' rather than 'in them' which
would require the dative autois, and against Shorey (1935 vol. 2) p. 186 note a that
the final clause is not a 'trailing anacoluthon'.
75. Cf. Sorabji (1980) pp. 61-2.
76. Sorabji (1980) pp. 61-2.
77. That is something which I argue changes as Plato's cosmology develops -
see Chapters 4 and 6.
78. Attempts to see a more complex astronomy in the actions of the daughters
of necessity (specifically some account of the retrograde motions of the planets, see
Knorr (1990)) will not do as they do not discriminate in their treatment of the stars,
sun, moon and five planets while only the five planets show retrograde motion. I
discuss this matter in greater detail in later chapters.
79. See Philebus 56d ff.
80. See Philebus 57d.
81. Sedley (1997) p. 333, cf. Dampier-Whetham (1948) p. 28, 'He roundly
condemned experiment as impious or a base mechanical art', and Crombie (1963)
vol. 2 pp. 228-9.
82. Lloyd (1987) p. 140 note 133, see also Cornford (1937) pp. 276-8.
83. See e.g. Cornford (1937) p. 277 note 1.
84. See Gorgias 485c ff. and 500e ff.
85. See Phaedrus 268c ff.
86. See Philebus 55e ff.
87. One out of many too, if we can prepare pigments from starting materials
which are a mixture of colours.
88. See Aristotle Meteorologica 375al ff., and Meteorologica 1.5 generally, see
also the pseudo-Aristotelian On Colours, and Democritus as reported by Theo-
phrastus De Sensu 73 ff.
89. See Aristotle Meteorologica 374a3 ff.
90. See Lloyd (1987) p. 140 note 133.
91. Philebus 15b ff. is a good example here.
92. Or perhaps 'self-generating' or 'self-existent' or 'spontaneous' for autophue.
93. Cf. the comments on the liver at Timaeus 7lab.
94. The one exception to this might be argued to be the Hippocratics.
95. Farrington (1936) p. 127.
96. Heath (1913) note 2 p. 135, cf. Neugebauer (1957) p. 152.

3. Meno's paradox and underdetermination

1. See e.g. the articles and commentary under 'The Duhem-Quine Thesis' in
Curd and Cover (1998), pp. 255-408, and relevant articles i n Newton-Smith (2000).
2. There are various positions here. For instrumentalism, see Duhem (1914),
and Van Fraassen (1980). For conventionalism, see Poincare (1952). For pragma-
282 Notes to pages 75-79
tism see Quine (1953). For relativism, now generally leading to a sociology of
science/epistemology/knowledge, see Hesse (1980), Bloor (1976), (1981), Collins
(1981).
3. Laudan refers to this as the 'Humean Underdetermination Problem' (HUD),
which he formulates as:
HUD For any finite body of evidence, there are indefinitely many mutually
contrary theories, each of which logically entails the evidence.
Laudan (1998) p. 323. Cf. Newton-Smith (2000) pp. 532-6.
4. See here Harre (1972) pp. 43 ff., (1981) pp. 6 ff. and Sober (2000).
5. The scale of the axes can be chosen arbitrarily (e.g. squared, cubed, log, or
something more exotic) to produce data in a rough line. There is nothing which
determines this choice of function, and recalcitrant data could be brought into line
by the use of ad hoc or even discontinuous functions.
6. Again one can manipulate the axes to produce such data.
7. One might also argue that the amount of error is underdetermined.
8. One might also question how the use of error bars can be justified on purely
empirical criteria.
9. I f our only criteria are empirical, what does i t mean for an experiment not
to work?
10. Millikan's oil drop experiment is a notorious example of a scientist using
ad hoc criteria to exclude results he is not happy with.
11. See e.g. Curd and Cover (1998) pp. 383-5.
12. So Einstein comments: ' I do not by any means find the chief significance of
the general theory of relativity in the fact that it has predicted a few minute
observable facts, but rather in the simplicity of its foundation and in its logical
consistency.' See Hetherington (1996) p. 273.
13. The modern underdetermination problem, and the modern problem of
induction generally, are often held to be descendants of Hume's discussion of
causation.
14. Duhem (1954) p. 183, cf. p. 187.
15. Quine (1969) p. 79.
16. Gillies (1998) pp. 302 ff.
17. See Duhem (1954) p. 214 ff. Duhem considered there to be a scientific 'good
sense' which guided our choice of theories (but did not of course determine them),
without ever saying a great deal about the nature of this 'good sense'.
18. Quine (1953) p. 42.
19. Quine (1953) p. 43. Laudan refers to this as the 'Quinean Underdetermi­
nation Problem' (QUD) which he formulates as:
QUD Any theory can be reconciled with any recalcitrant evidence by making
suitable adjustments in our other assumptions about nature.
Laudan (1998) p. 328.
20. See Laudan (1998) p. 325.
21. Quine (1953) p. 44.
22. Quine (1953) p. 43.
23. Indeed he does so specifically in the context of the statement quoted above,
see Quine (1951) p. 43, and (1986) pp. 618/19.
24. Quine (1953) p. 42.
25. Quine (1953) p. 45.
26. Quine (1986) pp. 618/19.
27. Quine (1953) p. 45.
28. Quine (1953) p. 46.
29. Quine (1953) p. 46.
 Notes to pages 79-88 283
30. Assuming that dropping the law of non-contradiction is not one of the
'drastic adjustments' to the system! I f i t is, anything goes.
31. I take one of the general points of the discussion of Protagoras' theory to
be that one must assert something to be true in order to assert at all.
32. Quine needs either to drop these three underdetermination principles or
explain why their scope is restricted (and proximity to the edge of the fabric, as
Quine might put it in Two Dogmas would be rather too vague). Similar problems
might be argued to apply to the problem about negative instances.
33. So too with negation and predication, as with the Sophist.
34. Often citing Quine, see e.g. Hesse (1980) p. 33.
35. Hesse (1980) p. 33.
36. i.e. as there is some matter in zero volume the density is infinite.
37. See Meno 82b ff., so too Phaedo 72e-77a, 91e-92d.
38. Quine (1960) p. 23.
39. See e.g. Barrow (1991), and cf. Hawking (1988).
40. Burnet (1911) p. 99 comments This is the oldest name for what we call
"natural science" '; cf. Gallop (1975) p. 234 note 56 and Hackforth (1955) p. 122.
41. Three possibilities canvassed for the protos pious on a non-ironical
interpretation have been Anaxagoras' teleology, teleology of any sort, and
divinely revealed knowledge; cf. the discussions of Crombie (1963), Gallop
(1975), Goodrich (1903) and (1904), Williamson (1904), Burnet (1911), Murphy
(1936), Hackforth (1955), Huby (1959), Rose (1966), Sayre (1969), Shipton
(1979), Bostock (1986).
42. Whether this passage has anything to do with the historical Socrates is
another matter; see Hackforth (1955) p. 127 ff. for discussion.
43. Cf. G. Fine (1987) p. 88 f f , MacKenzie (1988) p. 41 ff.
44. Throughout I shall translate nous as 'intelligence', though i t might equally
well be translated as 'mind'.
45. Cf. Aristotle Metaphysics 984b 15-17.
46. See Guthrie (1978) p. 317, cf. McCabe (1993) p. 54.
47. So Menander, fragment 241; see Hackforth (1955) p. 137.
48. Cf. MacKenzie (1986b) p. 17 ff, Vlastos (1991) ch. 1 on riddling and
complex irony.
49. Reading kardopio (lid of a kneading trough) as against kardopo (kneading
trough). Cf. ad loc. notes in Burnet (1911) and Gallop (1975). Kardopio is more in
tune with Aristotle's report at De Caelo 294bl5.
50. Possibly the theories of Empedocles and Anaxagoras. See Sedley (1989) p.
363.
51. Sedley (1990) p. 370. Cf. Lennox (1985) pp. 199 f f , G. Fine (1990).
52. See e.g. G. Fine (1990).
53. See Wiggins (1986) p. 3 and Vlastos (1969) p. 82 note 15.
54. Sedley (1989) p. 370.
55. G. Fine (1990) p. 386.
56. Cf. the discussions of R. Robinson (1953), Sayre (1969), Bostock (1986).
57. See Gentzler (1991) p. 266.
58. See Robinson (1950) p. 128.
59. See e.g. Robinson (1950) pp. 129-31, Bostock (1986) p. 169.
60. R. Robinson (1953) p. 127, cf. Sayre (1969) pp. 15-28.
61. Cf. Gorgias 479c5,496e5,508b3, Parmenides 136a8,136b3,142b3, Sophist
251e5, Philebus 22a7. I n what follows I might be accused of assimilating the
Republic and Phaedo too closely, but note that sumphonein and homolegein rather
284 Notes to pages 88-96
than sumbainein are used in the Republic especially in discussion of the line
allegory.
62. Gentzler (1991) p. 268.
63. Cf. Gentzler (1991) p. 269 ff.
64. See Gentzler (1991) p. 270.
65. Gentzler (1991) p. 271.
66. Cf. Philebus 56a, Sophist 254a, Phaedrus 260e and 270b, Gorgias 462c ff.,
Epinomis 976b.
67. This would rule out many of the propositions that Gentzler is concerned
about. Of course, there may be some accounts which are consistent, seemingly
suitably teleological and have not yet been shown to stand in a suitable relation.
Here Gentzler may be correct in suggesting that sumphonein and diaphonein are
contraries and not contradictories, such that we only reject as false those accounts
which do not suffice teleologically or do not stand in a suitable relation, reserving
judgement on others.
68. Vlastos (1975) pp. 30-1. I t is possible that Anaxagoras would not give an
unequivocal 'No' to this.
69. Cf. Timaeus 55cd, Philebus 17e.
70. Hawking (1993) ix says: ' I have written in the belief that the universe is
governed by an order that we can perceive partially now, and that we may
understand fully in the not-too-distant future.'
71. See Pais (1982) p. 6. Nor, according to Einstein, does god play dice.
72. Aristoxenus, Elements of Harmony I I , 30-1, Gaiser's translation, cf. Sim­
plicius Physics p. 151, p. 453, Themistius Oratio 21, 245cd, Proclus, Philebus
Commentary p. 688.
73. See Sayre (1983).
74. Gaiser (1980) p. 12.
75. See Cooper (1977) p. 155.
76. See Lennox (1985) p. 206.
77. M.R. Wright (1995) p. 47.
78. See Duhem (1969) p. 5. Several modern commentators have followed
Duhem here, most recently Cushing (1998) pp. 54-5.
79. See Lloyd (1978) and Musgrave (1991).
80. Duhem (1969) pp. 5-6. Duhem (1906) p. 103 asks us to 'Pause a moment at
this fundamental passage, which so clearly shows us the doctrines of Plato touch­
ing the method of astronomy.'
81. Duhem (1969) p. 7.
82. Duhem (1969) p. 28.
83. Duhem (1954) p. 299.
84. See Duhem (1954) p. 299.
85. See Republic 519c: 'Our job as lawgivers is to compel the best minds to
attain what we have called the highest form of knowledge, and to ascend to the
vision of the good as we have described' and 520e: the guardians will descend
because We are making a just demand of just men'.
86. See Republic 529c ff.
87. See e.g. Duhem (1906) pp. 105-11 and cf. (1969) pp. 5-24.
88. See e.g. Republic 510b3, b5, c4, c6, 511a4, a6, b5, b6, c7, d2.
89. See Republic 511b.
90. Duhem (1906) p. 103.
91. See Timaeus 40a.
92. See Laws 897b ff.
 Notes to pages 96-102 285
93. Duhem (1969) p. 7. Against this he contrasts the methods of Eudoxus and
Callipus, who 'Followed the method of the astronomer, as laid down by Plato'.
94. I n one context explanation A is most meritworthy, in another context or
from another perspective explanation B may have the greater merit, and there is
no independent perspective from which to judge which is in fact the better
explanation to adopt. See van Fraassen (1980).
95. See Leplin (2000) pp. 398-9.
96. Fine is influenced here by Hilbert's comments on set theory - a proof of the
consistency of a set theory ought not to come from that theory itself, but be derived
from other, more stringent considerations. See A. Fine (1984), Curd and Cover
(1998), Newton-Smith (2000).
97. Fine's reason for doubting decision making procedures is the historical
argument so Plato may well not share Fine's doubts.
98. Hilbert's criteria are often considered to be too stringent, and we might
well say the same of Fine. I f Fine's argument is generalised to 'we always need
higher level justification' then it is regressive and may lead to the conclusion that
we have no knowledge at all. Fine's ultimate position (the Natural Ontological
Attitude) is arguably a realist one (see Musgrave (1989)).
99. Simplicius in De Caelo 488, 18 f f , cf. 492, 31 ff.
100. See Mittlestrass (1962) p. 154 note 214, von Fritz (1971) p. 179 note 375,
Knorr (1990) pp. 319 ff.
101. Zhmud (1999) p. 220, cf. Vlastos (1975) Appendix L p. I l l on whether we
can put so much faith in Simplicius' 'fastidiousness'.
102. A t Simplicius in De Caelo 493, 1 it is engkuklion kai homales kai tetag-
menon.
103. See Vlastos (1975) Appendix L pp. 110-11.
104. See Simplicius Physics p. 60 Diels.
105. See Zhmud (1999) p. 218 note 23.
106. Cf. Zhmud (1999) p. 218 note 23.
107. Especially so i f we look to Simplicius in De Caelo 492, 31 ff. where a
different order of presentation means that this problem does not occur.
108. Geminus, Isagoge I , 19-21, Heath's (1913) p. 269 translation.
109. Van der Waerden (1978).
110. See Zhmud (1999) p. 219 note 25, and Zhmud (1997).
111. Zhmud (1998) p. 211.
112. Zhmud (1999) p. 219 ff.
113. Neugebauer (1957) p. 152. Cf. Adam (1902) vol. 2 p. 125, Shorey (1935)
vol. 2 p. 187 note c, Samburksy (1956) p. 45, Mourelatos (1981a) notes on pp. 63-4,
and Bulmer-Thomas (1984) p. 107 for further references to assessments of Plato's
astronomy.
114. I also discuss the possible relationship and direction of influence between
Plato and Eudoxus in greater detail.

4. Celestial motion in the Timaeus

1. See most recently Cavagnaro (1997), with whom I am in broad agreement
on the relation of the Republic, Timaeus and Laws on this matter while differing
on many details.
2. See Phaedo 109c ff.
3. This passage supports the idea that Plato holds an imperfection view of the
physical world or that he has not yet fully worked out his views on participation.
It may be that the visual analogue of what a fish might see is meant to apply in
286 Notes to pages 102-110
intellectual terms for humans, but that too could be used to support the sort of
conclusions the Republic draws.
4. Note the emphatic ton nuni instead of ton nun at Politicus 272b2.
5. Paralattein does occur at Timaeus 22d in Critias' account, which in turn is
an older myth that has come down to him (Timaeus 21b). As Lloyd (1991) p. 342
has pointed out, there are some oddities about the nature and status of Critias'
account; perhaps all we need here though is to note that Critias is entirely willing
to defer to Timaeus i n matters of astronomy, i.e. 'Timaeus is the best astronomer
among us, and has made a special study of the nature of the universe' (Timaeus
27a).
6. Cf. Popper (1945) vol. 1 p. 219 note 5.
7. Lucretius 2.1172-4, M.R. Wright (1995) translation, p. 91, cf. Politicus myth!
8. See Brisson (1995) pp. 349-63, Rowe (1995b) pp. 186-99, and cf. Erler (1995)
pp. 375-80, Ferrari (1995) pp. 389-98, especially p. 394 note 17.
9. See Politicus 269d5 ff., cf. McCabe (1997) pp. 100 and 103.
10. See McCabe (1997), pp. 104 ff., and see Politicus 272bc and 274e5 ff.
11. See Lane (1995).
12. Burkert (1972) p. 324, cf. Heath (1913) p. 171, Mittlestrass (1962) pp. 130
ff., Owen (1953) pp. 325 ff.
13. e.g. Timaeus 38c6, 39dl, 40b6, and 43b3. The cognate verb planao has
associations of to mislead, or to lead into error.
14. The six motions are up, down, left, right, backwards and forwards. Accord­
ing to the latest evidence, the universe does not revolve. I f i t did, we could expect
to see systematic anisotropics related to the axis of rotation in COBE maps, but
we do not.
15. We are also told that sight is the greatest gift from the gods to men.
16. Aristotle comments at Physics 218b that 'Some have identified time with
the revolution of the whole, and others with the sphere itself' and is thought to be
referring to Plato and the Pythagoreans respectively (see Simplicius' references to
Eudemus, Theophrastus and Alexander, at in De Caelo 700, 18). His general
criticism of Plato at Physics 219b ff. is that time, while intimately related to
motion, should not be identified with motion. Whether Aristotle is quite correct
here is another matter. I would agree with Cherniss (1944) p. 418 and note 349,
and Vlastos (1964) p. 409 ff., that the key phrase is that the stars 'mark off and
preserve the numbers of time' (Timaeus 38c6), and so implicitly distinguish clock
and time as measure and measured.
17. Adam (1902 vol. 2) pp. 297/8.
18. e.g. the uses of'calculate' at Timaeus 47c2 and 40dl.
19. Cf. Timaeus 37d, 38a9, 38c, and note the addition of arithmos at 39c8 in
comparison to Republic 530a8.
20. Cf. other uses of planes in the Timaeus, at 19e4 to describe those who
'wander' from city to city, at 43b4 to similarly describe our mental revolutions, and
in some biological contexts describing the random wanderings going on in our
bodies, of phlegms at 86e7, of atomic particles at 88e2, and of the womb in women
at 91c4.
21. There might be a worry here about reading the views of the Laws back into
the Timaeus, but the Timaeus advocates similar, i f not quite so well expressed or
developed ideas; see e.g. Timaeus 34a, 89a.
22. See Timaeus 34a.
23. Note that the motion of the whole is regular but not perfect, too; it fails to
be self-generating. See Timaeus 34a.
24. While we would regard this as mathematically incorrect, as we would
 Notes to pages 110-114 287
resolve the circular motion of the stars on the same into two normal linear motions,
the sense is clear enough; the fixed stars have only the translational motion
imparted by the same.
25. Laws 822a might also be taken as a clarification that the 'contrary power'
has either been rejected by this stage or was never a problem for regular celestial
motion.
26. See too e.g. Laws 897b.
27. See Timaeus 40a ff. and 40b ff.
28. See e.g. Philebus 32d.
29. A t Philebus 43a ff. the idea that nothing is free from the processes of
deterioration and restoration is raised in language that is reminiscent of the
discussions of Heraclitean flux at Theaetetus 179-83 and Cratylus 439-40. Socrates
avoids discussion of the truth of this theory by suggesting that there are some
changes that are so small that we do not perceive them, and hence we sometimes
feel that we are in a state that does not involve deterioration or restoration.
30. See e.g. Epinomis 982b-d, 983c, 987a, 991c.
31. See e.g. Timaeus 34a on motion and cf. Politicus 269d-270b, and Timaeus
32c, 38b, and 41a on indissolubility.
32. See Politicus 273b ff.
33. Cf. Politicus 270a9 and Timaeus 33d9, a point made by Brumbaugh (1961)
p. 523, although his concern is with models rather than uniform motion. Timaeus
34a is a better passage for my purposes. I t is worth noting that prior to the Timaeus
Plato always has the cosmos supported; a pivot in the Politicus, the spindle of
necessity in the myth of Er (Republic 616b).
34. Cf. Timaeus 33a9 and Politicus 273e4.
35. The myth is introduced to illustrate a point in the hunt for the definition
of the statesman, see Politicus 268de.
36. See e.g. Phaedo 109a f f , where the 'fish's eye view' is indeed indicative of
Plato's current epistemology.
37. See e.g. Laws 897b, 967a etc.
38. One might make the case that the Politicus is transitional, rather than the
last use of the old cosmology. Perhaps the two cycles of the myth reflect Plato's first
dissatisfaction with degeneration, and at Politicus 284a ff. there is the concern
with mensuration that we find in both Timaeus and Philebus, and some of the
Pythagorean ontological themes which with Sayre I shall argue in later chapters
are typical of the Timaeus and Philebus are to be found in the Politicus too.
39. See McCabe (1997) p. 99.
40. Skemp (1952) p. 147 uses 'thousands of revolutions', cf. Rowe (1995b) p. 65
'many tens of thousands of revolutions'.
41. Cf. Timaeus 39c.
42. Emphronos can mean sensible both in the sense of alive and i n the sense
of reasonable or prudent.
43. e.g. Timaeus 46e.
44. See Archer-Hind (1888), A.E. Taylor (1928), Cornford (1937) and Morrow
(1950) for further discussion here.
45. Cf. Long (1977).
46. There are possibly some conservative assumptions concerning art in play
here; 'good' in the sense aesthetically pleasing, perhaps, or in the sense of being an
accurate representation of the subject.
47. Cf. Aristotle on good and chance: 'Something good might come about by
chance; but in respect of chance, and insofar as it results from chance, i t is not
good.' Aristotle Protrepticus Fr. 11, translation of Lennox (1985) p. 198.
288 Notes to pages 114-120
48. One must be careful of anachronism here, as Plato never expresses his
views in terms of laws.
49. See e.g. Timaeus 30a5 et passim, and cf. Aristotle Metaphysics 10 on taxis
and armies.
50. See e.g. Timaeus 42d ff., where the demiurge is said to be blameless for
anything that may happen after the initial ordering; cf. Republic 380c ff.
51. i.e. as in my example, if we throw a bucket of paint at a canvas.
52. See Timaeus 61d ff. on the shape and properties of fire.
53. See Cornford (1937) p. 165.
54. Sorabji (1980) pp. 61-2, cf. pp. 206-8
55. Sorabji (1980) p. 61, cf. pp. 207-8.
56. Sorabji (1980) pp. 61-2.
57. See Vlastos (1969), cf. Sedley (1998).
58. See Timaeus 67c ff.
59. See Timaeus 65c ff. on taste.
60. See e.g. Timaeus 56d ff.
61. Is the sorting of like to like an effect of the motion of particles (as with
Descartes on gravity) or a cause in its own right (as with Newton on gravity)? I
would argue that Plato is closer to the former and so would see like to like more as
an effect than cause.
62. So A.E. Taylor (1939) p. 25 ff. The passage is open to several interpreta­
tions, though note that the standard middle period view that 'All things that come
to be must also decay (genomenoi panti phthora estin)' (Republic 546a) is being
pushed here and would be a suitable driving force for political and social corrup­
tion; cf. Adam (1902) vol. 2 p. 202, who asks What was the originating cause of
degeneration? Plato finds the cause, not in anything peculiar to the Ideal city, but
in a law which prevails through the whole of Nature - the law that everything
created is doomed to decay'.
63. Taylor (1939) p. 25.
64. Popper though would disagree. He comments that 'Plato summed up his
social experience, exactly as his historicist predecessor had done, by proffering a
law of historical development. According to this law ... all social change is corrup­
tion or decay or degeneration' (1945 vol. 1) p. 19, cf. p. 37. This seems far too strong
though. Both Plato's writing (e.g. Republic 473b, 502b ff., Laws 676b ff., 769d7-8)
and his practice, if we are to believe Epistle VII (and i t may be the case that only
the digression is a forgery), show considerable optimism that actual societies can
(at least temporarily) be improved rather than merely stopped from decaying.
65. As Plato once more invokes the (paidia Ispoude) contrast at Laws 688bc,
we must beware of his 'serious play' as with Republic 545d ff.
66. The Athenian Stranger hopes to root out ignorance and vices by an ade­
quate education system (see books I , I I and VII), and their are various strictures
aimed against financial acquisitiveness. No citizen may possess gold or silver
(Laws 74le ff.), one family may possess only one landholding, which is non-trans­
ferable (741b ff.), and there are strict limits on personal wealth.
67. I t is perhaps worth noting that in the Republic the decline begins with
incorrect breeding (Republic 546a ff.). The Laws too discusses this (775c ff.) but
does not connect poor breeding with any inevitable decline.
68. See Timaeus 4 I d ff. These matters will be discussed in more detail i n the
next chapter.
69. The language of mastery/being mastered here is precisely that of the
discussion of the relation of intellect, the revolutions and perception. Cf. Timaeus
44a where we find a similar use ofkratein.
 Notes to pages 121-131 289
70. For an excellent to reply to Owen (1953) on the relative politics of the
Republic, Politicus and the Timaeus-Critias which in his view indicate that the
Timaeus is earlier, see C. Gill (1979).
71. One might note that the receptacle is entirely characterless in order not to
distort what is in i t (Timaeus 50d ff.), and that what is in the receptacle is marked
out by forms and numbers (Timaeus 53b5).
72. See Hetherington (1996) pp. 275 ff. and p. 288.
73. Hetherington (1996) p. 275.
74. Some modern scientific realists, and indeed many seventeenth-century
scientific realists, are realists about god, and might well say that god is most real,
while still being scientific realists.
75. See e.g. Timaeus 40c and 47c2 on calculation and astronomy.

5. Plato and the development of Greek astronomy

1. See Republic 616b-617d.
2. See here Burkert (1972) pp. 330-1. See Anaxagoras Fr. 78, for Hippocrates,
Aristotle Meteorologica 343a5 ff. and for Democritus, Aetius I I , 16, 1.
3. Cf. Republic 617a, Timaeus 38e, Laws 822a, Epinomis 987b.
4. There is special relationship between sun, Venus and Mercury not shared
by the other planets. See below Chapter 5.IV.
5. Cf. Knorr (1990) and van der Waerden (1951), (1974a), (1982).
6. On Philolaus and the Pythagoreans, see Aristotle De Caelo 293a 18 ff., Aetius
I I , 7, 7 ff. See also Burkert (1972) p. 299 f f . Heath (1913) p. 99 ff. and Dicks (1970).
The sirens on top of each circle making a note of constant pitch (Republic 617b)
which together make a harmony would seem to be a Pythagorean influence.
7. Diagram following Adam (1902) and H.P.D. Lee (1955).
8. I agree with Dicks (1970) pp. 70-1 that the only evidence we have for an
offset ecliptic in Philolaus is Aetius (DK 44 A21), and his evidence is suspect and
should be rejected. The one suspect fragment appears to have sun, moon and earth
all moving i n an inclined orbit, and need special pleading to have the orbit of the
earth offset to that of sun and moon. Against this one suspect fragment is the fact
that longer accounts in Aetius, Aristotle and Simplicius give no indication of any
offset orbits. Cf. Dreyer (1906) and Heath (1913).
9. See below Chapter 5.Ill on the retrograde motion of the planets.
10. See Vlastos (1975) Appendix H p. 106 on the etymology. But Heath (1913)
p. 149 translates as 'third in speed of counter revolution the fourth appears to
move'.
11. Burnet (1924) p. 304 note 1.
12. Dicks (1970) p. 112.
13. See Dicks (1970) pp. 112-13.
14. Knorr (1990) p. 326 note 11.
15. See Timaeus 36c and Epinomis 987b, cf. Laws 760de on right and left. Plato
may not be entirely consistent in his usage.
16. See Timaeus 36b ff. on how the demiurge creates same and different and
produces the angle between them.
17. The Greeks, probably drawing on Babylonian data, reckoned the maxi­
mum deviation in latitude to be that of the moon, 6° either side of the ecliptic. See
Neugebauer (1975) p. 626.
18. See Burkert (1972) p. 306, Aetius I I I , 14, 1.
19. I reject Burnet and Taylor's theory of a moving earth in the Timaeus. See
Gregory (2000) and Chapter 6.
290 Notes to pages 133-138
20. Deviation in latitude is exaggerated here for effect.
21. Modern explanation of retrogression for superior and inferior planets. I f
one follows the projection of the apparent position of the planets relative to the
stars from earth, the planet will be seen to progress, stop, retrogress, stop and
progress again due to the relative motions of the earth and the planet, moving in
different orbits of the sun at different speeds.

22. Inclinations to the ecliptic are earth 0°00' (by definition), Mercury 7°00',
Venus 3°24', Mars 1°51', Jupiter 1°18', Saturn 2°29'.
23. See Neugebauer (1975) p. 626.
24. There is debate as to which of the second and third spheres generates
motion along the ecliptic and which generates deviation. See Dicks (1970) pp.
179-81.
25. I t also of course generates some longitudinal motion.
26. We have only a partial account of the actual values Eudoxus used for the
construction of his system, so reconstructions have to be conjectural.
27. See above on the myth of Er.
28. They are i n fact on orbit around the sun.
29. Certainly he refers to the relation of Venus and Mercury to the sun at 38d-6
(and quite possibly he does so at Timaeus 40c8-9 as well), almost certainly Timaeus
40c5 refers to retrogression, and Republic 617b2 may well do so (see Heath (1913),
Dicks (1970) and Knorr (1990) for debate here.
30. Vlastos (1975) Appendix J p. 108.
31. To give a full account of the orbits of Mercury and Venus one has of course
to talk of retrogression, but that on its own will not explain why Mercury and
Venus are always seen near the sun.
32. There are two interpretations of the contrary power which I do not discuss
here as there are overwhelming objections to them. Martin argues that Mercury
 Notes to pages 138-149 291
and Venus orbit at the same speed as the sun but in the opposite (contrary)
direction. He tries to support this with similar uses of enantia at Timaeus 36d and
Republic 617a7. However, the fact that Mercury and Venus do not orbit in a
contrary direction from the sun is evident, and the contrary power is supposed to
explain why Mercury, Venus and the sun overtake and are overtaken by each other,
and this interpretation of the contrary power does no such thing. Alternatively,
Mercury and Venus may orbit with the same direction and speed as the sun, but
the contrary power may occasionally act in opposition to this movement. Depend­
ing on the strength of the contrary power, this might merely slow Mercury and
Venus, or it might cause them to stop or even reverse their direction. However,
such an interpretation could only have Mercury and Venus overtaken by the sun
and not overtaking it, and they would appear at all angles relative to the sun.
33. See Knorr (1990).
34. Knorr (1990) p. 316.
35. Cornford (1937) p. 137.
36. See Empedocles Fr. 42.
37. The only modern discussion of the eclipse problem is in Taylor (1928) pp.
236-7.
38. There is debate as to which of the second and third spheres generates
motion along the ecliptic and which generates deviation. See Dicks (1970) pp.
179-81.
39. I t also of course generates some longitudinal motion.
40. That is our best conjecture on why Callippus added these spheres. See
Neugebauer (1975) pp. 625-8, and cf. Mendell (1998) and Yavetz (1998).
41. The model of the Timaeus, with the moon and then the sun as the nearest
celestial bodies, will not generate the transit of Mercury and Venus across the face
of the sun, but such a phenomenon was unknown to the ancients.
42. See Cornford (1937) p. 135 on Proclus, Taylor (1928) p. 243.
43. See Dreyer (1906) p. 68 on this phenomenon.
44. The Babylonians were aware of the motion of the moon in latitude, plotting
it against stars falling within the zodiac, see van der Waerden (1974b) pp. 98-9.
45. I t should be noted that this is not a term that is later used for retrogression
in the astronomical tradition.
46. See Simplicius in De Caelo 497, 18-22.
47. Neugebauer (1975) pp. 627-8.
48. However there is other evidence for 92, 89, 89, 95. See Neugebauer (1975)
p. 627 note 9. The modern values are (days.hours) 93.15, 89.19, 92.19, 89.1 (Ricker
(1967) p. 87).
49. See Simplicius in De Caelo 497, 18-22, cf. Simplicius' report of Eudemus
497, 17-22.
50. See Schiaperelli (1875), Neugebauer (1975) p. 625.
51. Simplicius in De Caelo 504, 17-20.
52. There is a problem with the text here, but the general meaning seems clear
enough. See Taran (1975) ad loc. and Lamb (1927) ad loc.
53. The Loeb text omits these words, without comment, and for no good reason.
They are in Burnet's text and are omitted in only one of the sources.
54. See Van der Waerden (1951), (1974a) and (1982), and see Gregory (2000).
55. Zeyl (2000) xliv-xlv.
56. The point of dividing the different is most likely to allow the seven planets
to have different speeds, cf. Zeyl (2000) vlx note 89.
57. See e.g. Schiaperelli (1875), Heath (1913) pp. 210 ff., Dicks (1970), Mendell
(1998), Yavetz (1998).
292 Notes to pages 150-160
58. Taylor (1928) p. 202.
59. Cornford (1937) p. 107 note 2.
60. There is also a problem of the degree of sophistication we allow Plato with
van der Waerden's theory. I f we are to attribute epicycles to Plato, how far do we
go with this? On what grounds do we accept or reject epicycles for Mercury and
Venus, for retrograde motion generally, for deviation from the plane of the differ­
ent, or for variations in planetary distance? Such considerations also apply if we
attribute hippopedes to Plato.
61. See Taylor (1928) pp. 196-202.
62. Assuming we reject the attribution of either epicycles and hippopedes to
Plato.
63. Philolaus probably accepted regular circular motion (see Geminus Ele-
menta Astronoma 10.2-21), but did not have offset orbits (see above, section I).
There is the matter of the central fire as well.
64. Vlastos (1975) p. 59.
65. See Gregory (2000).
66. See Neugebauer (1975) pp. 588 ff.
67. Burkert (1972) p. 324 considers Eudoxus to have influenced Plato, convert­
ing him to regular circular motion after the Timaeus, and Cavagnaro (1997)
believes such an influence to have occurred before the Timaeus.
68. See Owen (1953) p. 325
69. See Skemp (1942) pp. 142-3.
70. Owen (1953), p. 325.
71. Apollodorus' practice of equating the high point of someone's achievement
with the age of forty is of course highly dubious. Here he may have associated
Eudoxus' acme with his temporary presidency of the academy in 367/366.
72. Gisinger (1921) pp. 5-6.
73. De Santillana (1949) pp. 248-62.
74. Lasserre (1966) pp. 137-42.
75. See Pliny, Natural History XXX, 3.
76. Owen (1953) p. 325.
77. For some further conjectures on Eudoxus and Plato, see Dicks (1970) p. 143
note 210.
78. See Owen (1953) pp. 325-6.
79. See Owen (1953) p. 326.
80. Dreyer (1906) p. 91.
81. Heath (1913) p. 196.
82. Wright (1974) p. 73.
83. There is no need for a mechanism here, i f this is an instrumentalist
solution.
84. Wright (1974) p. 167.
85. Wright (1974) p. 168.
86. Epinomis 99Id is a good example here.
87. Cf. the comments on geometrical atomism at Timaeus 54ab.

6. Plato and the development of Greek cosmology

1. For debate here see Neugebauer (1957), (1975), Toulmin and Goodfield
(1961), Vlastos (1975), Oppenheim (1978), Van der Waerden (1978), Lloyd (1979),
(1987), Olsen (1982), Pingree (1992), Von Staden (1992).
2. Pedersen (1974) p. 26.
3. Burkert (1972) p. 335.
 Notes to pages 160-164 293
4. Aristotle De Caelo 294a28 ff. Cf. Metaphysics 1.3.
5. See Popper (1963), cf. Kirk (1960), (1961) and Lloyd (1991) ch. 5.
6. Aristotle De Caelo 294bl3 ff. While Furley (1987) p. 20 has called this the
'frisbee' theory of the stability of the earth, there are some important dissimilari­
ties. A frisbee flies, but will not stay aloft when i t is stationary. Moreover, when i t
flies the lift i t requires is generated by the leading edge being slightly higher than
the trailing edge, such that i t is not parallel to the ground.
7. See Furley (1987) pp. 53 ff.
8. See Furley (1987) p. 24.
9. Though with a central earth and a like-to-like principle, Plato is practically
centri-focal
10. See Burkert (1972) pp. 299 ff.
11. Aristotle De Caelo 295bl0 ff.
12. Simplicius in De Caelo 532, 13, Furley's translation and emphasis.
13. See Panchenko (1994), Schofield (1997), and Kahn (1999).
14. Hippolytus, Refutatio I , 6, 3, who says that for Anaximander 'the earth is
aloft, not held in the power of anything but remaining where it is because of its
like distance from all things'.
15. That the earth is a sphere would seem to be he likeliest reading of
peripheres here, though Morrison (1959) raises doubts; cf. Calder (1958), Rosen-
mayer (1959), and Guthrie vol. 4 p. 336 note 1.
16. Plato uses piptein in this sense in other contexts, e.g. Laches 181b and
Philebus 22e.
17. See Adam (1902) vol. 2 pp. 443-7 on the nature of ta hupozomata, and cf.
Laws 945c.
18. See Morrison and Williams (1968) and Morrison and Coates (1986).
19. See e.g. Timaeus 36e.
20. For the main contributors to the debate see Heath (1913) pp. 174 ff.,
Cornford (1937) pp. 120 ff., Cherniss (1944) pp. 540 ff., Dicks (1970) pp. 132 ff.,
Burkert (1972) p. 326 note 16, Vlastos (1975) pp. 31 ff.
21. I side with Cornford (1937) p. 120 note 1 in removing the ten after
illomenen de in Burnet's text.
22. See Burdach (1922) on the il- and eil- word groups (rolled around vs.
packed around) which become confused in meaning and spelling. I do not see that
either reading can support the up-and-down motion of Burnet and Taylor.
23. See Dicks (1970) p. 136.
24. See Burnet (1924) pp. 302-3 and Taylor (1928) pp. 226-40.
25. Taylor (1928) pp. 236-7.
26. See Timaeus 40c2-3.
27. All that is required here in fact is relative motion between the earth and
the same. The simplest and most reasonable solution is that the same rotates once
a day and the earth does not move at all, but the same could rotate twice and the
earth once, etc. Certainly it is reasonable to assume that day is one revolution of
the same, but note that month for Plato is not one revolution of the moon relative
to itself, but relative to the sun, so perhaps day is one revolution of the same
relative to the earth.
28. Aristotle De Caelo 293b30.
29. Plutarch, Quaestiones Platonicae V I I I , 1, 1006c, Numa c. I I .
30. See e.g. Heath (1913) pp. 183 ff. One would expect Aristotle to know of this
and record it. The idea that the sun was the worthiest body and should be at the
centre was also common in the Renaissance and the early part of the scientific
294 Notes to pages 164-169
revolution. Copernicus, Kepler and Harvey all gave the sun a central position on
these grounds.
31. See Aristotle De Caelo 293al5 ff.
32. That is certainly the case prior to Galileo and Newton, when there were not
the physics to deal with a moving earth.
33. See Cleanthes reported by Plutarch, De Facie in Orbe Lunae 922f-923a.
34. As indeed was the case up to the seventeenth century. I n general, heliocen-
trism was not considered a serious possibility.
35. I t is worth remembering that for both Plato and Aristotle the earth would
not have been moving through a vacuum. The Greeks were also quite capable of
realising that the earth would be moving at speeds unheard of for them.
36. The Copernican system was generally regarded as a useful computational
tool but not a serious cosmological possibility. I t took the work of Galileo both on
the physics of motion and with the observational evidence of the telescope, along
with the critical simplifying work of Kepler on ellipses to make the Copernican
hypothesis a real challenge to geocentrism.
37. This despite the fact that the Greeks had very good observational astro­
nomy. If they could discover the inequality of the seasons and the precession of the
equinoxes, then being unable to detect parallax (especially in the relatively small
Greek cosmos) was a serious matter.
38. This was also a problem for Galileo, as the telescopic evidence cut both
ways. While there was evidence to support Copernicus, the failure to detect
parallax meant that the cosmos had to be many times the previously supposed size.
39. Indeed the first step in establishing a distance ladder for beyond the solar
system is being able to detect parallax for relatively near stars.
40. Even then a great deal of the telescopic evidence goes against Aristotle but
does not specifically support Copernicus. All the telescopic evidence can be accom­
modated by the geoheliocentric thesis of Tycho Brahe.
41. Nor was it better observationally than the Ptolemaic system.
42. One can see Copernicus struggling to accommodate the idea of a moving
earth within a revised Aristotelian physics.
43. Heavy objects would continue to move to the centre of the cosmos when in
natural motion.
44. This is a further reason why Copernicanism was not accepted, and not even
seen as a serious candidate until Galileo and Kepler.
45. Apology 26cd, Laws 886e, cf. Hippolytus Refutatio I , 8, 3-10.
46. Cf. Empedocles Fr. 31 (Hippolytus Refutatio V I I , 29, 13) 'Of the shape of
the cosmos, what is was like when ordered by Love, he speaks like this: "No twin
branches spring from its back, there are no feet nor nimble knees, no parts of
generation, but it was a Sphere and in all directions equal to itself." ' See too
Xenophanes Fr. 15 & 16 for early criticism of anthropomorphism in religion.
47. Stobaeus 1.18.1, M.R. Wright's translation and brackets (1995) p. 62.
48. See e.g. Plutarch On the Phenomenon of a Face in the Moon's Circle 926d
ff.
49. The six motions are up, down, left, right, backwards and forwards.
50. Cf. Phaedrus 245c ff. where self-motion is the essence and definition of soul
and Timaeus 37b where i t is that which is self moved.
51. Cf. Epinomis 982b on intelligence and regularity.
52. Both Plato and Aristotle would hold that these entities are fully alive,
intelligent, etc. but would want to distinguish them from terrestrial living, intelli­
gent entities.
53. Cf. Lloyd (1991) p. 149.
 Notes to pages 169-176 295
54. The Laws is also adamant that prayer or other offerings will be of no avail
in trying to change the minds of the gods.
55. Cf. Epinomis 986e on observations having been proved true by the test of
a great amount of time.
56. See e.g. Burtt (1932), Milton (1998). Leibniz comments that The general
principles of physics and mechanics themselves depend upon the action of a
sovereign intelligence and cannot be explained without taking it into considera­
tion' (Letter Explaining the Laws of Nature, cf. Discourse on Metaphysics 19-20).
57. I t was typical of the natural magic tradition that while they postulated
unseen (and so occult, literally 'hidden') relations between things, these relations
(forces, sympathies, correlations) behaved in a law-like manner, as opposed to the
demonic and spiritual magic traditions.
58. For Aristotle too the heavens are divine, see e.g. De Partibus Animalium
644b22 ff.
59. See Lennox (1985).
60. While Plato's god is omnibenevolent he is not omnipotent and so Plato is
not troubled by many of the paradoxes of omnipotence that trouble Christian
theologians.
61. Shorey (1935) vol. 2 p. 127 note e calls this a 'Humian, Comtian, positivist,
pragmatist view of causation'.
62. See below and Laws 720c, 938a.
63. Vlastos (1969) p. 106.
64. Vlastos (1969) p. 106.
65. I cannot see Plato subscribing to a great many of the assumptions which
underpin Hume's analysis of causation.
66. Vlastos (1969) p. 106.
67. On the following issues, see Curd and Cover (1998) pp. 879 f f , Ayer (1956),
Dretske (1977), Mellor (1980), Harre (2000a), (2000b).
68. I f F-ness G-ness, and a is F, why can we say that a is G as well?
69. See Dretske (1977) pp. 839 ff.
70. Vlastos (1969) p. 106, see Harre (2000b) and Newton-Smith (2000) on the
current state of the debate.
71. As science continues to develop, one would not expect nor seek an exact
match.
72. See Milton (1998). Plato may wish to challenge this line of thought though,
as Socrates in the Gorgias believes Callicles to be incorrect in considering nomos
and phusis to be opposed.
73. See Phaedo 103d.
74. See Phaedo 104a.
75. Both the relations between forms and the model of participation may well
change during Plato's career. I t is interesting to note in relation to the Phaedo that
Dretske (1977) also refers to the relation between universals as 'brings with it'.
76. Note that fire (that is a collection of fire atoms) is not necessarily hot, a
perceptual quality. I t will be hot to humans, as it will divide up our bodies, but may
not be so for the star gods whose bodies are mostly fire.
77. Vlastos (1969) p. 107.
78. See Cohen (1994) ch. 4 and Sambursky (1956).
79. Whether we consider these to be based on regularities or laws proper. See
here on magic, scholasticism and the scientific revolution; Hutchison (1982),
Webster (1982), Henry (1990), Nadler (1998), Copenhaver (1998).
80. Even if this process is not held to be fully reductive at each stage, it is held
to be highly informative.
296 Notes to pages 176-184
81. So for such influential thinkers as Descartes and Boyle. The later romantic
reaction to the mechanical philosophy of the seventeenth century was enthusiastic
about biological analogues, and the idea of the clash of opposites in nature, and
opposed the view that nature was most beautiful in geometric and ordered forms
with the idea that nature was most beautiful when primeval and untrammelled,
storms and volcanoes being more beautiful than geometrically arranged gardens.
See Hankins (1985) pp. 1-17.
82. See Hahm (1977).
83. SeeFurley (1987) p. 13.
84. I t may be that it is not until one has metal machines that mechanism can
become a good analogue for regularity and order.
85. Democritus Fr. 164, Sextus Empiricus Against the Mathematicians V I I ,
116-18, Barnes' translation and italics, Barnes (1987).
86. See Lloyd (1991) p. 148.
87. See Lloyd (1991) p. 149.
88. See Diogenes Laertius IX, 31, Aetius I I , 7, 2, Simplicius Physics 327, 24.
89. Mechanical processes are in general reversible.
90. See e.g. De Partibus Animalium 644b20 ff.
91. On Anaxagoras see Hippolytus Refutatio I , 8, 3-10, Plato Apology 26de and
Laws 886de, on the atomists see Diogenes Laertius IX, 31, Simplicius Physics 327,
24.
92. When Descartes revives the idea of vortices in the seventeenth century he
has the advantage of doing so within heliocentrism. This mean that he needs only
one vortex for the planets, created by the spin of the sun, and a vortex for each of
the planets (again created by their spin, to explain terrestrial gravity and why
planets such as earth and Jupiter carry moons with them). Note in respect to the
discussion below that each vortex has a physical entity on which to centre itself.
93. I would agree in general with Taub (1993) that the Platonic influence on
Ptolemy is significantly underestimated.
94. See Ptolemy Planetary Hypotheses I , 2, Goldstein p. 8.
95. See Ptolemy Planetary Hypotheses I I , 7 et passim.
96. See Ptolemy Planetary Hypotheses I I , 12.
97. See Ptolemy Planetary Hypotheses I I , 7.
98. See Ptolemy Planetary Hypotheses I I , 7.
99. I agree with Taub (1993) p. 123 and Pedersen (1974) p. 393 that Ptolemy's
distinction between the terrestrial and the celestial is much less sharp than that
of Aristotle.
100. See Vlastos (1975) p. 54 note 86.
101. See too the opening words of Copernicus Commentariolus (p. 57).
102. Copernicus De Revolutionibus I , 4.
103. Copernicus Commentariolus pp. 57-8.
104. Copernicus De Revolutionibus, Preface and Dedication to Pope Paul I I I .
105. Copernicus De Revolutionibus, Preface and Dedication to Pope Paul I I I .
106. Copernicus De Revolutionibus I , 1.
107. Copernicus' own system was on the whole no better than Ptolemy's, being
in some respects better and in others worse. A comment from Galileo is interesting
here: ' I cannot sufficiently admire the eminence of those men's intelligence who
have received and held the Copernican system to be true, and with the sprightli-
ness of their judgements have done such violence to their senses, that they have
been able to prefer that which their reason dictated to them to that which sensible
experiences represented most manifestly to the contrary.'
 Notes to pages 184-192 297
108. See e.g. Kuhn (1957), Koyre (1973), Ravetz (1990) and cf. Koestler (1959)
for an alternative view.
109. Such a return to a strict interpretation of ancient thought is typical of
Renaissance Neoplatonism.
110. See here, Kuhn (1957), Koyre (1973), Ravetz (1990), Field (1988).

7. Geometrical atomism - flux and language

1. Pyle (1997). Plato fares little better in other histories of atomism, see e.g.
Gregory (1931).
2. Aristotle Metaphysics 1.4 985b4 ff., KRS translation.
3. I t is important to remember that such views are also expressed in other late
works i n non-mythological settings and without being framed with such strong TW
assumptions.
4. Difficult, but note that whenever the world soul encounters anything sensi­
ble (aistheton, 37b6), opinions and beliefs which are stable and true occur (doxai
kai pisteis gignontai bebaioi kai alethes, 37b8), something I say more about in
Chapter 9. See Lee (1967) pp. 2-3 and Zeyl (1975) pp. 127-8 on whether Plato
considers the task too difficult.
5. Interesting to compare Plato's use of an ou mallon argument here with that
of the atomists. See Simplicius Physics 28, 8, and above Chapter l . V I I .
6. See e.g. Zeyl (1975) p. 127 ff.
7. I follow most commentators in removing kai ten tode at 49e3.
8. This formulation comes three times, at 49d4-5, 49d5-6 and 49e4. For Sim­
plicius on the passage, see in De Caelo 637, 4-5.
9. So Gill (1987) p. 36, Silverman (1992) p. 89, Hunt (1998) p. 70.
10. Cf. Gill (1987) p. 36, Hunt (1998) p. 71.
11. Gill (1987) p. 36, and see Cherniss (1954) p. 128.
12. On this see Gill (1987) pp. 40 ff., cf. Silverman (1992) pp. 91 ff. and Algra
(1994) 99-102. As will become clear, I too think there is something beyond the
tripartite ontology of forms, receptacle and phenomena, but not until the imper­
ceptible entities of geometrical atomism are introduced at 53c ff.
13. See Owen (1953) p. 322 ff., Reed (1972) p. 66 ff.
14. This argument defeats the lame plea of the Tim. that even if we cannot say
what any mere gignomenon is we can describe i t as to toiouton' Owen (1953) p. 323
note 3.
15. See Cherniss (1954).
16. See e.g. Owen (1953) p. 322 ff., Bostock (1988) p. 147 ff.
17. There would seem to be change every instant for all the properties under
consideration. A radical line here has been suggested to me by Verity Harte.
Perhaps i t is significant that Plato uses gold in this analogy, as opposed to say clay,
for while one can easily imagine clay being instantly remodelled by hand and so
being a reasonable analogue for radical flux, it is much harder to do so for gold,
which would have to be worked, stamped or cast into a different shape.
18. See Guthrie (1962) vol. 5 pp. 263-4, Gill (1987) p. 50, Algra (1994) pp. 76
ff., Reed 1972) p. 78.
19. This may also be a useful way of approaching the comments on flux in the
receptacle passages, which may not be consistent with each other and may present
a variety of options up to a strong and radical flux. I t may be that none of these
are Plato's position, but that we are offered a range of alternatives for any proposed
solution to the reference problem to be able to deal with.
20. Prior (1985) p. 113, cf. McCabe (1994) pp. 180 ff., Reed (1972) p. 75.
298 Notes to pages 192-201
21. See Gill (1987) p. 46.
22. Contra Cornford (1937) pp. 184 and 200, I do not see the problem with
translating morphe as shape here. The gold analogy begins specifically with
triangles and other shapes being formed, and a verb in the following sentence,
diaschematizomenon at 50c7, might well be though to echo schemata at 50b2.
Diaschematizomenon does not have to mean shaped (Cornford (1937) p. 185 note
1) but i n context that would seem likeliest with morphe as shape following on from
the talk of shapes in the gold example. That morphe cannot mean shape in the
Phaedo, i f true, does not seem to me to be of any great relevance here.
23. Interesting to note here that the Greek piles up four negatives in the last
part of the sentence and so is highly emphatic.
24. Cf. Aristotle, Categories V and Metaphysics V.8 and VII.3.
25. Cf. McCabe (1994) p. 187, Reed (1972) p. 75.
26. McCabe (1994) p. 187.
27. Gill (1987) p. 51.
28. Gill (1987) p. 51.
29. Note that Theaetetus 182c allows that we can speak of something if it is not
undergoing both phora and alloiosis.
30. See Zeyl (1975) pp. 146-7.
31. See Gill (1987) p. 51.
32. Kinesis is often rendered as 'motion', but I prefer change, as kinesis is quite
general, covering change in nature as well as change of position.
33. I take 'grows old' here to entail some sort of change and decay rather than
mere temporal existence.
34. Reading phoran rather than periphoran at Theaetetus 181b6 with Burnet.
Cf. Parmenides 138b on phora and alloiosis.
35. Let us for simplicity here exempt the account which describes the flux from
change in itself, to avoid the problems that Plato ascribes to Heraclitus, and
restrict our flux to the physical world. One can allow change into the account in a
limited way, as long as i t is ordered change, as then one can take as many
derivatives with respect to time as are required to produce a temporally invariant
description. The higher order account is then free from change as long as one is not
bound by an 'everything, whatever i t is, is in flux' clause.
36. See Theaetetus 182c ff.
37. See Cratylus 439b ff.
38. See Irwin (1977). On p. 4 he gives the following definition of a- and
s-change: 'x s-changes iff at Time t i x is F and at Time t2 x is not F, and x is not in
the same condition at t2 as it was at t i ... x a-changes iff x is F in one aspect, not
F in another, and x is in the same condition when i t is F and when it is not R'
39. Cf. e.g. Theaetetus 202e f f , Politicus 277e f f , Philebus 18b f f , and cf.
McCabe (1994) pp. 186-7.
40. See Timaeus 54d6, 55a8, 55b4, 57c9 and 61a7.
41. See Timaeus 54d ff. on how stoicheics form complex triangles.
42. See Friedlander (1958) pp. 246-60.
43. Cf. O'Brien (1984) pp. 341-58.
44. See Cornford (1937) pp. 230 ff. The problem with the Greek is whether
sustasin at Timaeus 57c9 refers to the putting together of the stoicheic triangles in
the sense of their internal construction or their external arrangement with other
stoicheics into complexes. Cornford argues convincingly in favour of the latter.
45. See De Caelo 306al-7 and De Generatione et Corruptione 332a27 ff. for
Aristotle's criticisms of Plato's theory of transmutation.
46. See Timaeus 81b f f , and especially 89bc.
 Notes to pages 202-211 299
47. There is a second variable here in that bonds of the same durability may
last different amounts of time depending on how much wear and tear they are
subject to.
48. Earth, on this account, says Aristotle has the best right to be called an
element because earth alone is indestructible (306al8-21).
49. See Cornford (1937) p. 329 note 2, Taylor (1928) pp. 585-6, cf. LSJ II.2 on
hriza.
50. See Taylor (1928) p. 586, LSJ strongly supports Taylor on chalao.
51. ek druochon was a nautical metaphor for something newly made, like a
ship 'fresh off the stocks'.
52. I f someone asks what the timber and then the subsequent boat are, the
safest answer I can give is wood; cf. the gold example, etc.
53. LSJ gives 'to place or set together, to combine, to organise, pass, to stand
together', cf. Cornford (1937) pp. 230 ff who argues for this sense.
54. Cf. Timaeus 56e, 81b and 89c.
55. Archer-Hind (1888) notes to pp. 304-6 simply substitutes 'particles' for
trigona.
56. This directly relates to Timaeus 8Id; both passages discuss the marrow.
57. There is one difficulty with the interpretation that trigona at Timaeus 81b
refers exclusively to the complex triangles which constitute atoms, which is that
not all atoms are constituted out of complex triangles. This is the case for air, fire
and water of course, but for the smallest 'isotope' of earth, the atomic faces are
squares made of four stoicheic triangles. My interpretation requires that this is a
slight oversight on Plato's part.
58. See for instance Timaeus 53d duoin archetai trigonoin, 54a duoin
trigonoin, 55b and 57c ton stoicheon, 55e ton kat'archas trigonon hupothenton.
59. Cf. Timaeus 81c ff.
60. Zeyl (1975) p. 125.
61. See Timaeus 56bc.
62. Cf. McCabe (1993) p. 181.
63. See Silverman (1992) p. 87, cf. Schulz (1966) pp. 65-86 who considers that
Plato reduces sensible qualities and ultimately has shapes and the receptacle.
64. As the Theaetetus is rather indeterminate on what it may be to give a logos
at this point, I hedge between account and analysis for logos here.
65. Cf. e.g. Theaetetus 2 0 l e i , 202b7 and Timaeus 48bc, and passim references
to the two basic triangles as stoicheia.
66. So contra McCabe (1994) pp. 186-7 Plato does not repudiate the letters and
syllables analogy, nor does the Timaeus hold that analysis by decomposition is the
exclusive approach to cosmology (p. 182).
67. I take i t that the teleological account here would count, as G. Fine (1979b)
p. 317 has put it, as logosk, that is explanations that will help to produce knowl­
edge rather than logos , mere sentences about elements which may or may not be
s

true.
68. I would add the omission of teleology from the Theaetetus is as striking as
the omission of forms cf. Cornford (1935) pp. 106-11,158-63, Ross (1951) pp. 124-5.
69. Vlastos (1975) p. 68.
70. See Sayre (1983) pp. 113 and 154-5. So too one can characterise an instant
of time as a Dedekind cut on a time line; cf. Parmenides 156d and Sayre p. 114.
71. Cf. Aristotle De Anima 404b23 and Metaphysics 1086al0 on forms and
numbers.
72. See Gill (1987) pp. 51-2.
73. See Gill (1987) pp. 51-3.
300 Notes to pages 211-220
74. The Sophist use of the letters and syllables analogy looks to the commun­
ion of forms, the Theaetetus and Philebus are (arguably!) more neutral and about
world generally, the Politicus looks to the structure of politics and society.
75. See Sophist 253a5, Philebus 18c8
76. See especially Timaeus 81d4, 81d5.
77. See desmos at Timaeus 31cl, 36a9, 38e5, 41b5, 43a2, 73b3, 73d6, 77e3,
84a3 and cf. Politicus 309c.
78. Cf. Aristotle Metaphysics V.8 and VII.2-3.
79. Certainly a different ontology including non-perceptible physical entities,
as I have argued.
80. See Chapter 1 for more on this passage.

8. Geometrical atomism - matter and space

1. See Algra (1994) p. 78, Sorabji (1988) pp. 32-5.
2. For problems with the matter theory, see Algra (1994) pp. 78-83.
3. Cf. Algra (1994) p. 86.
4. See the next section for occurrences of chora and hedra outside Timaeus
48e-53b.
5. For problems with the space theory, see Algra (1994) pp. 83-9. Algra does not
discuss the plokanon metaphor, though i t seems strongly spatial.
6. See Descartes Principia Philosophia II.4 ff.
7. On this see Algra (1992) pp. 81 ff. et passim, and cf. Guthrie (1962) vol. 5 p.
265.
8. See Algra (1994) p. 74 and Lee (1966) for an opposing view.
9. See O'Brien (1984) pp. 359-65.
10. See O'Brien (1984) p. 361.
11. See Cherniss (1944) pp. 154 ff.
12. See O'Brien (1984) p. 360.
13. Aristotle, De Caelo 306b3 ff.
14. Compare Descartes, who also has a packing problem, who allows all shapes
and sizes of corpuscles, and so has a very fine matter i n which the larger pieces are
able to swim around.
15. So Taylor (1928) p. 399 and Cornford (1937) pp. 242-5.
16. See Taylor (1928) pp. 398-9, Cornford (1937) pp. 243-6, and see O'Brien
(1984) pp. 361-2 for a critique of their positions. Cornford's translation of Timaeus
81b6 ff. plays down both the compressive effect of the body/cosmos and the
possibility that chopping up blood particles will result i n voids.
17. So O'Brien (1984) p. 362.
18. See here Aition (1972) on both ancients and moderns.
19. One might argue though that this cuts both ways, and the surrounding
material metaphors are supposed to be explained by en hoi. There are possibly
more spatial than material metaphors, but I doubt that a count will help here.
20. Cornford (1937) p. 187 note 1.
21. See Aristotle Generation of Animals 763b ff.
22. See Algra (1994) p. 94 note 49.
23. See Theaetetus 191c, 194de, 196ab, and cf. Timaeus 72c. Keros is also used
in the Theaetetus, see e.g. 194c5 ff.
24. Cf. the use of apotopousthai 191d6 (like the impression from a seal ring).
25. Mohr (1985) p. 94.
26. Algra (1994) p. 85.
27. Algra (1994) p. 86.
 Notes to pages 222-229 301
28. See Timaeus 57cl, 58a7, 79d6, 82a3 and 83a4.
29. See Timaeus 59a3, 60c4, 62a, 67b, 72c, and 79b2.
30. See Timaeus 59d6 and 62b7.
31. See Timaeus 70c5, 74cl, 74dl and 78d3.
32. I t is perhaps important that geometrical atomism has its own analysis of
these phenomena. I f we want to explain ignification and liquefaction, geometrical
atomism will tell us that we perceive fire and water when sufficient numbers of
their respective atoms have accumulated together, and would also have a tale to
tell about change in terms of the stoicheic triangles and their complexes which go
to make up these atoms. So too gold, for geometrical atomism, gold can be analysed
into elements, atoms and ultimately stoicheic triangles.
33. There is also a serious problem concerning the localisation of phenomena,
as explained in the next section.
34. I n some commentators this is explicit (e.g. Cornford (1935), (1937), Yolton
(1949)), in many others i t is implicit.
35. See Burnyeat (1991) and below.
36. See Taylor (1928) p. 430 ff.
37. See Cornford (1935) pp. 48-51 and (1937) p. 261 note 1, and cf. Nahnikian
(1955) and O'Brien (1984) pp. 144-74. I n my view it is significant that Cornford has
an 'A' reading of the Theaetetus - see next chapter.
38. See McDowell (1973) pp. 138-40.
39. See G. Fine (1988a) pp. 20-1.
40. See e.g. Theaetetus 157ab.
41. See Timaeus 56bc on the perceptual status of the entities of geometrical
atomism.
42. See Theaetetus 153e and cf. Timaeus 61c ff.
43. Timaeus 67c5-7, Fine's translation.
44. A. Fine (1988a) p. 21.
45. Cf. Timaeus 62b ff. on hard and soft.
46. Cf. O'Brien (1984), (1997).
47. See Timaeus 61e on the nature of fire.
48. Aristotle is critical of the role of geometrical atomism in Plato's conception
of weight. A t De Caelo 299b 15 he gives us the principle that Weight cannot be
constituted out of that which has no weight... if it is weight that makes one weight
heavier than another weight, then each indivisible part must have weight.' Aris­
totle argues that if the reason for a difference in weight is not this, but that earth
is heavier than fire, then some surfaces will be light and some heavy. Aristotle's
example would seem i l l chosen, as Plato could accept that the surfaces that make
up earth are heavier than those of fire (or water or air) as earth does not transmute
and has different stoicheic triangles. Aristotle also needs to be more discriminating
about what he is referring to by epipedos. Does he mean complex triangles, which
might well differ from each other under the 'isotope' theory, or does he have the
stoicheic triangles in mind? The basic principle he proceeds from seems suspect,
as in general we cannot infer from a whole having a property that all the parts do
as well, so some more argument in the case of weight is required.
49. Cf. Burnyeat (1991) p. 8, cf. Theaetetus 160de.
50. Assuming a 'B' reading here.
51. See Theaetetus 152a.
52. On this, and whether the Theaetetus theory is Plato's, see Day (1997).
53. Cf. Theaetetus 182c ff.
54. See Timaeus 43e ff., and next chapter.
302 Notes to pages 229-240
55. We are likely to mix attributions of sameness and difference, or judge left
to be right.
56. Aristotle, Physics 209bl3 ff.
57. See Aristotle, Physics 209b23 f f , 211a36 f f , and cf. Sorabji (1988) p. 36.
58. Aristotle, De Caelo 305b29-31.
59. Aristotle, De Caelo 306bl5 ff.
60. Aristotle, De Generatione et Corruptione 329al3-17.
61. Cf. E.N. Lee (1966) pp. 349 ff.
62. Aristotle De Generatione et Corruptione 329al4-15, 329a21-24.
63. I t is worth noting that the evidence here too is from De Caelo and De
Generatione et Corruptione, rather than the Physics.
64. See e.g. Silverman (1992) p. 97 note 26.
65. Cf. T.M. Robinson (1970) p. 93, Mohr (1985) pp. 92-3.
66. See Timaeus 45a f f , and 71a.
67. See Kung (1988), cf. Guthrie (1962) vol. 6 p. 265 note 3.
68. See Algra (1994) p. 92 note 43.
69. See Kung (1988) p. 167. Indeed, Plato also describes the relation as
aporotata.
70. Cornford (1937) p. 209, Cornford's capital for space.
71. Cf. O'Brien (1984), (1997).
72. See e.g. Descartes Principia Philosophia 11.10, 11, 16.
73. See e.g. Descartes Principia Philosophia 11.20.
74. See Burtt (1932) pp. 96 ff.
75. See e.g. Descartes Principia Philosophia IV.20 ff. (gravity).
76. Generated by rubbing amber, here electron 81c2.
77. The lodestone, here ton Heracleion lithon, 81c2.
78. Crombie (1963) vol. 2 p. 222.
79. Though Descartes is careful to say that his particles are impenetrable.
80. See Archer-Hind (1888) pp. 244-5, Taylor (1928) pp. 470-2, Cornford (1937)
pp. 271-5.
81. Vlastos (1967).
82. Taylor (1928) p. 370.
83. Cornford (1937) pp. 231-9.
84. Artman and Schafer (1993) p. 258.
85. See Artman and Schafer (1993) pp. 256-61.
86. See Meno 82b ff.
87. Diagrams here following Artman and Schafer (1993).
88. See Artman and Schafer (1993) p. 260.
89. Vlastos (1975) pp. 68-9.
90. Vlastos (1975) pp. 71-2. Vlastos does not suggest this. Cf. M.R. Wright
(1995) pp. 102-3.
91. Vlastos (1975) pp. 85 ff.
92. Though Aristotle is critical that Plato gives no proper account here, see De
Caelo I I I .
93. Boyle, A Disquisition about the Final Causes of Natural Things, Works
(1688) vol. 5 p. 409.
94. Newton, Optics IV.260.
95. Descartes, Principia Philosophia IV. 133 ff.
96. See Popper (1952) pp. 75 f f , and e.g. Aristotle Metaphysics 985b31 ff.
97. See e.g. Aristotle Metaphysics 987bll-13.
98. Plato is well aware of the irrationality of several numbers; see Theaetetus
 Notes to pages 240-246 303
147c f f , where the square roots of three, five, and seventeen are mentioned, along
with a general term ('oblong numbers') for all numbers with irrational roots.
99. Cf. also Epinomis 990d ff. on numbers and modelling.
100. See Popper (1945) p. 248 note 9 and (1952) p. 87ff, also M.R. Wright
(1995) p. 54.
101. See Popper (1952) p. 88.
102. Proclus Euclid Commentary p. 71, Popper's translation.
103. Popper (1952) p. 88.

9. Epistemology in the Timaeus and Philebus

1. Cf. Theaetetus 189e f f . Sophist 263e ff. and Philebus 38a ff. on the relation
of judging as silent statement and statement as spoken judgement.
2. There is a question here as to whether this 'pair' refers to intelligence and
knowledge or to (opinions and beliefs) and (intelligence and knowledge); see
Cornford (1937) p. 96 note 1.
3. Cf. Timaeus 35a ff.
4. See Timaeus 42a ff. As well as desire, fear and anger, mortals are subject to
'all such emotions as follow with them, and all that are of a contrary nature'.
5. Cf. Timaeus 34a, 47c, 90cd.
6. See Timaeus 43b.
7. See Timaeus 36e.
8. Cf. Sophist 266c ff.
9. But woe betide the indolent! Timaeus continues at 44d 'Should he be
neglectful, he lives a lame existence, he returns to Hades unperfected and lacking
in intelligence.' This translation of holokleros hugies te pantelos as 'perfect and
entirely sound in mind' is justified by comparison with the state of the negligent
who have to return to Hades who are left ateles kai anoetes, 'unperfected and
lacking in intelligence'.
10. Cf. Theaetetus 176bc and Sedley (1997) here.
11. I t might be argued that Sophist 255c ff. explicitly distinguishes between
one and two place predicates, although I tend to the view that the distinction there
is between the 'is' of identity and the 'is' of predication; cf. Owen (1971) and Reeve
(1985) pp. 54-5, Cornford (1935) pp. 282-5, Malcolm (1967) p. 143 note 24. Cf.
Bostock (1986) pp. 49-51 and 181-3 on properties and relations in the Phaedo.
12. For complete uses, see Timaeus 35a and passim, for incomplete see e.g.
Timaeus 37a and 44a.
13. Cf. Frede (1996).
14. See Timaeus 37a f f , 44a ff.
15. Plato treats thought and judgement as i f they have the logical charac­
teristics of speech and assertion; cf. Theaetetus 189a f f . Sophist 263e f f , Burnyeat
(1991) pp. 84-5.
16. See e.g. Philebus 38c5 f f , 42a ff. and Sophist 234b and the discussion of
what we can see at a distance. Note though that Plato's concern there is to develop
accounts of what we can visualise.
17. I side with Cooper (1970) p. 123 ff. against Cornford (1935) pp. 102-9 that
the main contrast at Theaetetus 184-6 is between perception and judgement rather
than objects of opinion (sensibles) and objects of knowledge (forms).
18. See Theaetetus 185a, 185c.
19. See Theaetetus 185b, 185c.
20. See Theaetetus 186a.
304 Notes to pages 247-255
21. Cf. Epinomis 978b f f , which if anything is even clearer and more forthright
on this matter.
22. See too Aristotle De Anima 404bl6, cf. 406b25 ff., and cf. Sophist 254a ff.
23. The theme of bebaiotes, 'fixity' or 'security' plays an important part in later
Plato. Cf. Philebus l i e , 14c, 15b and 62b, Sophist 250c, the discussion of the
Cratylus concerning the fixity of names (and the digression of Epistle VII), and
MacKenzie (1986a) pp. 141-2.
24. Cf. Theaetetus 186c, where Socrates states that: 'Knowledge is not then i n
the sensations, but in the process of reasoning (sullogismoi) concerning them; for
it is possible to apprehend being and truth in this manner, but impossible other­
wise.'
25. There is of course an important parallel and contrast here with Republic
V I I ; in both Plato first treats eyesight and then moves to draw conclusions about
hearing. I n the Timaeus he continues 'Again, the same account applies to both
sound and hearing, as they were given by the gods with the same intent and
purpose' (Timaeus 47c). Cf. especially Republic 530d, 'As our eyes are framed for
astronomy, so are our ears for harmony, and these two are in some way kindred
disciplines'.
26. A t 37bc where the world soul is producing stable and true opinion, and
knowledge and reason, the same's motion is described as eutrochos, and that of the
different is orthos.
27. See Frede (1988), cf. Zeyl (2000) xxx ff.
28. Frede's translation (1988) p. 41, cf. Republic 507b.
29. See Phaedo 96d, and the solution at lOOe ff. I t is sometimes suggested that
the myth of the Phaedo is a 'likely story' in the manner of the Timaeus. However,
while Socrates in the Phaedo does not claim certainty for his tale, there is no
suggestion that all that can be achieved is a likely story, merely that this is an
hypothesis that would take an inordinate amount of time to prove; see Phaedo
108d, 114d.
30. See Aristotle De Caelo 306a9 ff.
31. Cf. especially Parmenides Fr. 1/28 ff. and Timaeus 29b.
32. Parmenides Fr. 8/60, ' I tell you the entire arrangement as seems likely, so
that no opinion of mortal men will ever surpass you'.
33. See Parmenides Fr. 8/50 and cf. Timaeus 29c ff.
34. Cf. Parmenides Fr. 8/60 and Timaeus 29c, and passim comments in the
Timaeus that this is the most likely account.
35. See Theaetetus 18Id ff.
36. See Aristotle Metaphysics 1078b 16 ff. on Plato's early Heraclitean influ­
ences.
37. I t is also worth noting that Socrates allows a long speech rather than
question and answer, and is well dressed, again contrary to his usual custom, cf.
Schoos (1999) p. 103.
38. Cf. McCabe (1994) p. 185.
39. Sayre(1983)ix.
40. This is not to say that Plato does not have views, either positive or
negative, that he wishes to get across. See Burnyeat (1980) pp. 186 f f , cf. Annas
(1981), Burnyeat (1981) pp. 97-139, MacKenzie (1986b) p. 22 ff., McCabe (1993).
41. To use Burnyeat's example, it is one thing to know that root 3 is irrational,
but quite another to understand why that should be so.
42. Cf. Phaedrus 275e.
43. I f we are to believe Dionysius of Halicarnassus in his On Literary Compo­
sition, Plato painstakingly wrote and re-wrote each of his works.
 Notes to pages 255-263 305
44. Paradigm cases here are the Laches, with physical and intellectual concep­
tions of courage, and the Cratylus with its conventionalist and naturalist accounts
of naming. Cf. Sophist 251a, where we must fend off both arguments at once, and
force a passage between them.
45. See here MacKenzie (1982a).
46. See my comments in Chapter 3 on complex irony.
47. One implication of this is that we cannot attribute the book paradox in its
fully-fledged form to Plato. I t is employed as a tool to provoke the reader to
investigate the text at a different level.
48. Note also the resonances between the fertility metaphors between
Theaetetus at 149b ff. and 210b, and those at Phaedrus 276b ff.
49. So I would disagree with Arieti (1991) who sees the Timaeus as a playful
parody of scientific thinkers.
50. See Phaedrus 229d ff. Note also that there are no criteria here for judging
the matter one way or the other.
51. Typho was a monster with a hundred heads, so I translate haplousteron as
single-minded.
52. See Phaedrus 229bc; the questions are whether the myth is to be believed,
and the exact spot where the abduction is supposed to have happened.
53. See Phaedrus 229c. Was Boreas in fact the north wind which blew Orethuia
to her doom?
54. Cf. Stewart (1960) pp. 98 ff. and T.M. Robinson (1970) p. 60 on myths
'which engage a man'.
55. Cf. Osborne (1988), (1996).
56. Cf. Lloyd (1991) pp. 342-3.
57. Cf. Timaeus 21d and 26d.
58. See especially Phaedrus 275bc, cf. Vlastos (1939) p. 380.
59. Cf. Phaedrus 274a ff.
60. See Timaeus 27a.
61. Here I would concur with Graham (1990) p. 21, who comments that 'If we
think of the Timaeus as setting out the general conditions for a teleological
programme of explanation in nature, there is no particular reason why Plato
should have to interpret further the model for us. Here, he might say, are the tools
for a teleological natural science. I present them in mythical form as a kindiof
archetype for you. You must decide on how to interpret the model and the
categories of the model.'
62. Cf. Burnyeat (1982) pp. 4-8 and especially the notes on pp. 6-7.
63. Burnyeat (1991) p. 8.
64. Burnyeat (1991) p. 8.
65. Burnyeat (1991) p. 10.
66. See Cornford (1937).
67. Note that while I believe we can have similar 'A' readings of the Theaetetus
and Timaeus, the alternative readings need not be similar.
68. See Theaetetus 210bc.
69. Though the Timaeus does not offer a definition of knowledge.
70. Both Theaetetus and Timaeus would appear to allude to the Phaedo at
many points and provide revised views.
71. Cf. Gosling (1975) pp. 153-4, 170-80, 196-205 for a discussion of techne in
the Philebus.
72. Cf. my earlier discussions of techne, empeiria and tribe.
73. See e.g. Gorgias 465a, 500e.
74. Xenophon Oeconomicus IV, 2.
306 Notes to pages 263-264
75. Cf. Philebus 56bc.
76. According to Aristotle, Plato thought that 'Concrete things (ta pragmata)
are apprehended in some cases by mind, in others by knowledge, by opinion, or by
sensation, and the numbers are the forms (eide) of these things (pragmata)' (De
Anima 404b26), though cf. Metaphysics 987a32, 1078bll.
 Bibliography

Adam, J. (1902) The Republic of Plato, 2 vols, Cambridge U.R

Aition, E.J. (1972) The Vortex Theory of Planetary Motions. Baird.
Alexander, H.G. (1956) The Leibniz-Clarke Correspondence. Manchester U.P.
Algra, K. (1994) Concepts of Space in Greek Thought. Brill, Leiden.
Annas, J. (1981) An Introduction to Plato's Republic. Oxford U.P.
(1982a) Aristotle on Inefficient Causes. Philosophical Quarterly 32, pp.
311-26.
(1982b) Knowledge and Language: The Theaetetus and the Cratylus. I n M .
Schofield and M.C. Nussbaum (eds.), Language and Logos: Studies in Ancient
Greek Philosophy Presented to G.E.L. Owen, Cambridge U.P, pp. 95-114.
Anton, J.P. (1981) Science and the Sciences in Plato. Buffalo N.Y.
Archer-Hind, R.D. (1888) The Timaeus of Plato. Macmillan and Co, London.
Arieti, J.A. (1991) Interpreting Plato. Rowman and Littlefield.
Ariew, R. (1984) The Duhem Thesis, British Journal for the Philosophy of Science
35, pp. 313-25.
Artman, B. and Schafer, L. (1993) On Plato's 'Fairest Triangles' (Timaeus 54a).
Historia Mathematica 20, pp. 255-64.
Ayer, A.J. (1956) What is a Law of Nature? Revue Internationale de Philosophic 67,
pp. 144-65
(1963) What is a Law of Nature? I n Curd and Cover (1998).
Ayers, M . and Garber, D. (1998) The Cambridge History of Seventeenth-Century
Philosophy. Cambridge U.P.
Baeumker, C. (1890) Das Problem der Materie in der Griechieschen Philosophie.
Munster.
Ballew, L. (1974) Straight and Circular in Parmenides and the 'Timaeus'. Phrone-
sis 19, pp. 189-209.
Baltussen, H . (1993) Theophrastus on Theories of Perception: Argument and
Purpose in De Sensibus. Utrecht U.P.
Barker, A. (1978) Sumphonoi Arithmoi: A Note on Republic 531c 1-4. Classical
Philology 73, pp. 337-42.
(1994) Ptolemy's Pythagoreans, Archytas, and Plato's Conception of Mathe­
matics. Phronesis 39, 1994, pp. 113-35.
Barnes, J. (1979) The Presocratic Philosophers. 2nd ed, 1982, Routledge and
Kegan Paul, London and N.Y.
(1987) Early Greek Philosophy. Penguin.
Barrow, J.D. (1991) Theories of Everything. Vintage.
Barrow, J.D. and Tipler EJ. (1986) TheAnthropic Cosmological Principle. Oxford
U.P.
Bergstrom, L. (2000) Quine. I n Newton-Smith (2000).
308 Bibliography
Bedu-Addo, J.T. (1979) The Role of the Hypothetical Method i n the Phaedo.
Phronesis 24, pp. 111-32.
Bobrovnikov, N.T. (1984) Astronomy Before the Telescope I . Pachart, Tucson.
(1990) Astronomy Before the Telescope II. Pachart, Tucson.
Bodemar, C.W. (1964), Regeneration and the Decline of Preformationism in Eight­
eenth-Century Embryology. Bulletin of the History of Medicine 38, pp. 20-31.
Bodnar, I.M. (1997) Alexander of Aphrodisias on Celestial Motions. Phronesis 42,
pp. 190-205.
Bolton, R. (1975) Plato's Distinction between Being and Becoming. Review of
Metaphysics 29, 1975/6, pp. 66-95.
Bonjour, L. (1986) The Structure of Empirical Knowledge. Harvard U.P.
Borelli, G.A. (1680) De Motu Animalium.
Bostock, D. (1978) Plato on Time and Change in the Parmenides. Phronesis 23,
1978, pp. 229-41.
(1986) Plato's Phaedo. Oxford U.P.
(1988) Plato's Theaetetus. Oxford U.P.
Bowen, A.C. (1991) Science and Philosophy in Classical Greece. Garland, N.Y.
Boyd, R. (1973) Realism, Underdetermination, and a Causal Theory of Evidence.
Nous 7, pp. 1-12.
Boyle, R. (1688) The Works of the Honourable Robert Boyle, ed. T. Birch, London.
Brandwood, L. (1990) The Chronology of Plato's Dialogues. Cambridge U.P.
Brisson, L. (1987) Platon, Lettres. Paris.
(1992) Timee: Critias. Flammarion, Paris.
(1994) Le Meme et Vautre dans la structure ontologique du Timee du Platon.
Academia Verlag, Sankt Augustin.
(1995) Interpretation du mythe du Politique. I n Rowe (1995a), pp. 349-63.
(1997) Perception sensible et raison dans la Timee. I n Calvo and Brisson
(1997).
(1999) Plato the Myth Maker. Chicago U.P.
Brisson, L. and Meyerstein, F.W. (1995) Inventing the Universe: Plato's Timaeus,
the Big Bang, and the Problem of Scientific Knowledge. SUNY.
Brocker, W. (1963) Der Philosophische Exkurs in Platons Siebtem Brief. Hermes
91, pp. 416-25.
(1965) Nachtrag zum Philosophischen Exkurs in Platons Siebentum Brief.
Hermes 93, pp. 132.
Brown T.M. (1974) From Mechanism to Vitalism in Eighteenth-Century English
Physiology. Journal of the History of Medicine 7, pp. 179-216.
Bruins, E.M. (1951) La Chimee du Timee. Revue de Metaphysique et Morale 56, pp.
269-82.
Brumbaugh, R.S. (1961) Plato and the History of Science. Studiem Generale 14/9,
pp. 520-7.
Bulmer-Thomas, I . (1984) Plato's Astronomy. Classical Quarterly 34, pp. 107-12.
Burdach, K. (1922) Die Lehre des Platonischen Timaios (40b) von der kosmischen
Stellung der Erde. Neue Jahrbuch fur der klassischen Altertum 49, pp. 254-78.
Burge, E.L. (1971) The Ideas as Aitiai in the Phaedo. Phronesis 16, pp. 1-13.
Burkert, W. (1972) Lore and Science in Ancient Pythagoreanism. Harvard U.P.
Burnet, J. (1911) Plato's Phaedo. Oxford U.P.
(1924) Greek Philosophy: Thales to Plato. London.
Burnyeat, M.F. (1976) Plato on the Grammar of Perceiving. Classical Quarterly,
NS 26, pp. 29-51.
(1980) Socrates and the Jury: Paradoxes in Plato's Distinction between
 Bibliography 309
Knowledge and True Belief. Proceedings of the Aristotelian Society, Supp. Vol.
54, pp. 173-91.
(1981) Aristotle on Understanding Knowledge. I n E. Berti (ed.), Aristotle on
Science: The Posterior Analytics, Padua U.P.
(1982) Idealism and Greek Philosophy: What Descartes Saw and Berkeley
Missed. Philosophical Review 91, pp. 3-40.
(1987) Platonism and Mathematics: A Prelude to Discussion. I n A. Graeser
(ed.), Mathematics and Metaphysics in Aristotle, Bern, pp. 213-40.
(1991) The Theaetetus of Plato. Hackett, Indianapolis/ Cambridge.
Burtt, E.A. (1932) The Metaphysical Foundations of Modern Physical Science.
Kegan Paul, Trench and Trubner, London.
Bury, R.G. (1897) The Philebus of Plato. Cambridge U.P.
Bynum, W.F., Browne E.J. and Porter, R. (1981) The Dictionary of the History of
Science. Macmillan, Hong Kong.
Calder, W.M. (1958) The Spherical Earth in Plato's Phaedo. Phronesis 3, pp. 121-5.
Callahan, J.F. (1948) Four Views of Time in Ancient Philosophy, Harvard U.P.
Calvert, B. (1970) Forms and Flux in Plato's Cratylus. Phronesis 15, pp. 26-47.
Calvo, C. and Brisson, L. (1997) Interpreting the Timaeus-Critias. Proceedings of
the IV Symposium Platonicum. Academia Verlag, Sankt Augustin.
Carone, G.R. (1997) The Ethical Function of Astronomy in Plato's Timaeus. I n
Calvo and Brisson (1997).
Cavagnaro, E. (1997) The Timaeus of Plato and the Erratic Motion of the Planets.
In Calvo and Brisson (1997), pp. 351-62.
Cherniss, H.F. (1936) The Philosophical Economy of the Theory of Ideas, in R.E.
Allen (ed.), Studies in Plato's Metaphysics, Routledge and Kegan Paul, London
andN.Y. 1965, pp. 1-12.
(1944) Aristotle's Criticism of Plato and the Academy, Baltimore U.P.
(1945) The Riddle of the Early Academy. California U.P.
(1954) A Much Misread Passage in the Timaeus (Timaeus 49c7-50b5).
American Journal of Philology 75, pp. 113-30.
(1957a) The Relation of the Timaeus to Plato's Later Dialogues. I n R.E.
Allen (ed.), Studies in Plato's Metaphysics, Routledge and Kegan Paul, London
and N.Y, 1965, pp. 339-78.
(1957b) Timaeus 38a8-b5. Journal of Hellenic Studies 77, pp. 18-23.
(1975) The Sources of Evil According to Plato. American Philosophical
Quarterly 12, pp. 244-58.
Cherry, R.S. (1967) Timaeus 49c-50b. Apeiron 2, pp. 1-11.
Claghorn, G.S. (1954) Aristotle's Criticism of Plato's Timaeus. Nijhoff.
Cleary, J.J. (1987) Science, Universals and Reality. Ancient Philosophy 7,1987, pp.
95-130.
(1997) Plato's Teleological Atomism. I n Calvo and Brisson (1997).
Clegg, J.S. (1976) Plato's Vision of Chaos. Classical Quarterly NS 26, pp. 52-61.
Code, A. (1988) Reply to Michael Frede's 'Being and Becoming in Plato'. Oxford
Studies in Ancient Philosophy 6, pp. 53-60.
Cohen, H.F. (1994) The Scientific Revolution. Chicago U.P.
Cohen, S.M. and Keyt, D. (1992) Analysing Plato's Arguments: Plato and Plato­
nism. Oxford Studies in Ancient Philosophy, Supp. Vol. 1992, Methods of
Interpreting Plato and his Dialogues, pp. 173-200.
Cooper, J.M. (1970) Plato on Sense Perception and Knowledge: Theaetetus 184 to
186. Phronesis 15, pp. 123-46.
(1977) The Psychology of Justice in Plato. American Philosophical Quar­
terly 14, pp. 151-7.
310 Bibliography
Copenhaver, B. (1998) The Occultist Tradition and its Critics. I n Ayers and Garber
(1998).
Cornford, F.M. (1932) Mathematics and Dialectic i n the Republic V I - V I I . Mind 41,
(I) pp. 37-52, (II) pp. 173-90.
(1935) Plato's Theory of Knowledge. Routledge and Kegan Paul, London.
(1937) Plato's Cosmology. Routledge and Kegan Paul, London.
(1939) Plato and Parmenides. Routledge and Kegan Paul, London and N.Y.
(1941) The Republic of Plato. Clarendon, Oxford.
(1950) Before and After Socrates. Cambridge U.P.
(1952) Principium Sapientiae. Cambridge U.P.
Cottingham, J , Stoothoff, R. and Murdoch, D. (1985) The Philosophical Writings
of Descartes, 3 vols, Cambridge U.P.
Cresswell, M.J. (1971) Plato's Theory of Causality: Phaedo 95-106. Australian
Journal of Philosophy 49, pp. 244-9.
Crombie, I.M. (1963) An Examination of Plato's Doctrines, 2 vols, Routledge and
Kegan Paul, London.
Curd, M . and Cover, J.A. (1998) Philosophy of Science: The Central Issues, Norton,
N.Y.
Cushing, J.T. (1998) Philosophical Concepts in Physics. Cambridge U.P.
Dampier-Whetham, W.C. (1948) A History of Science. Cambridge U.P.
Davies, P.C.W. (1982) The Accidental Universe, Cambridge U.P.
(1992) The Mind of God. Penguin.
Davis, M . (1967) The Statesman as a Political Dialogue. American Journal of
Philology 88, pp. 319-31.
Davis, P.J. (1979) The Fourfold Classification in Plato's Philebus. Apeiron 13, pp.
124-34.
Day, J.M. (1997) The Theory of Perception i n Plato's Theaetetus 152-183. Oxford
Studies in Ancient Philosophy 15, pp. 51-80.
Day, T. and Kincaid, H . (1994) Putting Inference to the Best Explanation in its
Place. Synthese 98, pp. 271-95.
De Santillana, G. (1949) Eudoxus and Plato, Isis 32, pp. 248-62.
(1961) The Origins of Scientific Thought, vol. I , From Anaximander to
Proclus - 600 BC to 300 AD. Weidenfeld and Nicolson, London.
De Vogel, C.J. (1955) The Present State of the Socratic Problem. Phronesis 1, pp.
26-35.
Dicks, D.R. (1970) Early Greek Astronomy to Aristotle. Thames and Hudson,
Bristol.
Diels, H . (1922) Die Fragmente der Vorsokratiker. Weidman, Berlin.
Dillon, J. (1995) The Neoplatonic Exegesis of the Statesman Myth. I n Rowe
(1995a), pp. 364-74.
Dombrowski, D.A. (1981) Atlantis and Plato's Philosophy. Apeiron 15, pp. 47-60.
Dretske, F.I. (1977) Laws of Nature. I n Curd and Cover (1998).
Dreyer, J.L.E. (1906) A History of Astronomy. 2nd ed, Dover 1953.
Driscoll, J. (1979) The Platonic Ancestry of Primary Substance. Phronesis 24, pp.
253-69.
Duhem, P. (1906a) Le Systeme du monde: histoire des doctrines cosmologiques de
Platon a Copernic, vol. I . Herman, Paris 1954.
(1954) The Aim and Structure of Physical Theory. Originally La Theorie
physique, son objet et sa structure, translated by P. Wiener, from the second
edition (Riviere, Paris 1914), Princeton.
(1969) To Save the Phenomena. Originally 'Sozein ta Phainomena; Essai sur
 Bibliography 311
la Notion de Theorie Physique de Platon a Galilee' (Annales de Philosophie
Chretienne, 1908), trans. E. Doland and C. Maschler, Chicago,
Edwards, P. (1972) The Encyclopaedia of Philosophy, 8 vols, Macmillan, N.Y. 1972.
Einstein, A. (1973) Ideas and Opinions. Souvenir Press.
Erler, M . (1995) Kommentar zu Brisson und Dillon. I n Rowe (1995a), pp. 375-80.
Evans, J. (1998) The History and Practice of Ancient Astronomy. Oxford U.P.
Everson, S. (1990) Epistemology (Companions to Ancient Thought 1). Cambridge U.P.
(1991) Psychology (Companions to Ancient Thought 2). Cambridge U.P.
(1994) Language (Companions to Ancient Thought 3). Cambridge U.P.
(1998) Ethics (Companions to Ancient Thought 4). Cambridge U.P.
Fahrnkopf, R. (1977) Forms in the Philebus. Journal of the History of Philosophy
15, pp. 202-7.
Farrington, B. (1936) Science in Antiquity. Oxford U.P.
Ferguson, J. (1963) Sun, Line and Cave Again. Classical Quarterly NS 13, pp.
188-93.
Ferrari, G.R.F. (1995) Myth and Conservatism in Plato's Statesman. I n Rowe
(1995a), pp. 389-98.
Field, G.C. (1933) Plato and Natural Science. Philosophy 8, pp. 131-41.
Field, J.V. (1988) Kepler's Geometrical Cosmology. Athlone Press, London.
Findlay, J.N. (1974) Plato: The Written and Unwritten Doctrines. Humanities
Press, N.Y.
Fine, A. (1984) The Natural Ontological Attitude. I n Curd and Cover (1998).
(1996) The Shaky Game: Einstein, Realism and the Quantum Theory.
Chicago U.P.
Fine, G. (1977) Plato on Naming. Philosophical Quarterly 27, 1977, pp. 289-301.
(1978) Knowledge and Belief in Republic V. Archiv filr Geschichte der
Philosophie 60, pp. 121-39.
(1979a) False Belief in the Theaetetus. Phronesis 24, pp. 70-80.
(1979b) Knowledge and Logos in the Theaetetus, Philosophical Review 88,
pp. 366-97.
(1983) Plato and Aristotle on Form and Substance. Proceedings of the
Cambridge Philological Society 209, pp. 23-47.
(1984) Separation. Oxford Studies in Ancient Philosophy 2, pp. 31-87.
(1986) Immanence. Oxford Studies in Ancient Philosophy 4, pp. 71-98.
(1987) Forms as Causes: Plato and Aristotle. I n A. Graeser (ed.), Mathemat­
ics and Metaphysics in Aristotle, Bern pp. 69-112.
(1988a) Plato on Perception: A Reply to Professor Turnbull, 'Becoming and
Intelligibility in Plato'. Oxford Studies in Ancient Philosophy 6, pp. 15-28.
(1988b) Owen's Progress. Philosophical Review 97, pp. 373-99.
(1990) Knowledge and Belief in Republic V-VII. Companions To Ancient
Thought 1, Epistemology, ed. S. Everson, Cambridge U.P. pp. 85-115.
(1992) Inquiry in the Meno. I n Kraut, R. (1992), pp. 200-26.
Frank, E. (1923) Plato und die Sogennaten Pythagoreer. Halle an der Salle.
Frankel, H . (1942) Zeno of Elea's Attacks on Plurality. American Journal of
Philology 63, pp. 1-25 and 193-206.
Frede, D. (1989) The Soul's Silent Dialogue: A Non-Aporetic Reading of the
Theaetetus. Proceedings of the Cambridge Philological Society 215, pp. 20-49.
(1996) The Philosophical Economy of Plato's Psychology: Rationality and
Common Concepts in the Timaeus, pp. 29-58 in Frede and Striker (1996).
Frede, M . (1967) Pradikation und Existenzaussage. Vandenhoeck and Ruprecht,
Gottingen.
312 Bibliography
(1980) The Original Notion of Cause. I n J. Barnes et al. (ed.) Doubt and
Dogmatism. Oxford U.P.
(1988) Being and Becoming in Plato. Oxford Studies in Ancient Philosophy
6, pp. 37-52.
Frede, M . and Striker, G. (1996) Rationality in Greek Thought. Oxford U.P.
Friedlander, P. (1958) An Introduction to Plato. Routledge and Kegan Paul, London
and N.Y., ch. 14 pp. 246-60.
(1969) Plato, The Dialogues, 2 vols, Princeton U.P.
Furley, D.J. (1967) Two Studies in the Greek Atomists. Princeton U.P.
(1981) The Greek Theory of the Infinite Universe. Journal of the History of
Ideas 42, pp. 571-86.
(1986) The Cosmological Crisis in Classical Antiquity. Proceedings of the
Boston Area Colloquium on Ancient Philosophy 2, pp. 1-19.
(1989) Cosmic Problems. Cambridge U.P.
(1987) The Greek Cosmologists. Cambridge U.P.
Gaiser, K. (1963) Platons Ungeschriben Lehre, 2nd ed., Stuttgart U.P, 1968.
(1980) Plato's Enigmatic Lecture 'On the Good'. Phronesis 25, pp. 5-37.
(1986) Platons Zusammenschau der Mathematischen Wissenschaften, A &
A 32, pp. 89-124.
Gallop, D. (1963) Plato and the Alphabet, Philosophical Review 72, 364-76.
(1975) Plato's Phaedo. Oxford U.P.
Gentzler, J. (1991) Sumphonein in Plato's Phaedo. Phronesis 36, pp. 265-76.
Gill, C. (1979) Plato and Politics: The Critias and the Politicus. Phronesis 24, pp.
148-67.
Gill, C. and McCabe, M.M. (1996) Form and Argument in Later Plato. Oxford U.P.
Gill, M.L. (1987) Matter and Flux in Plato's Timaeus. Phronesis 32, pp. 34-53.
Gillies, D. (1998) The Duhem Thesis and the Quine Thesis. I n Curd and Cover
(1998), pp. 302-19.
Gillispie, C.C. (1970-1980) The Dictionary of Scientific Biography (ed.), 16 vols,
Scribner, N.Y.
Gisinger, F. (1921) Die Erdbeschreibung des Eudoxus von Knidos. Stoicheia, Heft
6.
Goldstein, B.R. and Bowen, A.C. (1983) ANew View of Early Greek Astronomy. Isis
74, 1983, pp. 330-40.
Goodrich, W.J. (1903) On Phaedo 96a-102a and on the Deuteros Pious 1. Classical
Review 17, pp. 5-11.
(1904) On Phaedo 96a-102a and on the Deuteros Pious 2. Classical Review
18, pp. 381-5.
Gosling, J.C.B. (1960) Republic V: ta polla kola. Phronesis 5, pp. 116-28.
(1968) Doxa and dunamis in Plato's Republic. Phronesis 13, pp. 119-30.
(1975) Plato: Philebus. Oxford U.P.
Graham, D.W. (1990) Socrates, the Craft Analogy and Science. Apeiron 23, pp. 1-24.
Gregory, A.D. (1996) Astronomy and Observation in Plato's Republic. Studies in
History and Philosophy of Science 27, pp. 451-71.
(2000) Plato and Aristotle on Eclipses. Journal for the History of Astronomy
xxxi, pp. 245-59.
Gregory, J.C. (1931) A Short History of Atomism: From Democritus to Bohr. Black.
Gribbin, J. (1986) In Search of the Big Bang. Heinemann, London.
(1993) In the Beginning. Penguin, London.
Gribbin, J and Rees, M . (1990) Cosmic Coincidences. Heinemann, London, 1990.
Grote, G. (1875) Plato and the Other Companions of Socrates. John Murray,
London.
 Bibliography 313
Gulley, N . (1954) Plato's Theory of Recollection. Classical Quarterly NS 4, pp.
194-213.
(1960) The Interpretation of Plato, Timaeus 49d-e. American Journal of
Philology 81, pp. 53-64.
(1962) Plato's Theory of Knowledge. Butler and Tanner, Frome and London.
(1968) The Philosophy of Socrates. London.
Guthrie, W.K.C. (1962-81) A History of Greek Philosophy, 6 vols., Cambridge U.P.
(1976) Rhetoric and Philosophy: The Unity of the Phaedrus. Paideia,
Special Plato Issue, pp. 117-24.
Hackforth, R. (1952) Plato's Phaedrus. Cambridge U.P.
(1955) Plato's Phaedo. Cambridge U.P.
(1959) Plato's Cosmogony. Classical Quarterly NS 9, pp. 17-22.
(1972) Plato's Philebus. Cambridge U.P.
Hahm, D.E. (1977) The Origins of Stoic Cosmology. Ohio.
Hahn, R. (1978) On Plato's Philebus 15bl-8. Phronesis 23, pp. 158-72.
(1980) Material Causality, Non-Being, and Plato's Hypodoche: A Re-View of
the Timaeus in Terms of the Divided Line. Apeiron 14, pp. 59-67.
Hall, T.S. (1969) Ideas of Life and Matter: Studies in the History of General
Physiology, 600 BC - 1900 AD, 2 vols., Chicago U.P.
Hankins, T.L. (1985) Science and the Enlightenment. Cambridge U.P.
Harman, G. (1965) Inference to the Best Explanation. Philosophical Review 74, pp.
88-95.
Harre, R. (1972) The Philosophies of Science. Oxford U.P.
(1981) Great Scientific Experiments. Oxford U.P.
(1993) Laws of Nature. Duckworth.
(1995) Realism Rescued: How Scientific Progress is Possible. Duckworth
and Open Court.
(2000a) Dispositional Powers, pp. 97-101 in Newton-Smith (2000).
(2000b) Laws of Nature, pp. 213-23 i n Newton-Smith (2000).
Hawking, S. (1988) A Brief History of Time. Guild, London.
(1993) Black Holes and Baby Universes and Other Essays. Bantam, London.
Heath, T. (1913) Aristarchos ofSamos. Dover edition, New York 1981.
(1921) A History of Greek Mathematics. Oxford U.P.
(1947) Euclid: The Elements. St. Johns College Press, Annapolis.
Henry, J. (1990) Magic and Science in the Sixteenth and Seventeenth Centuries.
I n Olby, Cantor, Christie and Hodge (1990).
Hetherington, N.S. (1993a) Cosmology: Historical, Literary, Religious and Scien­
tific Perspectives. Garland, N.Y
(1993b) Encyclopedia of Cosmology: Historical, Philosophical and Scientific
Foundations of Modern Cosmology. Garland, N.Y.
(1996) Plato and Eudoxus: Instrumentalists, Realists or Prisoners of The-
mata? Studies in History and Philosophy of Science 27, pp. 271-89.
Hicken, W.F. (1958) The Character and Provenance of Socrates' 'Dream' in the
Theaetetus, Phronesis 3, pp. 126-45.
Hirsch, U . (1997) Sinnesqualitdten und ihre Namen (zu Tim. 61-69). I n Calvo and
Brisson (1997).
Hocutt, M . (1974) Aristotle's Four Becauses. Philosophy 49, pp. 385-9.
Holland, A.J. (1973) An Argument i n Plato's Theaetetus: 184-6. Philosophical
Quarterly 23, pp. 10-116.
Hoskin, M . (1997) The Cambridge Illustrated History of Astronomy. Cambridge
U.P.
(1999) The Cambridge Concise History of Astronomy. Cambridge U.P.
314 Bibliography
Huby, P.M. (1959) Phaedo 99d-102a. Phronesis 4, pp. 12-14.
Hume, D. (1738) A Treatise on Human Nature. Everyman, London 1955.
Hunt, D.P. (1998) The 'Problem of Fire': Referring to Phenomena in Plato's Ti­
maeus. Ancient Philosophy 18, pp. 70-80.
Hutchison, K. (1982) What Happened to Occult Qualities in the Scientific Revolu­
tion. Isis 73, pp. 233-53.
Irwin, T.H. (1974) Plato's Moral Theory. Oxford U.P.
(1977) Plato's Heracleiteanism. Philosophical Quarterly-27, pp. 1-13.
(1988) Reply to David. L. Roochnik. Platonic Writings, Platonic Readings,
ed. C.L Griswold, Routledge, London and N.Y., pp. 194-9.
Joad, C.E.M. (1933) Plato's Theory of Forms and Modern Physics. Philosophy 8,
pp. 142-54.
Jones, W.H.S. (1948) Hippocrates, vol. I . Loeb, Harvard U.P.
Jordan, R.W. (1984) Plato's Arguments for Forms. Cambridge Philological Society
Supp. Vol. 9.
Jowett, B. (1953) The Dialogues of Plato, 4 vols., 4th ed., Oxford U.P.
Kahn, C.H. (1970) On Early Greek Astronomy. Journal of Hellenic Studies 90, pp.
99-116.
(1974) Pythagorean Philosophy Before Plato. I n The Pre-Socratics, ed.
A.P.D. Mourelatos, Anchor Books, Garden City.
(1981) Some Philosophical Uses of'to be' in Plato. Phronesis 26, pp. 105-34.
(1991) Some Remarks on the Origins of Greek Science and Philosophy, pp.
1-10 in Bowen (1991).
(1999) Review of Greek Philosophy from the Beginning to Plato. Oxford
Studies in Ancient Philosophy, pp. 325-41.
Keyt, D. (1961) Aristotle on Plato's Receptacle. American Journal of Philology, 82,
pp. 291-300.
(1963) The Fallacies of the Phaedo. Phronesis 8, pp. 167-71.
(1969) Plato's Paradox that the Immutable is Unknowable. Philosophical
Quarterly 19, pp. 1-14.
(1971) The Mad Craftsman of the Timaeus. Philosophical Review 80, pp. 230-5.
Kirk, G.S. (1960) Popper on Science and the Presocratics. Mind NS 69, pp. 318-19.
(1961) Sense and Common-sense i n the Development of Greek Philosophy.
Journal of Hellenic Studies 81, pp. 105-17.
Kirk, G.S., Raven, J.E. and Schofield, M . (1983) The Presocratic Philosophers. 2nd
edition, Cambridge U.P.
Klagge, J.C. (1992) Editor's Prologue, Oxford Studies in Ancient Philosophy, Supp.
Vol., Methods of Interpreting Plato and his Dialogues, pp. 1-12.
Kline, M . (1972) Mathematical Thought from Ancient toModern Times. Oxford U.P.
Knorr, W. (1982) The Interaction of Mathematics and Philosophy i n Antiquity. I n
N . Kretzman, (ed.), Infinity and Continuity in Ancient and Medieval Thought.
Ithaca 1982.
(1989) The Practical Element in Ancient Exact Sciences. Synthese 81, pp.
313-28.
(1990) Plato and Eudoxus on the Planetary Motions. Journal for the History
of Astronomy 21, pp. 313-29.
(1993) Plato's Cosmology. I n Hetherington (1993b).
Koestler, A. (1959) The Sleepwalkers: A History of Man's Changing Vision of the
Universe. Penguin.
Koyre, A. (1943) Galileo and Plato. Journal of the History of Ideas 4, pp. 400-28.
(1973) The Astronomical Revolution. Methuen, London.
Krafft, F. (1965) Der Mathematikos und der Physikos. Bemerkungen zu der Ange-
 Bibliography 315
blichen Platonischen Aufgabe, die Phanomene zu Retten. Beitrage zur
Geschichte der Wissenschaft und Technik 5, pp. 5-24.
Kramer, H.J. (1959) Arete bei Platon und Aristoteles. Heidelberg U.P.
(1964) Retraktionen zum Problem des Esoterischen Platon. Museum
Helveticum 21, pp. 137-67.
(1966) Aristoxenus liber Platons peri tagathou. Hermes 94, pp. 111-26.
Kraut, R. (1992) The Cambridge Companion to Plato. Cambridge U.P.
Kung, J. (1981) Aristotle on Thises, Suches and the Third Man Argument. Phrone­
sis 26, pp. 207-47.
(1985) Tetrahedra, Motion and Virtue. Nous 19, pp. 17-27.
(1988) Why the Receptacle is Not a Mirror. Archiv fur Geschichte der
Philosophic 70, pp. 167-78.
Kuhn, T.S. (1957) The Copernican Revolution. Random House, N.Y.
Lamb, W.R.M. (1927) Epinomis. Loeb Plato vol. X I I .
Lane, M . (1995) A New Angle on Utopia: The Political Theory of the Statesman. I n
Rowe (1995a) pp. 276-91.
Lang, H.S. (1992) Aristotle's Physics and its Medieval Varieties. SUNY.
Lasserre, F. (1964) The Birth of Mathematics in the Age of Plato. London.
(1966) Die Fragmente des Eudoxus von Knidos. De Gruyter, Berlin.
Laudan, L. (1998) Demystifying Underdetermination. I n Curd and Cover (1998)
pp. 320-53.
Ledger, G.R. (1989) A Computer Analysis of Plato's Style. Oxford U.P.
Lee, E.N. (1966) On the Metaphysics of the Image in Plato's Timaeus. Monist 50,
pp. 341-68.
(1967) On Plato's Timaeus, 49d4-e7. American Journal of Philology 88, pp.
1-28.
(1971) On the Gold Example in Plato's Timaeus. I n Essays in Ancient
Philosophy, ed. J.P. Anton and G.L. Kustas, Albany, N.Y. pp. 219-35.
(1976) Reason and Rotation. Circular Movement as the Model of Mind
(Nous) in Later Plato. I n Facets of Plato's Philosophy, Phronesis Supp. Vol. I I ,
pp. 70-102.
Lee, H.P.D. (1955) The Republic. Penguin.
(1965) Plato: Timaeus and Critias. Penguin.
Leibniz, G.W.F. (1686) Discourse on Metaphysics. I n G.H.R. Parkinson (ed.) Leib­
niz: Philosophical Writings. Dent, London, 1973.
(1687) Letter Explaining the Laws of Nature. I n G.H.R. Parkinson (ed.)
Leibniz: Philosophical Writings. Dent, London, 1973.
Lennox, J.G. (1985) Plato's Unnatural Teleology. I n D.J. O'Meara (ed.) Platonic
Investigations. Washington D.C. U.P, pp. 195-218.
Leplin, J. (2000) Realism and Instrumentalism. I n Newton-Smith (2000).
Leslie, J. (1989) Universes. Routledge, London.
Letwin, O. (1981) Interpreting the Philebus. Phronesis 26, pp. 187-206.
Liddle, A.R. (1999) An Introduction to Modern Cosmology. Wiley, John and Sons.
Lipton, P. (2000) Inference to Best Explanation. I n Newton-Smith (2000).
Lloyd, G.E.R. (1966) Polarity and Analogy. Cambridge U.P.
(1968) Plato as a Natural Scientist, Journal of Hellenic Studies 28, pp.
78-92.
(1970) Early Greek Science to Aristotle. Chatto and Windus, London.
(1978) Saving the Appearances. Classical Quarterly 88, pp. 202-22. See also
Lloyd (1991) pp. 248-53.
(1979) Magic, Reason and Experience. Cambridge U.P.
(1987) The Revolutions of Wisdom. California U.P.
316 Bibliography
(1990) Plato and Archytas in the Seventh Letter. Phronesis 35, pp. 159-74.
(1991) Methods and Problems in Greek Science, Cambridge U.P, pp. 333-51.
(1996) Aristotelian Explorations. Cambridge U.P.
Lombard, L.B. (1978) Relational Change and Relational Changes. Philosophical
Studies 34, pp. 63-79.
Long, A.A. and Sedley, D. (1987) The Hellenistic Philosophers, 2 vols, Cambridge
U.P.
Long, A.A. (1977) Chance and Law in Epicureanism. Phronesis 22, pp. 63-88.
Lovibond, S. (1991) Plato's Theory of Mind. I n Everson (1991).
McCabe/ MacKenzie, M.M. (1982a) Paradox in Plato's Phaedrus. Proceedings of the
Cambridge Philological Society, NS 28, pp. 64-76.
(1982b) Parmenides' Dilemma. Phronesis 28, pp. 1-13.
(1986a) Putting the Cratylus in its Place. Classical Quarterly 36, pp.
124-50.
(1986b) Impasse and Explanation: from the Lysis to the Phaedo. Archiv fur
Geschichte der Philosophie 88, pp. 15-45.
(1988) The Virtues of Socratic Ignorance. Classical Quarterly 38, pp.
331-50.
(1993) Myth, Allegory and Argument. Apeiron 25, The Language of the
Cave, pp. 47-67.
(1994) Plato's Individuals. Princeton U.P.
(1997) Chaos and Control: Reading Plato's Politicus. Phronesis 42, pp.
94-117.
McDowell, J.H. (1973) Plato: Theaetetus. Oxford U.P.
MacKenzie, see McCabe.
Malcolm, J. (1962) The Line and the Cave. Phronesis 7, pp. 38-45.
(1967) Plato's Analysis of'to on' and 'to me on' in the Sophist. Phronesis 12,
pp. 130-46.
Martin, A. and Primaversi, O. (1999) U Empedocle de Strasbourg. Bibliotheque
Nationale et Universitaire de Strasbourg/Berlin. De Gruyter.
Mellor, D.H. (1980) Necessities and Universals in Natural Laws. I n Mellor (ed.),
Science, Belief and Behaviour: Essays in Honour of R.B. Braithwaite. Cam­
bridge U.P, pp. 105-25.
Mendell, H . (1998) Reflections on Eudoxus, Callippus and their Curves: Hip-
popedes and Callippopedes. Centaurus 40, pp. 177-275.
Milton, J.R. (1998) Laws of Nature. I n Ayers and Garber.
Mittlestrass, J. (1962) Die Rettung der Phanomene. De Gruyter, Berlin.
(1988) On Socratic Dialogue. I n Platonic Writings, Platonic Readings, ed.
C.L Griswold, Routledge, London and N.Y, pp. 126-42.
Modrak, D.K. (1981) Perception and Judgement in the Theaetetus. Phronesis 26,
pp. 35-54.
Mohr, R.D. (1977) Plato, Statesman 284c-d: An Argument from the Sciences.
Phronesis 22, pp. 232-4.
(1978a) The Formation of the Cosmos in the Statesman Myth. Phoenix 27,
pp. 250-2.
(1978b) The Gold Analogy in Plato's Timaeus (50a4-b5). Phronesis 23, pp.
243-52.
(1980a) Image, Flux and Space in Plato's Timaeus. Phoenix 34, pp. 138-52.
(1980b) The Sources of Evil Problem and the arche kineseos Doctrine in
Plato. Apeiron 14, pp. 41-56.
(1980c) The Sources of Flux in Plato's Timaeus. Apeiron 14, pp. 96-114.
(1981) Disorderly Motion in Plato's Statesman. Phoenix 35, pp. 199-215.
 Bibliography 317
(1983) Philebus 55c-62a and Revisionism. Canadian Journal of Philosophy
9, Supp. Vol., pp. 165-70.
(1985) The Platonic Cosmology. Brill, Leiden.
(1985a) Plato's Theology Reconsidered: What the Demiurge Does. History
of Philosophy Quarterly 2, pp. 131-44.
(1986) Plato on Time and Eternity. Ancient Philosophy 6, pp. 39-46.
Moravscik, J. (1974) On Explanation. Synthese 28, pp. 22-5.
Morgan, M.L. (1984) Sense-Perception and Recollection in the Phaedo. Phronesis
29, pp. 237-51.
(1989) How Does Socrates Solve the Paradox of Inquiry in the Meno? Essays
in Ancient Greek Philosophy III, Plato, ed. J.P. Anton and A. Preus, New York
U.P, pp. 169-82.
Morrison, J.S. (1959) The Shape of the Earth in Plato's Phaedo. Phronesis 4, pp.
101-19.
(1977) Two Unresolved Difficulties in the Line and Cave. Phronesis 22, pp.
212-31.
Morrison, J.S. and Coates, J.F. (1986) The Athenian Trireme: The History and
Reconstruction of an Ancient Greek Warship. Cambridge U.P.
Morrison, J.S. and Williams, R.T. (1968) Greek Oared Ships, 900-322 BC, Cam­
bridge U.P.
Morrow, G.R. (1950) Necessity and Persuasion in Plato's Timaeus. I n Studies in
Plato's Metaphysics, ed. R.E. Allen, Routledge and Kegan Paul, London 1965,
pp. 421-38.
(1962) Plato's Epistles. Bobbs-Merrill, Indianapolis and N.Y
(1970) Plato and the Mathematicians: An Interpretation of Socrates' Dream
in the Theaetetus. Philosophical Review 79, pp. 309-33.
(1968) Plato's Theory of the Primary Bodies in the Timaeus and the Later
Doctrine of Forms. Archiv fur Geschichte der Philosophic 60, pp. 12-28.
Mourelatos, A.P.D. (1981a) Plato's 'Real Astronomy': Republic 527d-531d. I n Sci­
ence and the Sciences in Plato, ed. P. Anton, Buffalo N.Y, pp. 33-73.
(1981b) Astronomy and Kinematics in Plato's Project of Rationalist Expla­
nation. Studies in the History and Philosophy of Science 12, pp. 1-32.
(1991) Plato's Science - His View and Ours of His, pp. 11-42, in Bo wen
(1991).
Mueller, I . (1981) Ascending to Problems: Astronomy and Harmonics i n Republic
V I I . Science and the Sciences in Plato, ed. P. Anton, Buffalo N.Y, pp. 103-22.
(1989) Joan Kung's Reading of the Timaeus. Apeiron 22, pp. 1-25.
(1992) Mathematical Method and Philosophical Truth. I n Kraut (1992).
Murphy, N.R. (1936) The Deuteros Pious in the Phaedo. Classical Quarterly 30, pp.
40-7.
Musgrave, A. (1985) Realism versus Constructive Empiricism. I n Curd and Cover
(1998).
(1989) NOA's Ark - Fine for Realism. I n Curd and Cover (1998).
(1991) The Myth of Astronomical Instrumentalism. I n G. Munevar (ed.)
Beyond Reason: Essays on the Philosophy of Paul Feyerabend. Klusener Aca­
demic, Dordrecht.
Naddaf, G. (1997) Plato and the Peri Phuseos Tradition. I n Calvo and Brisson
(1997).
Nadler, S. (1998) Doctrines of Explanation in Late Scholasticism and in the
Mechanical Philosophy. I n Ayers and Garber (1998).
Nahnikian, G. (1955) Plato's Theory of Sensation. Review of Metaphysics 9, pp.
129-48.
318 Bibliography
Nehamas, A. (1975) Plato on the Imperfection of the Sensible World. American
Philosophical Quarterly 12, pp. 105-17.
(1984) Episteme and Logos in Plato's Later Thought. Archiv fur Geschichte
der Philosophic 66, pp. 11-36.
(1985) Meno's Paradox and Socrates as a Teacher. Oxford Studies in Ancient
Philosophy 3, pp. 1-30.
Neugebauer, O. (1957) The Exact Sciences in Antiquity. 2nd edition, Providence.
(1975) A History of Ancient Mathematical Astronomy. Springer-Verlag,
Berlin.
Newton, I . (1687) Principia Mathematica Philosophia Naturalis. Third edition
(1726) with variant readings assembled and edited by A. Koyre and LB. Cohen,
Cambridge U.P, 1972.
(1704) Opticks. London, 1730.
Newton-Smith, W. (1980) The Underdetermination of Theories by Data. I n N.R.
Hilpinen (ed.), Rationality in Science. Dordrecht, Reidel.
(1981) The Rationality of Science. Routledge, London.
(2000) A Companion to the Philosophy of Science. Blackwell, Massachu­
setts.
Nikolic, M . (1974) Eudoxus and Dedekind. Boston Studies in the History and
Philosophy of Science 15.
North, J. (1994) The History of Astronomy and Cosmology. Fontana.
O'Brien, D. (1968) Derived Light and Eclipses i n the Fifth Century. Journal of
Hellenic Studies 88, pp. 114-327.
(1984) Theories of Weight in the Ancient World, vol. I I , Plato: Weight and
Sensation. Brill.
(1997) Perception et intelligence dans la Timee de Platon. I n Calvo and
Brisson (1997).
Olby, R.C., Cantor, G.N., Christie, J.R.R. and Hodge, M.J.S. (1990) Companion to
the History of Modern Science. Routledge, London.
Olsen, R. (1982) Science Deified and Science Defied, 2 vols., California U.P.
Oppenheim, A.L. (1978) Man and Nature in Mesopotamian Civilisation. I n Gil-
lispie vol. 15, 1978.
Osborne, C. (1988) Topography in the Timaeus: Plato and Augustine on Mankind's
Place in the Natural World. Proceedings of the Cambridge Philological Society
214, pp. 104-14.
(1996) Space, Time, Shape and Direction: Creative Discourse in the Timaeus.
In Gill & McCabe (1996), pp. 179-211.
Owen, G.E.L. (1953) The Place of the Timaeus in Plato's Dialogues. Reprinted in
R.E. Allen (ed.), Studies in Plato's Metaphysics, Routledge and Kegan Paul,
1965, pp. 313-39.
(1957a) A Proof in the Peri Ideon. Reprinted in Logic, Science and Dialectic,
ed. M . Nussbaum, Duckworth, London, 1986, pp. 165-79.
(1957b) Logic and Metaphysics in Some Earlier Works of Aristotle. Re­
printed in Logic, Science and Dialectic, ed. M . Nussbaum, Duckworth, London,
1986, pp. 180-99.
(1966a) The Platonism of Aristotle. Reprinted in Logic, Science and Dialec­
tic, ed. M . Nussbaum, Duckworth, London, 1986, pp. 200-20.
(1966b) Plato and Parmenides on the Timeless Present. The Monist 50, pp.
317-40.
(1970) Notes on Ryle's Plato. Reprinted in Logic, Science and Dialectic, ed.
M . Nussbaum, Duckworth, London, 1986, pp. 85-103.
 Bibliography 319
(1971) Plato on Not-Being. I n Logic, Science and Dialectic, ed. M . Nuss-
baum, Duckworth, London, 1986, pp. 104-37.
Pais, A. (1982) Subtle is the Lord. Oxford U.P.
Palter, R. (1970) An Approach to the Early History of Astronomy. Studies in History
and Philosophy of Science 1, pp. 93-133.
Panchenko, D. (1994) Homoios and Homoiotes in Thales and Anaximander. Hy-
perorboreus 1, pp. 28-55.
Pangle, T.L. (1980) The Laws of Plato. Chicago U.P.
Pannekoek, A. (1961) A History of Astronomy. George Allen and Unwin.
Parry, R.D. (1979) The Unique Worlds of the Timaeus. Journal of the History of
Philosophy 17, pp. 1-10.
(1991) The Intelligible World Animal in Plato's Timaeus, Journal of the
History of Philosophy 29, pp. 13-32.
Parry, R.D. (1979). The Unique World of the Timaeus. Journal for the History of
Philosophy 17, pp. 1-10.
(1983) The Craft of Justice. Canadian Journal of Philosophy 9, pp. 19-38.
(1991) The Intelligible World-Animal in Plato's Timaeus. Journal for the
History of Philosophy 29, pp. 13-32.
(1996) Plato's Craft of Justice. SUNY, Albany.
Patterson, R. (1981) The Unique Worlds of the Timaeus. Phoenix 35, pp. 105-19.
(1985) On the Eternality of Platonic Forms. Archiv fur Geschichte der
Philosophic 67, pp. 27-45.
Pedersen, O. (1974) Early Physics and Astronomy,, 2nd edition 1993, Cambridge
U.P.
Penner, T. (1987) The Ascent from Nominalism, Dordrecht.
Philip, A. (1981) Recurrences thematiques et topologie dans le Thedre'. Revue de
Metaphysique et Morale 86, pp. 452-76.
Pingree, D. (1992) Hellenophilia vs. the History of Science. Isis 83, pp. 554-63.
Plass, P. (1960) Socrates' Method of Hypothesis in the Phaedo. Phronesis 5, pp.
103-15.
Pohle, W. (1971) The Mathematical Foundations of Plato's Atomic Physics. Isis 62,
pp. 36-46.
(1973) Dimensional Concepts in Plato's Physics. In Exegesis and Argument,
ed. Lee, Mourelatos and Rorty, Assen 1973, pp. 306-23.
Popper, K.R. (1945) The Open Society and its Enemies, vol. I , The Spell of Plato,
5th edition, Routledge and Kegan Paul, London 1966.
(1963) Conjectures and Refutations, Routledge and Kegan Paul, London.
Post, L.A. (1963) Letter V I I . I n E. Hamilton and H . Cairns, Plato: The Collected
Dialogues. Princeton U.P.
Prior, W.J. (1985) Unity and Development in Plato's Metaphysics. Croom Helm,
Beckenham.
Proclus. (1966) Commentaire sur le Timee, 5 vols, ed. A.J. Festugiere, Libraire
Philosophique, J. Vrin, Paris.
Ptolemy. (1898) Suntaxis Mathematica, 2 vols, ed. J.L. Heiberg, Teubner Series.
(1980) Tetrabiblos, ed. & trans. EE. Robbins, Loeb Series, Harvard U.P.
(1907) Planetary Hypotheses - Opera Astronomica Minora, ed. J.L. Heiberg,
Teubner Series.
The Arabic Version of Ptolemy's Planetary Hypotheses, ed. B.R. Goldstein,
Transactions of the American Philosophical Society, NS 57, 1967, pp. 3-55.
Pyle, A. (1997) Atomism and its Critics. Thoemmes, Bristol.
Quine, W.V.O. (1953) Two Dogmas of Empiricism. I n From a Logical Point of View.
Harper.
320 Bibliography
(1960) Word and Object. Cambridge, Massachusetts, M.I.T. Press.
(1969) Ontological Relativity and Other Essays, New York, Columbia U.P.
(1975) On Empirically Equivalent Systems of the World, Erkenntnis 9, pp.
313-28.
(1986) The Philosophy of W.V.O. Quine. Open Court.
Raven, J.E. (1953) Sun, Divided Line and Cave. Classical Quarterly NS 3, pp.
131-87.
Ravetz, J. (1990) The Copernican Revolution. I n Olby, Cantor, Christie and Hodge
(1990).
Reale, G. (1997) Plato's Doctrine of the Origin of the World, with Special Reference
to the Timaeus. I n Calvo and Brisson (1997).
Reed, N . H . (1972) Plato on Flux, Perception and Language. Proceedings of the
Cambridge Philological Society 18, pp. 65-77.
Reeve, C.D.C. Motion, Rest and Dialectic in the Sophist. Archiv fur Geschishte der
Philosophie 67, pp. 47-64.
Reydam-Schils, G. (1997) Grasping Sensibles Without Sense-Perception. I n Calvo
and Brisson (1997).
Richardson, H . (1926) The Myth of Er. Classical Quarterly 20, pp. 113-33.
Ricker, R.A.R. (1967) The Paths of the Planets. London.
Ritter, C. (1910) Platon, seine Leben, seine Schriften, seine Lehre. Munich U.P.
Robin, L. (1908) La Theorie Platonicienne des Idees et des Nombres d'apresAristote.
Presses Universitaires de France, Paris 1908.
Robins, I . (1995) Mathematics and the Conversion of the Mind, Republic V I I
522cl-531e3, Ancient Philosophy 15.
Robinson, R. (1953) Plato's Earlier Dialectic. 2nd edition, Oxford U.P, 1984,
Introduction, pp. 1-6.
Robinson, T.M. (1970) Plato's Psychology. Toronto U.P.
(1979) The Argument of Timaeus 27d ff. Phronesis 24, pp. 105-9.
Roochnik, D. (1986) Socrates' Use of the Techne Analogy. Journal of the History of
Philosophy 24, pp. 295-310.
(1987) Plato's Use of Atechnos. Phoenix 41, pp. 255-63.
(1994) Counting on Number: Plato on the Goodness of Arithmos. American
Journal of Philology 115, pp. 543-63.
(1996) Of Art and Wisdom: Plato's Understanding of Techne. Pennsylvania
State U.P.
Rose, L.E. (1966) The Deuteros Pious in Plato's Phaedo. Monist 50, pp. 464-73.
Rosen, S. (1979) Plato's Myth of the Reversed Cosmos. Review of Metaphysics 33,
pp. 60-85.
(1983) Plato's Sophist. Yale U.P.
Rosenmayer, T.G. (1959) The Shape of the Earth in the Phaedo: A Rejoinder.
Phronesis, pp. 71-2.
(1960) Plato's Hypothesis and the Upward Path. American Journal of
Philology 81, pp. 355-66.
Ross, W.D. (1951) Plato's Theory of Ideas. Oxford U.P.
Rowan-Robinson, M. (1996) Cosmology, 3rd edition, Oxford U.P.
Rowe, C.J. (1993a) Explanation in Phaedo 99c6-102a8. Oxford Studies in Ancient
Philosophy 11, pp. 49-69.
(1993b) Plato: Phaedo. Cambridge U.P.
(1995a) Reading the Statesman. Proceedings of the HI Symposium Platoni-
cum. Academia Verlag, Sankt Augustin.
(1995b) Plato: Statesman. Aris and Phillips, Warminster.
 Bibliography 321
Runia, D. (1997) The Literary and Philosophical Status of Timaeus' Proemium. I n
Calvo and Brisson (1997).
Runciman, W.G. (1962) Plato's Later Epistemology. Cambridge U.P.
Ryle, G. (1960) Letters and Syllables in Plato. Philosophical Review 69, pp. 431-51.
(1966) Plato's Progress. Cambridge U.P.
(1990) Logical Atomism in Plato's Theaetetus. Phronesis 35, pp. 21-46.
Sambursky, S. (1956) The Physical World of the Greeks. Routledge and Kegan Paul,
London.
(1965) Plato, Proclus, and the Limits of Science. Journal of the History of
Philosophy 03, pp. 1-11.
Santa Cruz, M . I . (1997) Le Discours de la Physique: eikos logos. I n Calvo and
Brisson (1997).
Saunders, T.J. (1970) Plato: The Laws. Penguin, London.
Sayre, K.M. (1969) Plato's Analytic Method. Chicago U.P.
(1983) Plato's Late Ontology: A Riddle Resolved. Princeton U.P.
(1992) A Maieutic View of Five Late Dialogues. Oxford Studies in Ancient
Philosophy, Supp. Vol, Methods of Interpreting Plato and His Dialogues, pp.
221-44.
(1998) The Role of the Timaeus in the Development of Plato's Late Ontology.
Ancient Philosophy 18, pp. 93-124.
Schiaperelli (1875) Le Sfere Omocentriche di Eudosso, di Callippo et di Aristotele.
Publicazioni del R. Observatorio di Brera in Milano IX.
Schicker, R. (1995) Aspekte der Rezeption des Politikos im Mittel- und Neuplaton-
ismus. I n Rowe (1995a), pp. 381-8.
Schleiermacher, F. (1804) Introductions to Plato's Dialogues, trans. W. Dobson,
New York 1973.
Schofield, M . (1997) The Ionians. I n Taylor (1997).
Schoos, D.J. (1999) Timaeus' Banquet. Ancient Philosophy 19, pp. 97-107.
Schulz, D.J. (1966) Das Problem der Materie in Platons Timaios. Abhandlungen
zur Philosophic, Psychologie und Pddogogik 31, Bonn.
Schulz, D.J. (1966) Das Problem der Materie in Platons Timaios. Bonn.
Scott, D. (1987) Platonic Anamnesis Revisited. Classical Quarterly 37, pp. 346-66.
Scolnikov, S. (1997) Freedom and Education in Plato's Timaeus. I n Calvo and
Brisson (1997).
Sedley, D. (1982) Two Conceptions of Vacuum. Phronesis 27 pp. 175-93.
(1989) Teleology and Myth in the Phaedo. Proceedings of the Boston
Colloquium on Ancient Philosophy 5, pp. 359-83.
(1991) Is Aristotle's Teleology Anthropocentric? Phronesis 36, pp. 179-96.
(1997) Becoming Like God in the Timaeus and Aristotle. I n Calvo and
Brisson (1997), pp. 327-40.
(1998) Platonic Causes. Phronesis 43, pp. 114-32.
Shiner, R. (1974) Knowledge and Reality in Plato's Philebus. Assen.
Shipton, K.M.W. (1979) A Good Second-best: Phaedo 99b ff. Phronesis 24, pp. 33-53.
Shoemaker, S. (1984) Identity, Cause and Mind, Cambridge U.P.
Shorey, P. (1927) Platonism and the History of Science. Proceedings of the Ameri­
can Philosophical Association 46, pp. 159-82.
(1935) The Republic, 2 vols, Loeb Classical Library, Harvard U.P.
Silk, J. (1989) The Big Bang, 2nd edition, Freeman.
Silverman, A. (1990) Plato on Perception and 'Commons'. Classical Quarterly 40,
pp. 148-50.
(1992) Timaean Particulars. Classical Quarterly 42, pp. 87-113.
Simplicius. Aristotelis in De Caelo Commentaria, ed J.L. Heiberg, Berlin 1894.
322 Bibliography
Skemp, J.B. (1942) The Theory of Motion in Plato's Later Dialogues, 2nd edition,
Adolf M . Hakkert, Amsterdam, 1967.
(1952) Plato: The Statesman, 2nd edition, Bristol Classical Press, Bristol
1987.
Smoot, G. (1993) Wrinkles in Time. Little, Brown and Company.
Sober, E. (2000) Simplicity. I n Newton-Smith (2000), pp. 433-41.
Solmsen, F. (1940) Plato and the Unity of Science. Philosophical Review 44, pp.
566-71.
(1960) Aristotle's System of the Physical World. Cornell U.P.
(1963) Nature as Craftsman in Greek Thought. Journal of the History of
Ideas 24, pp. 473-96.
(1971) The Tradition about Zeno of Elea Re-examined. Phronesis 16, pp.
116-41.
Sorabji, R. (1980) Necessity, Cause and Blame. Perspectives on Aristotle's Theory.
Duckworth, London.
(1983) Time, Creation and the Continuum: Theories in Antiquity and the
Early Middle Ages. Duckworth, London.
(1988) Matter, Space and Motion: Theories in Antiquity and their Sequel.
Duckworth, London.
Stenzel, J. (1959) Zahl und Gestalt bei Plato und Aristoteles. Herman Gentner
Verlag, Bad Homburg vor der Hohe.
Stewart, J.A. (1960) The Myths of Plato. London.
Stocks, J.L. (1911) The Divided Line of Plato Rep. V I . Classical Quarterly 5, pp.
73-88.
Stokes, M.C. (1992) Plato and the Sightlovers of the Republic. I n The Language of
the Cave, ed. A. Barker and M . Warner, Apeiron 25, pp. 103-33.
Stough, C.L. (1976) Forms and Explanation in the Phaedo. Phronesis 21, pp. 1-30.
Strang, C. (1956) Plato and the Third Man. Proceedings of the Aristotelian Society,
Supp. Vol. 37, 1963, pp. 147-63.
(1986) Plato's Analogy of the Cave. Oxford Studies in Ancient Philosophy 4,
pp. 19-34.
Striker, G. (1970) Peras und Apeiron: Das Problem derFormen in Platons Philebus.
Vandenhoeck and Ruprecht, Gottingen.
Szlezak, T.A. (1979) The Acquiring of Philosophical Knowledge According to Plato's
Seventh Letter. In Arktouros, Hellenic Studies Presented toB.M.W. Knox, Berlin
U.P, pp. 354-63.
Taran, L. (1971) The Creation Myth in Plato's Timaeus. I n Essays in Ancient Greek
Philosophy, ed. J.P Anton and G.L Kustas, pp. 372-406.
(1975) Academica: Plato, Philip of Opus, and the Pseudo-Platonic Epino­
mis. American Philosophical Society, Philadelphia.
Taub, L. (1993) Ptolemy's Universe. Open Court.
Taylor, A.E. (1926) Plato: The Man and his Work. Meridian Books, Cleveland 1956.
(1928) A Commentary on Plato's Timaeus. Oxford U.P.
(1939) The Decline and Fall of the State i n Republic, V I I I . Mind, NS 48, pp.
23-38.
(1956) Plato: Philebus and Epinomis. Nelson and Sons, London.
Taylor, C.C.W. (1969) Forms as Causes in the Phaedo. Mind 78, pp. 45-59.
(1997) From the Beginning to Plato. Routledge History of Philosophy, vol. 1.
Routledge, London.
(1998) Plato's Ethics. I n Everson (1998).
Teloh, H . (1981) The Development of Plato's Metaphysics. University Park.
 Bibliography 323
Theiler, W. (1925) Zur Geschichte der Teleologischen Naturbetrachtung bis auf
Aristote. Zurich.
Thesleff, H . (1982) Studies in Platonic Chronology. Societas Humanarum Litter-
arum, Helsinki.
Toulmin, S. and Goodfleld, J. (1961) The Fabric of the Heavens. Hutchison, London.
Turbayne, C M . (1976) Plato's Fantastic Appendix: The Procreation Model of the
Timaeus. Paideia 50.
Turnbull, R.G. (1955) Aristotle's Debt to the Natural Philosophy of the Phaedo.
Philosophical Quarterly 8, pp. 131-43.
(1978) Knowledge and the Forms in the Later Platonic Dialogues. Proceed­
ings and Addresses of the American Philosophical Association 51, pp. 735-58.
(1988) Becoming and Intellegibility. Oxford Studies in Ancient Philosophy
6, pp. 1-14; Response to Professor Fine's Critique of'Becoming and Intelligibil­
ity'. Oxford Studies in Ancient Philosophy 6, pp. 29-36.
Vallejo, A. (1997) No, It's Not a Fiction. I n Calvo and Brisson (1997).
Van der Waerden, B.L. (1951) Die Pythagoreer. Amsterdam.
(1974a) The Earliest Form of the Epicycle Theory, Journal for the History
of Astronomy 5, pp. 175-85.
(1974b) Science Awakening I I . Oxford U.P.
(1978) Basic Ideas and Methods of Babylonian and Greek Astronomy. I n
Gillispie vol. 15, 1978.
(1982) The Motion of Venus, Mercury and the Sun in Early Greek Astro­
nomy. Archive for the History of the Exact Sciences 26, pp. 99-113.
Van Eck, J. (1994) Skopein en Logois. Ancient Philosophy 20, pp. 19-40.
Van Fraassen, B. (1980) The Scientific Image. Oxford U.P.
(1989) Laws and Symmetry. Oxford U.P.
Vartanian, A. (1957) Trembly's Polyp, La Mettrie and Eighteenth-Century French
Materialism. I n P.P. Wiener and A. Noland (eds.), Roots of Scientific Thought,
New York U.P.
Vidal-Naquet, P. (1978) Plato's Myth of the Statesman, the Ambiguities of the
Golden Age and History. Journal of Hellenic Studies 98, pp. 132-41.
Vlastos, G.L. (1939) The Disorderly Motion in the Timaeus. I n Studies in Plato's
Metaphysics, ed. R.E. Allen, Routledge and Kegan Paul, London, 1965, pp.
379-420.
(1950) The Physical Theory of Anaxagoras. Philosophical Review 59.
(1964) Creation in the Timaeus: Is it a Fiction? In Studies in Plato's Metaphys­
ics, ed. R.E. Allen, Routledge and Kegan Paul, London 1965, pp. 401-19.
(1965a) Anamnesis in the Meno. Dialogue 4, pp. 143-67.
(1965b) A Metaphysical Paradox. Reprinted in Platonic Studies, 2nd edi­
tion, Princeton U.P, 1981, pp. 43-57.
(1967) Plato's Supposed Theory of Irregular Atomic Figures. Isis 57, pp
204-9.
(1969) Reasons and Causes in the Phaedo. Reprinted with revisions in
Platonic Studies, 2nd edition, Princeton U.P. 1981, pp. 76-110.
(1972) The Unity of the Virtues in the Protagoras. Reprinted with revisions
in Platonic Studies, 2nd edition, Princeton U.P, 1981, pp. 415-58.
(1975) Plato's Universe. Clarendon, Oxford.
(1981) The Role of Observation in Plato's Conception of Astronomy. I n
Science and the Sciences in Plato, ed. P. Anton, Buffalo N.Y, pp. 1-31.
(1985) 'Socrates" Disavowal of Knowledge. Philosophical Quarterly 35, pp.
1-31.
(1991) Socrates. Cambridge U.P.
324 Bibliography
von Fritz, K. (1971) Grundproblem der Geschichte der antiken Wissenschaft. De
Gruyter, Berlin.
Von Leyden, W. (1964) Time, Number and Eternity in Plato and Aristotle. Philo­
sophical Quarterly 14, pp. 35-52.
Von Staden, H . (1992) Affinities and Elisions. Isis 83, pp. 578-95.
Wasserstein, A. (1962) Greek Scientific Thought. Proceedings of the Cambridge
Philological Society, NS 8, pp. 51-63.
Waterfield, R.A.H. (1980) The Place of the Philebus in Plato's Dialogues. Phronesis
25, pp. 270-305.
Webster, C. (1982) From Paracelsus to Newton. Cambridge U.P.
Wedberg, A. (1955) Plato's Philosophy of Mathematics. Stockholm.
Weinberg, S. (1983) The First Three Minutes, 2nd edition. Fontana, London.
Westerink, L.G. (1976) The Greek Commentaries on Plato's Phaedo, 2 vols, Amster­
dam.
White, N . (1974) Inquiry. Review of Metaphysics 28, 1974/75, pp. 289-310.
(1976) Plato on Knowledge and Reality. Hackett, Indianapolis.
Whitehead, A.N. (1933) Adventures of Ideas. The Free Press, N.Y.
(1941) Mathematics and the Good. I n P. Schilpp (ed.), The Philosophy of
Alfred North Whitehead, Open Court.
Wicksteed, P H . and Cornford, F.M. (1929) Aristotle: Physics, 2 vols, Loeb Classical
Library, Harvard U.P.
Wieland, W. (1982) Platon und die Formen des Wissens, Gottingen, pp. 13-94.
Wiggins, D. (1986) Teleology and the Good in Plato's Phaedo. Oxford Studies in
Ancient Philosophy 4, pp. 1-18.
Wightman, W.P.D. (1950) The Growth of Scientific Ideas. Edinburgh.
Williamson, H . (1904) The Phaedo of Plato. Macmillan and Co, London.
Wittaker, J. (1968) The 'Eternity' of the Platonic Forms. Phronesis 13, pp. 131-44.
Wolfsdorf, D. (1999) Plato and the Mouth-Piece Theory. Ancient Philosophy 19, pp.
13-24.
Worrall, J. (1982) Scientific Realism and Scientific Change. Philosophical Quar­
terly 32, pp. 210-31.
Wright, L. (1974) The Astronomy of Eudoxus: Geometry or Physics? Studies in
History and Philosophy of Science 4, pp. 165-72.
(1976) Teleological Explanations. Berkeley U.P.
Wright, M.R. (1995) Cosmology in Antiquity. Routledge, London.
Yavetz, I . (1998) On the Homocentric Spheres of Eudoxus. Archive for the History
of the Exact Sciences 52, pp. 221-78.
Yoh, M . (1975) On the Third Attempted Definition of Knowledge, Theaetetus
201c-210b, Dialogue 14, 420-42.
Yolton, J. (1949) The Ontological Status of Sense Data in Plato's Theory of
Perception. Review of Metaphysics 3, pp. 21-58.
Zembaty, J.S. (1983) Plato's Timaeus. Mass Terms, Sortal Terms and Identity
Through Time in the Phenomenal World. Canadian Journal of Philosophy,
Supp. Vol. 9.
Zeyl, D. (1975) Plato and Talk of a World of Flux: Timaeus 49a6-50b5. Harvard
Studies in Classical Philology 79, pp. 125-48.
(2000) Timaeus. Hackett.
Zhmud, L. (1994) Die Beziehungen zwischen Philosophie und Wissenschaft in der
Antike, Sudhoffs Archiv 78, pp. 1-13.
(1997) Wissenschaft, Philosophie und Religion im Friihen Pythagoreismus,
Berlin.
(1999) Plato as 'Architect of Science'. Phronesis 43, pp. 211-44.
 Index Locorum

References in bold are to the works cited; references in

italics are to the pages of the Notes section.

AELIAN De Partibus Animalium 644b29,

Nat. Anim. XVI, 29, 37 295, 296
Metaphysics 984bl5, 283; 985b4,
AETIUS 297; 985b31, 302; 987a32, 306;
I, 26, 2, 276; I I , 7,1, 160; I I , 7, 2, 296; 987bll, 302; 1000M2, 38;
II, 7, 7, 289; I I I , 14,1, 289; V, 18, 1078bll, 306; 1078bl6, 304;
1, 277; V, 19, 5, 37 1086M0,299
Meteorologica 343a5, 289; 343b,
ANAXAGORAS 142-3; 374a3, 281; 375al 281
Fr.2, 29; Fr.4, 29; Fr.12, 27; Fr.13, 27; Physics 187a23,278; 196a24, 26;
Fr.78,289 198b29, 37; 203b25, 277; 209M3,
302; 209b23, 302; 211a36, 302;
ARISTOTLE 218b,286; 219b, 286
De Anima 404M6, 304; 404b23,299; Protrepticus F r . l l , 287
404b26, 306; 406b26, 247, 304
De Caelo 270bll, 281; 270M3, 280; ARISTOXENUS
292a, 143; 292a8,280; 293al5, Elements of Harmony II, 30-1,284
293; 293al8, 289; 293b30, 293;
294a28, 160; 294M3, 160; 294M5, DEMOCRITUS
283; 295a, 218; 295M0, 161; Fr.2, 26-7; Fr.164, 276, 296
299M5, 301; 299b25, 237; 302bl0,
278; 303a, 278; 305b29, 302; 306a, DIOGENES LAERTIUS
202,298, 299; 306a9, 304; 306b, II, 8-9, 276; IX, 31, 25, 27, 276, 296;
202, 300; 306bl5, 302 IX, 45, 27,276
De Generatione Animalium 763b,
300 EMPEDOCLES
De Generatione et Corruptione Fr.31, 294; Fr.38, 218; Fr.57, 37;
314a8,278; 315b25, 202; 316a, Fr.59, 37; Fr.61, 37; Fr.98, 37
202; 325b20, 40; 325b25, 203;
325b30, 217; 329al4, 302; 329a21, EUSEBIUS
302; 332a27,298 Preparatio Evangelica XIV, 23, 3,
39
326 Index Locorum
GEMINUS 889b, 20, 115; 893b, 96, 123, 109;
Isagoge I,19-21, 285 895c, 169; 897b, 24, 169, 280, 287;
Elementa Astronomiae 10.2-21, 291 898a, 95, 184, 186; 898b, 96; 938a,
295; 945c, 293; 945e, 148; 966e,
HIPPOLYTUS 123, 249; 967a, 110, 287; 967b, 167
Refutatio I, 6, 3, 293; I, 8, 3-10, 276, Meno 81d, 253; 82b, 28, 302
294, 296; I, 13, 2, 27; VII, 29, 13, Parmenides 133b-134e, 252; 136a,
294 283; 136b, 283; 137b, 257; 142b3,
283; 156d, 299
LUCRETIUS Phaedo 63e, 280; 72e-77a, 283;
2.11 72-4, 286 73b-75c, 245; 74a, 65; 90d, 253;
91e-92d, 283; 96a, 83, 275-276;
MENANDER 96b, 83, 85; 96d, 83, 304; 97d, 85,
Fr.21, 283 86; 98a, 148; 98b, 83, 276; 98c,
276; 99a, 276; 99b, 86; 99c, 84, 86;
PARMENIDES 99e, 86; 100a, 87, 88; 100b, 84;
F r . l , 304; Fr.8, 160, 252, 304; Fr.10, 100c, 276; lOOe, 304; 101b, 276;
160; Fr.12, 160 lOlcd, 280; lOld, 88, 208; lOlde,
278; 102b-105b, 174; 103d, 295;
PLATO 104a, 295; 108d, 304; 108e, 85, 86,
Apology 26cd, 276, 294; 26de, 296 161; 109a, 287; 109b, 66; 110a,
Cratylus 439-440, 287; 439b, 298; 103; 114d, 304
439d, 191 Phaedrus 229bc, 305; 229c, 305;
Epinomis 976b, 284; 978b, 249, 304; 229d, 305; 230a, 258; 245c, 294;
978d, 147, 148; 978e, 142; 982b, 246a, 258; 260e, 284; 268a-269a,
294; 982b-d, 287; 982c, 139; 982d, 51; 268c, 281; 268de, 287; 268e,
169; 983c, 111, 287; 986e, 62, 137, 53; 270b, 171, 284; 274a, 305;
295; 987a, 287; 987b, 289; 990a, 275bc, 305; 275d, 255, 258; 275e,
62; 990b, 137, 148; 990d, 303; 304; 276b, 256, 257, 305; 276d,
991c, 107, 287; 991d, 89, 292; 256; 276e, 257; 277e, 257;
991e, 96 277e-278a, 256; 278a, 257
Episf/e V / / 342b, 50; 342, 279; 343a, Philebus l i e , 304; 14c, 304; 15b,
50; 343d, 50 281, 304; 17a, 212; 17e, 42, 284;
Gorgias 462c, 284; 465a, 51, 305; 18b, 211, 298; 18c, 300; 22a, 283;
479c, 283; 482b, 253; 485c, 51, 22e, 293; 28c, 257; 28d-33a, 111;
281; 496e, 283; 500e, 51, 96, 171, 30a, 111; 30c, 111; 30d, 111; 30e,
173, 281, 305; 501a, 51; 503d, 89; 257; 31d, 111; 32d, 287; 38a, 303;
508b, 283 38c, 229, 303; 42a, 303; 43a, 287;
Laches 181b, 293 55c, 262; 55d, 262; 55e, 51, 279,
Laws 676b, 119, 288; 688b, 257; 281; 56a, 51, 262-3, 284; 56bc, 306;
688bc, 288; 688c, 119; 691a, 119; 56cd, 263; 56d, 28i; 57c-e, 263;
697c, 119; 704d, 119; 713a, 120; 57d, 281; 61d, 263; 62a, 263-4; 62b,
713b, 112; 720c, 295; 741b, 288; 304; 64e, 276, 278; 66a, 92; 66b, 92
741e, 288; 760de, 289; 767c, 148; Politicus 268d, 112, 257; 269a, 102,
769d, 119, 288; 775c, 288; 819d, 113; 269b, 113; 269c, 108, 113;
240; 821a, 110, 139; 822a, 110, 112, 269d, 115-16, 286; 269d-270b, 287;
123, 287, 289; 867b, 62; 886e, 294; 270a, 113, 287; 272b, 102, 286;
 Index Locorum 327
272bc, 286; 272d, 108; 272d, 113; 157ab, 301; 160de, 301; 161e, 257;
272e, 103; 273b, 65, 102, 103-4, 176bc, 303; 179-3, 287; 181b, 298;
287; 273c, 118; 273d, 102; 273e, 181c, 196; 181d, 304; 182c, 190,
23,287; 274d, 113; 274e, 286; 191, 298, 301; 183a, 191; 183b,
275b, 112; 277d, 211; 277e, 298; 278; 184-6, 303; 185a, 303; 185b,
284a, 287; 288c, 257 303; 185c, 246, 303; 186a, 303;
Protagoras 348e, 280 186bc, 304; 189a, 303; 189e, 303;
Republic 380c, 288; 462b, 93; 473b, 191c, 300; 191d, 300; 193c, 300;
288; 479a, 279; 502b, 288; 506c, 194c, 300; 194de, 300; 196ab, 300;
52; 507b,304; 510ab,279; 510b, 201d, 208-10, 211; 201e, 193,299;
50, 51, 208, 284; 510c, 284; 510d, 202b, 209,299; 202e, 298; 207a,
65,278; 511a, 208,284; 511b, 279, 209; 207b, 244; 210b, 305; 210bc,
284; 511c, 284; 511d, 52, 284; 305
514b, 208; 515b, 170-1; 516c, 171; Timaeus 19b, 258; 19b-27a, 259;
518b, 208; 519c, 284; 520b, 58; 19e, 286; 20d, 259; 21b, 286; 21d,
520c, 51, 52-3, 94, 279; 520e, 284; 259, 305; 22d, 280, 286; 23a, 259;
521c, 55, 280; 523a, 52, 279, 280; 26a, 259; 26d, 305; 27a, 28ft 305;
524e, 279; 526c, 94; 526d, 54, 55, 27c-29d, 12-13, 241, 244, 247-53,
56; 526de, 279, 280; 527d, 279; 278; 27d, 188, 250; 28a, 248; 29b,
536e, 57; 528a, 49, 66; 528c, 49; 247, 250, 252, 304; 29c, 188, 248,
528e, 48, 49; 528e-530d, 55, 59; 250, 252-3, 304; 29d, 44; 30a, 18,
529a-530c, 280; 529b, 56; 529c, 23, 43, 114,288; 30b, 167; 31a, 43,
56, 58, 60, 64, 65, 66, 107,279, 280, 44-5; 31c, 236,299; 31c-32a, 236;
284; 529cd, 280; 529d, 49, 65-6, 32c, 278, 287; 33a, 287; 33b, 167;
102-3; 529d-530b, 279, 280; 530a, 33d, 287; 34a, 106, 110, 168, 286,
64, 89, 106, 107, 108, 111, 115-16, 303; 34b, 167; 35a, 303; 35a-37d,
279, 280, 286; 530b, 11, 49, 56, 58, 111; 35e, 83; 36a, 300; 36c, 289;
59, 60, 61, 63, 66, 99, 103, 127, 279; 36e, 163,293, 303; 37a, 109, 242,
530d, 53-4, 57, 304; 530e, 57; 531c, 245, 247-9, 253, 303; 37b, 250,294,
280; 536c, 257; 545d, 119, 120, 297; 37bc, 248, 304; 37d, 107, 286;
288; 545e, 257; 546a, 103, 118, 38a, 108, 286; 38b, 287; 38c, 107-8,
288; 546e, 4, 103; 602b, 257; 602d, 148, 163, 286; 38d, 110, 122, 137,
229; 608a, 257; 616b, 287; 138, 147-8, 151, 157, 290; 38e, 106,
616b-617d, 289; 616c, 66, 125, 168, 201, 300; 39c, 60, 62, 64, 106,
161; 617a, 66, 289, 291; 617b, 66, 107, 108, 142, 286, 287; 39d, 108,
126-7, 145, 289, 290; 617c, 66, 127 286; 40a, 106, 109-10, 162-3, 168,
Sophist 234b, 303; 237b, 257; 250c, 284, 287; 40b, 106, 109, 168, 201,
304; 251a, 305; 251e, 283; 253a, 286, 287; 40c, 143, 144, 145, 148,
211, 300; 254a, 284, 304; 255c, 303; 163, 288, 290, 293; 40cd, 201, 280;
262c, 229; 262e, 211; 263d, 243-4; 40d, 286, 288; 41a, 287; 41b, 300;
263e, 303; 265c, 69; 266a, 70-1, 72; 41c, 243; 41d, 288; 42a, 303; 42b,
266c, 303 120, 244; 42d, 201; 42e, 201; 43a,
Symposium 211b, 50, 563; 211c, 206, 300; 43b, 286, 303; 43d, 243;
278-9 43e, 243, 301; 44a, 244, 245, 288,
Theaetetus 147c, 302-3; 149b, 305; 303; 44b, 244; 44c, 38, 244; 44d,
152a, 301; 152c, 227; 152de, 250; 303; 45a, 302; 45d, 71; 46a, 243;
153e, 301; 155d, 233; 157a, 227; 46e, 287; 47a, 12, 51, 61, 107, 109,
328 Index Locorum
148, 246-7, 280; 47ab, 279; 47b, 78d, 301; 79b, 217, 222, 301; 79d,
174; 47bc, 249; 47c, 250, 286, 288, 301; 80bc, 18, 234; 80c, 71, 217,
303, 304; 48a, 105-6, 113-15; 48b, 222, 239; 81a, 203, 237, 276; 81b,
197, 209; 48bc, 299; 48c, 208; 48e, 203, 204, 205, 217, 297; 81c, 299,
194; 48e-53b, 14, 189, 192, 193, 302; 81c-e, 203; 81d, 204, 299; 82a,
194, 205-6, 211, 212, 216; 49a, 193, 301; 82d, 204; 83a, 301; 84a, 300;
194, 215, 221, 222, 223, 231; 49b, 86d, 120; 86e, 286; 87b, 120; 88d,
194, 206; 49c, 194; 49d, 189, 223, 222; 88de, 19, 195, 215, 219, 222;
253, 297; 49e, 195, 215, 297; 50a, 88e, 286; 89a, 286; 89bc, 297;
215, 220-1; 50b, 191, 206, 298, 89c, 205, 299; 89d, 205, 248;
50bc, 192, 223; 50c, 192, 193, 194, 90a-d, 244; 90cd, 248, 303; 90d,
215,216, 220, 223, 232,298; 120, 248; 91a, 219; 91c, 286; 91d,
50d-51b, 193; 50d, 215, 219, 220, 51, 219
288; 50e, 215, 220; 51a, 193, 194, 'On the Good , 92
9

215; 51b, 192, 220, 223; 51e, 194;

52a, 194; 52b, 193, 194, 215, 219; PLINY
52d, 19, 25, 193, 215, 219, 220; Natural History XXX, 3, 292
52de, 195, 215; 52e, 221; 53a, 22,
23, 210, 215, 219, 235, 278, 53b, PLUTARCH
210, 224, 288; 53c ff., 194, 200, Quaestiones Platonicae V I I I , 1, 293
204, 205, 206, 210, 215, 221, 222, Numa c. I I , 293
223.297, 53d, 26, 40-1, 195, 208, De Facie in Orbe Lunae
299, 54a, 81, 186, 208, 266, 299; 922f-9232a, 294; 926d, 294
54ab, 292; 54c, 201; 54d, 298; 55a,
287; 55b, 298, 299; 55c, 42-3, 212; PROCLUS
55cd, 284; 55e, 299; 56b, 200-1, Philebus Commentary 688, 284
250; 56bc, 194, 299, 301; 56c, 194; Euclid Commentary 71, 303
56d, 200-1, 288; 56e, 205, 299;
57a-c, 206; 57c, 19, 195, 204, 215, PTOLEMY
222.298, 299, 301; 57d, 198-9, 204; Planetary Hypotheses I, 2, 296; I I ,
58a, 18, 215, 217, 218, 301; 58b, 7, 296; I I , 12, 296
217, 222; 59a, 217, 301; 59c, 200;
59cd, 257; 59d, 257, 301; 60c, 217, SEXTUS EMPIRICUS
222, 301; 60e, 217, 222; 61a, 204, Against the Mathematicians VII
217, 222, 298; 61b, 217, 222; 61c, 116-18, 25, 276, 296
301; 61c-68d, 211; 61d, 116-17,
118, 174, 206,288; 61e, 211, 301; SIMPLICIUS
62a, 301; 62b, 301; 62c, 161, 167; in de Caelo 242.18,278; 242.21, 278;
63c, 17-18, 23; 63e, 19, 23; 65c, 68; 295.11, 278; 488.18, 285; 488.23,
66d, 235; 67b, 301; 67c, 68, 301; 95, 97,275; 492.3, 5; 492.31, 97,
67e, 211, 228; 68b-d, 48, 67-73; 285; 493.1, 285; 493.3, 95, 97;
68b, 69, 70, 71; 68b-d, 265-6; 68d, 497.18-22, 291; 504.17, 147;
12, 67, 69, 70; 68e, 13; 69a, 235; 504.17-20, 291; 532.13, 161; 587.1,
69b, 68; 69c, 22-3; 70c, 301; 71a, 37; 587.20, 277; 700.18, 286
302; 71ab, 281; 72c, 222, 300, 301; in Physica 28.8, 39, 278, 297; 32.6,
73b, 300; 74c, 301; 74d, 301; 75b, 37; 35.3, 29; 155.31, 29; 156.13, 27;
114; 75e, 114; 77bc, 246; 77e, 300;
 Index Locorum 329
164.24, 27; 178.25, 29; 300.21, 37; XENOPHANES
300.31, 27; 327.24,296 Fr.15,294; Fr.16,294

STOBAEUS XENOPHON
1.18.1, 294 Oeconomicus IV, 2, 305

THEMISTIUS
Oratio 21, 245cd, 284
 Index of Names

References i n italics are to the pages of the Notes section.

Aelian, 276 Callippus, 62, 128, 141, 144, 145, 146,

Aetius, 159, 275, 276, 288, 295 148, 149, 157, 178, 182, 290
Anaxagoras, 27, 28, 29, 30, 36, 38, 39, Cronus, 4, 102, 108, 111, 112, 120
40, 45, 82, 83, 84, 85, 86, 91, Clarke, 275
125, 159, 166, 173, 177, 208, Cleanthes, 163,293
267, 272,275, 282, 283, 288, 295 Cleinias, 110, 168
Anaximander, 90, 152, 156, 159, 160, Clotho, 66,127, 139
165,292 Copernicus, 46, 137, 163, 165, 181,
Anaximenes, 159 183, 269, 293, 295
Apollodorus, 153,291 Cudworth, 31
Aristarchus, 163,165, 181 Cypris, 37
Aristotle, 5, 7, 26, 30, 34, 38, 39, 40,
62, 70, 93, 96, 108, 142, 148, Daedalus, 65
159, 160, 162, 163, 164, 165, Democritus, 7, 14, 15, 25, 26, 27, 29,
167, 172, 175, 177, 178, 179, 30, 34, 36, 38, 39, 40, 41, 42, 44,
186, 201, 213, 216, 217, 218, 45, 125, 144, 145, 159, 174, 175,
228, 229, 230, 236, 239, 246, 186, 201, 231, 232, 233, 237,
251, 274, 275, 276, 277, 279, 238, 272, 275, 277, 280, 288, 295
280, 282, 285, 286, 287, 288, Dercyllides, 126
292, 293, 294, 295, 297, 298,
299, 300, 301, 303, 305 Descartes, 5, 14, 30, 31, 32, 169, 186,
Aristoxenus, 92, 283 215, 217, 232, 233, 234, 238,
Athenian Stranger, 110, 168,287 276, 287, 295, 299, 301
Atlas, 86 Dionysius, 277, 303
Atropos, 66, 127, 139 Diotima, 50
Autolycus, 149
Einstein, 91, 281, 283
Barrow, 31, 33,275, 276, 282, 306 Empedocles, 3, 14, 29, 30, 36, 37, 38,
Berkeley, 169, 222 39, 40, 45, 86, 103, 121, 140,
Bessell, 163 186,188, 194, 208, 217, 271,
Boreas, 257, 304 272,276, 282, 290, 293
Epicurus, 39, 41, 103, 186, 272
Boyle, 14, 35, 169, 238,277, 295, 301 Euclid, 239, 302
Euctemon, 144, 145, 146
Callicles, 294
 Index of Names 331
Eudemus, 97, 98, 285, 290
Eudoxus, 5, 62, 97, 98, 99, 128, 134, Orethuia, 257, 304
141, 144, 145, 146, 148, 149,
151, 152, 153, 154, 155, 156, Parmenides, 250, 251, 256, 260, 282,
157, 178, 182, 269, 284, 289, 291 297, 298, 303
Eusebius, 278 Philip of Opus, 3
Philodemus, 99
Galileo, 163, 174,293, 295 Philolaus, 125, 131, 150, 152, 156,
Gassendi, 5, 14 159, 165, 270,288, 291
Geminus, 98,284, 291 Planck, 32
Pliny, 153, 291
Harvey, 47,293 Plutarch, 163,292, 293
Hephaestus, 37 Proclus, 239, 283, 290, 302
Heraclitus, 196, 225, 226, 251, 260, Protagoras, 79, 85, 225, 226, 227, 254,
297 260,279, 282
Hermes, 137 Protarchus, 92, 262, 263
Hesiod, 3, 62, 103, 147, 208 Ptolemy, 155, 180, 181, 182, 183, 295
Hipparchus, 145, 182
Hippocrates of Chios, 125 Sextus Empiricus, 25, 275, 295
Hippolytus, 160, 275, 292, 293, 295 Simplicius, 5, 29, 93, 95, 97, 98,100,
Hobbes, 31 146, 148, 151, 152, 160, 274,
Hume, 78, 111,281, 294 275, 276, 277, 283, 284, 285,
288, 290, 292, 295, 296
Kepler, 46, 165, 181, 182, 183, 185, Socrates, 3, 6, 50, 51, 52, 53, 55, 56,
269, 293 57, 64, 65, 80, 82, 83, 84, 85, 86,
87, 88, 90, 92, 102, 103, 112, 118,
Lachesis, 66, 127, 138, 139 147, 160, 170, 189, 192, 195,
Laplace, 275 196, 207, 208, 210, 243, 245,
Leibniz, 28, 275, 277, 294 249, 251, 252, 254, 255, 256,
Leucippus, 7, 14, 15, 26, 27, 29, 30, 34, 257, 259, 261, 262, 263, 266,
36, 38, 39, 40, 41, 42, 44, 45, 278, 282, 286, 294, 303
174, 186, 201, 216, 233, 237, Solon, 257
238, 272, 277 Sosigenes, 97, 98, 99
Locke, 222 Stobaeus, 167, 293

Lucretius, 103, 272, 285 Thales, 152, 158, 159

Themistius, 283
Menander, 282 Theodorus, 189, 190
Meno, 6, 7, 41, 52, 72, 74, 79, 80, 81, Theon, 98, 126
82, 85, 88, 89, 90, 151, 156, 157, Theophrastus, 163,276, 280, 285
184, 198, 207, 208, 235, 237, Typhon, 257
252, 266, 270,282, 301, 306
Meton, 144, 145, 146 Zeus, 102, 108, 196
Newton, 28, 169, 171, 238, 276, 277, Zoroaster, 153
280, 281, 284, 287, 293, 294, 301
 General Index

References in italics are to the pages of the Notes section.

ageing, 105, 201, 203-5 Republic and myth of Er, 125-8

anthropic principles, 33-4 retrograde motion, 126-7, 131-6,
archai, 195, 202, 206, 208, 250, 252 138, 142, 144-5, 179-80
Aristotle: evidence on geometrical superior and inferior planets, 136-8
atomism, 200-3, 204, 229-31 Timaeus model, 128-31
astronomy, 1, 2, 3, 5, 6, 9, 11, 12, 46, atomists, 1, 2, 5, 7, 8, 10, 14, 15, 17,
48-67, 92, 93, 94, 95, 96, 97, 99, 20, 24, 25, 26, 38, 39, 40, 41, 44,
101-13, 118, 122, 124, 125, 128, 90, 91, 103, 164, 167, 168, 174,
130, 131, 133, 140, 141, 144, 176, 177, 178, 187, 196, 209,
148, 149, 151, 152, 153, 154, 232, 233, 238, 240, 267, 271,
156, 157, 158, 159, 160, 178, 272, 273, 274, 296, 297
179, 180, 182, 183, 185, 209,
238, 241, 244, 249, 255, 259, Babylonians, 5, 141, 143, 145, 157,
262, 265, 269, 270, 271, 274, 158, 291
278, 279, 280, 281, 284, 285,
286, 289, 294, 304 causation, 13, 61, 92, 98, 106, 115-18,
Copernican, 183-6 119, 206, 260, 269, 275, 288, 291
eclipses and occlusions, 139-45 celestial latitude, 131-6
great year, 108, 145-8, 149 celestial motion, 3-5, 64-7, 101-23,
inequality of the seasons, 108, 145-8 124-58
Meno's paradox and astronomy, Epinomis, 101, 111, 113, 121-3
157-8 Laws, 109, 110-11, 113, 121-3
parallax, 163, 164-6,293 Phaedo, 101-2, 113, 121-3
Philolaus, 125, 126-7, 131, 150, 152, Philebus, 101, 111, 113, 121-3
156, 159, 165,270, 288, 291 Politicus, 101, 102-5, 113, 121-3
planetary velocity and distance, Republic, 101, 106-7, 113, 121-3
145-8 Timaeus, 101-23
Plato and development of, 124-58, centrifocal theory, 4, 160, 161
178-83 chaos, 5, 17, 18, 19, 20, 21, 22, 23, 24,
Plato and Eudoxus, 155-7 25, 27, 31, 32, 34, 63, 90, 91,
Ptolemaic, 178-83 102, 104, 115, 192, 206, 210,
relation of astronomy and cosmology, 235, 259, 261, 271, 272
124-58 colours, 12, 30, 49, 68, 69, 70, 71, 72,
regularity of celestial motion, 101-11 117, 190, 192, 227-8 265
 General Index 333
contrary power (enantian ... desmoi, 201, 206, 207, 210-13, 203,
dunamin), 110, 136-9, 140, 141, 207,211,500
146, 149, 150, 151, 154, 158, deuteros pious, 82-7, 246, 246
270,287, 290, 291 Duhem-Quine thesis, 78-80
Copernican principle, 34
Copernican revolution, 164-7, 183-4 Earth, 4, 19, 20, 31, 34, 36, 37, 44, 60,
cosmogony, 4, 7, 17, 19, 20, 22, 24, 34, 85, 86, 90, 102, 103, 115, 125,
81, 91, 166, 177, 192, 233, 259, 127, 128, 130, 131, 133, 136,
268, 270, 271, 272, 274 137, 141, 142, 143, 146, 147,
cosmology, 1, 2, 3, 4, 6, 7, 13, 15, 17, 153, 155, 156, 159, 160, 161,
21, 22, 24, 26, 32, 34, 42, 44, 46, 162, 163, 164, 165, 166, 178,
67, 74, 89, 91, 96, 99, 100, 102, 180, 218, 270, 277, 289, 290,
103, 104, 112, 113, 118, 121, 122, 293, 294, 296
123, 124, 131, 132, 137, 149, eclipses, 139-45, 163, 291
151, 152, 153, 158, 159, 160, ecliptic, 5, 124, 125, 128, 129, 130,
162, 163, 164, 165, 166, 167, 131, 132, 133, 134, 141, 142,
175, 177, 178, 181, 182, 183, 143, 149, 152, 163, 270, 271,
198, 206, 239, 240, 255, 257, 289, 290, 291
258, 259, 266, 268, 270, 271, Egyptians, 62, 143, 259
272, 274, 276, 277, 281, 287, 299 elements, 4, 17, 19, 20, 22, 23, 24, 25,
animate cosmology, 167-70, 175-8 26, 27, 29, 39, 40, 42, 43, 71, 92,
Copernican cosmology, 183-6 115, 117, 181, 184, 189, 190, 191,
divine cosmology, 262-4 193, 194, 195, 198, 199, 201,
Greeks and geocentrism, 164-7 202, 206, 207, 208, 209, 212,
Plato and development of, 159-86 216, 217, 221, 223, 224, 230,
Plato and geocentrism, 162-4 231, 235, 240, 247, 261, 272,
Presocratic, 160-2 276, 278, 299, 301
subsequent cosmology, 183-6 earth, 4,19,20, 22,37,40,43,68,69,
71, 115, 169, 188, 189, 194, 195,
daughters of necessity, 66, 99, 127, 197, 198, 200, 201, 202, 205, 207,
139, 162, 281 209, 212, 218, 221, 235, 270, 277,
degeneration: cosmological, 3-5, 102-5 296, 299, 301
Phaedo, 102-3 water, 4,19, 20, 22, 23, 35,36, 37,40,
Politicus, 102-5 43, 68, 69, 71, 83, 101, 103, 115,
Republic, 103 118, 159, 160, 161, 166, 169, 188,
degeneration: political, social, moral, 189, 194, 195, 197, 198, 200, 201,
118-21 205, 207, 209, 215, 221, 229, 235,
Laws, 119-21 299, 301
Phaedo, 118 air, 4, 19, 20, 22, 27, 36, 37, 38, 40,
Politicus, 118-21 43, 68, 70, 71, 83, 86, 115, 160,
Republic, 118-21 161, 162, 167, 188, 189, 194, 195,
Timaeus, 120-1 197, 198, 200, 201, 205, 207, 209,
Demiurge, 30, 38, 44, 89, 106, 107, 221, 235,299,301
111, 114, 170, 210, 224, 225, 235, fire, 4, 12, 19, 22, 23, 37, 40, 43, 49,
236, 237, 243, 245, 247, 268, 68, 69, 70, 71, 72, 83, 96,106,115,
273,276, 288, 289 116, 117, 118, 125, 160, 165, 169,
174, 188, 189, 194, 195, 197, 198,
334 General Index
200, 201, 205, 206, 207, 208, 209, relation to receptacle paradoxes,
211, 212, 215, 220, 221, 223, 224, 194-6
228, 235, 244, 253, 265, 288, 292, relation to stoicheia, 196-200
295, 299, 301 significance of, 238-40
aether, 27, 29, 83, 155, 178, 278 Socrates' dream, 208-10
empeiria, 51, 69, 70, 71, 72, 89, 96, space, 214-16, 217-19, 221-3
171, 173, 262, 265, 269,279, 305 teleology and bonding, 235-8
epicycle, 149, 179, 180, 184, 186 vs. atomists, 187-8
epistemology, 12, 51, 61, 80, 82, 89, 90, geometry, 11, 49, 54, 56, 65, 121, 155,
151, 240, 248, 250, 252, 253, 175, 239, 240, 263, 279, 280
260,281, 286 geostatism, 133
equant, 179, 180, 181, 184, 186 God, Christian, 7, 15, 19-23, 28, 31,
equinox, 129, 145, 146, 148 32-3, 35, 41, 47, 170, 176-7, 239,
equipoise principle, 86, 161, 167 273,276, 284, 295
experiment, 12, 49, 67, 68, 70, 72, 73, God, Plato's conception of, 4, 12, 48-9,
74, 75, 77, 78, 206, 220, 266, 67-73, 103-5, 108, 111-14, 115,
268, 274, 281, 282 116, 152, 167-70, 174, 188, 244,
263, 269,289, 295
fish's eye view, 66, 101-2 gold example, 220-1, 230, 190, 191,
flux, 13, 14, 123, 174, 187, 188, 189, 192-3, 204, 205-7, 219, 220, 221,
190, 191, 192, 194, 195, 196, 229, 230,297, 298
197, 201, 202, 206, 207, 213, gravity, 4, 10, 18, 19, 20, 21, 30, 31,
214, 220, 221, 226, 229, 240, 32, 38, 165, 172, 177, 179, 181,
250, 252, 254, 259, 261, 262, 233, 238, 265,275, 287, 295, 301
272, 287, 297, 298 great year, 108, 147

geocentrism, 125, 164, 166, 167, 178, heliocentrism, 133, 164-7, 181, 184,
182, 185,294 293, 295
geometrical atomism, 1, 7, 13, 14, 15, Heraclitean flux, 14, 190-1, 205, 228,
29, 40, 41, 68, 71, 81, 117, 121, 249, 251,253,286
174, 175, 187, 188, 190, 193, hippopede, 134, 135-6, 177, 178, 179
194, 195, 198, 202, 205, 206, hypotheses, 5, 9, 12, 50, 72, 74, 81, 82,
207, 208, 209, 211, 212, 214, 88, 89, 90, 91, 94, 95, 96, 97, 98,
215, 216, 218, 220, 221, 223, 99, 156, 157-8, 207, 256, 266, 277
224, 225, 226, 227, 229, 234,
235, 237, 238, 239, 240, 241, inequality of the seasons, 145-8, 149
250, 252, 266, 272, 273,292, inferior/superior planets, 136-7
297, 301 instrumentalism, 8-10, 74, 93-100,
ageing process, 203-5 122, 155-8 169, 266, 274,280
Aristotle on, 200-3, 204, 229-31
basic nature of, 194-6 Jupiter, 125, 136, 137, 138, 139, 147,
Democritus and Descartes, 231-5 148,289, 295
gold example, 205-7
language, 211-13 letters and syllables, 186,196-200,
matter, 214-16, 218-21 208-10, 211, 246, 261, 263, 271,
relation to letters and syllables, 298, 299
196-200, 208-13 like-to-like principle, 4, 19, 20, 23, 24,
 General Index 335
25, 27, 115, 116, 118, 165, 166, 25, 30, 38, 39, 42, 44, 82, 84, 85,
167, 173, 176, 179, 207, 225, 87, 90, 91, 125, 237, 238, 239,
228, 234, 237, 288, 293 270, 271, 272, 273
line and cave, 11, 13, 49-67, 226 Presocratics, 73, 157, 167, 208, 254,
269, 270, 271, 273
Mars, 125, 126, 127, 136, 137, 138, protos pious, 82-7, 282
139, 142, 146, 147, 148, 289 Ptolemaic astronomy, 166
Meno's paradox, 6, 7, 41, 74, 79, 80, Pythagoreans, 152, 159, 239, 270,286,
81, 82, 85, 89, 90, 151, 156, 157, 288
184, 198, 207, 208, 237, 266, 270
Mercury, 110, 125, 131, 136, 137, 138, realism, scientific, 1, 2, 4, 6, 8, 9, 10,
139, 140, 142, 143, 146, 147, 72, 74, 80, 89, 91, 93, 97, 99,
148, 150, 151, 153, 269, 288, 122, 155, 156, 168, 172, 235,
289, 290, 291 234, 265, 266, 267, 268, 272,
Milesians, 172,186, 188 273, 274,281, 285
mirror: images, receptacle like a, 71, reason and necessity, 13, 26, 36, 101,
225, 228, 231, 243 113-15, 122, 187
Moon, 5, 20, 27, 29, 31, 44, 60, 66, 107, receptacle, 7, 13, 14, 19, 23, 25, 186,
110, 125, 127, 128,129, 130, 187, 189, 190, 191, 192, 193,
131, 132, 133, 134, 137, 139, 194, 204, 205, 206, 208, 209,
140, 141, 142, 143, 146, 147, 210, 211, 213, 214, 215, 216,
148, 150, 151, 152, 162, 164, 217, 218, 219, 220, 221, 222,
177, 179,280, 288, 290, 292 223, 224, 225, 226, 227, 228,
phases of, 140-1 229, 230, 231, 232, 233, 234,
multiverse, 7, 32-4, 42-5, 273 253, 258, 260, 288, 296, 298
myth, 23, 66, 67, 87, 96, 101, 102, 104, and qualities, 223-7
105, 112, 113, 118, 120, 121, 123, and language, 188-91
140, 170, 258-62, 270, 286, 287, and matter, 214-16, 218-21
304, 305 and mirrors, 71, 225, 228, 231, 232,
Plato's use of, 258-9 243
myth of Er, 64-7, 112, 125-8, 138-9, and paradoxes, 191
161, 286 and space, 214-16, 217-19, 221-3
myth of Politicus, 101-23 retrograde motion, 126-7, 131-6, 138,
142, 144-5, 179-80
necessity, 113-15, 117-18
Neoplatonism, 183, 184-6,295 Saturn, 125, 136, 137, 138, 139, 147,
148, 289
observation, 1, 9, 10, 11, 48, 49, 56, 57, scientific laws, 10, 114, 115-17, 170-5
59, 60, 61, 62, 63, 64, 73, 146, nomological realism, 170-3
149, 267, 270, 273 civil law and physical law, 173-5
scientific revolution, 47, 175, 273,295
paidia, 119, 255-8,287 Socratic irony, 85
paralattein, 64, 99,103,279, 286 Socrates' 'autobiography', 82-7
parallax, 163, 164-7,293 Socrates' dream, 6, 193, 208-10, 211,
perception: Timaeus and Theaetetus, 267
227-9 Solstice, 129-30, 145-8
physiologoi, 1, 2, 7, 8, 10, 14, 15, 17, Soul, 12, 24, 28, 51, 56, 83, 87, 88, 93,
336 General Index
96, 109, 110, 111, 120, 121, 161, 87, 89, 90, 93, 113, 114, 116, 122,
162, 167, 168, 170, 172, 200, 170, 174, 177, 209, 239, 267,
207, 210, 228, 232, 240, 241, 268, 274, 283, 299
242, 243, 244, 245, 246, 247, Theory of Everything (TOE), 33-4, 82,
248, 250, 252, 255, 257, 258, 91
260, 261, 268,293, 296, 303 tribe, 69-73, 89, 96, 171-3, 262-3, 264,
human soul, 241-4, 245-50 268, 304
world soul, 12, 93,109,120, 229, 233, two worlds (TW), 11, 48, 49, 50, 51, 52,
241-4, 247-50, 251, 253, 261, 262, 240, 241, 247, 249, 252, 253,
269, 297, 304 259, 260, 261, 263, 264, 268,
space, 33, 81, 143, 176, 179, 180, 190, 278, 279, 296
191-4, 213, 214, 215, 216, 217,
219, 220, 221, 222, 223, 224, underdetermination, 6, 72, 74, 75, 76,
228, 229, 230, 232, 233, 238, 77, 78, 79, 80, 81, 89, 91, 93, 97,
246, 277, 299, 301 156, 157, 208, 266, 267, 270,
stoicheia, 193, 194, 195, 196, 205, 207, 272,281, 282
208, 210, 220, 229, 235, 240,
252, 253, 259, 260, 261, 266, Venus, 110, 125, 131, 136, 137, 138,
270, 271, 298 139, 140, 143, 146, 147, 148,
Stoics, 103, 163 150, 151, 153, 269,288,289,
Sun, 5, 20, 27, 28, 29, 31, 34, 36, 44, 290, 291
60, 61, 66, 101, 107, 110, 125, vortex, 14, 25, 27, 29, 30, 86, 91, 175,
127, 128, 129, 130, 131, 132, 177, 178, 217, 231, 233, 238,
133, 134, 136, 137, 138, 139, 270,277, 295
140, 141, 142, 143, 144, 145,
146, 147, 149, 150, 151, 152, wandering cause, 13, 105-6, 113-18
162, 164, 165, 177, 179, 181, wandering motion, 105-9
184, 185, 275, 276, 280, 288, winnowing basket (plokanon), 19, 25,
289, 290, 292, 295 214, 215, 216, 218, 231, 232,
275, 299
taxis, 43, 89, 92, 95, 114,277, 287
techne, 12, 20, 49, 51, 69, 70, 96, 115, zodiac, 5, 43, 128, 129, 132, 133, 134,
171-3, 261, 262-3, 268, 274, 278, 137, 138, 142, 144, 269, 290
304 zoogony, 30, 34-6, 39, 270, 273
teleology, 1, 2, 5, 6, 7, 10, 14, 15, 17,
25, 26, 30, 32, 44, 46, 74, 85, 86,