ANCIENT GREEK LETTER WRITING

Ancient Greek
Letter Writing
A Cultural History
(600 bc–150 bc)

P A O L A CE C C A R E L L I

1
 3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Paola Ceccarelli 2013
The moral rights of the author have been asserted
First Edition published in 2013
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2013938774
ISBN 978–0–19–967559–3
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 Preface

My interest in letter writing began with the idea of a new edition with commentary
of the correspondence of the Seleucid kings: I wanted to look at the way the
medium inflected the meaning, at the influence of the epistolary form on the
content and its reception. But these letters functioned in a context, a context that
is lost to us, although we constantly try to reconstruct it. I thus realized that if
I wanted to understand something about the way royal letters functioned in the
Greek world, I had to widen the enquiry, and in two directions. On the one hand,
I needed to offer a contrastive analysis of the functioning of letters and other
forms of mediated communication, that is, to contextualize letter writing within
communicative practices, as well as more generally within the larger field of
writing. On the other hand, I needed to look at both ‘real’ letters and documents,
and at letters and documents embedded in a narrative: for only the latter would
provide an ‘original’ context, even if bent by the laws of the literary genre in
question. As a result, this book moves to and fro between real, epigraphically
preserved letters (from the archaic to the Hellenistic period, including both
everyday and royal correspondence); real documents (decrees for instance);
literary letters, in particular the letters embedded in the work of the ancient
historians, in drama, and in the speeches of the Athenian orators; and other
types of embedded documents, among which are oral messages as one of the
forms taken by mediated communication in the ancient world. Moreover, because
I was interested not only in writing and letter writing, but also in the connotations
of these documents, in how the Greeks had perceived them, I have given attention
to Greek narratives concerning the invention of writing, and the move towards an
epistolary format; these narratives have been contrasted with the actual (Greek)
documentation, but also occasionally with Near Eastern narratives of the origins
of writing and epistolary writing.
Interest in ancient letters has been growing in recent years; a quick glance at
Klauck’s masterful survey (2006) shows how much has been written on the subject
by both literary critics and historians. The former have produced a number of
studies on letter writing in specific authors and on epistolary collections, offering
literary and in particular narratological analyses of the way the letter functions in
the context of larger narratives (drama, historiography, and especially the novel).1
The renewed interest in ancient epistolary writing is also behind the recent
publications of collections of translated ancient letters.2
Ever since Bentley’s epoch-making dissertation on the letters of Phalaris,
the issue of the authenticity of the literary letters by famous historical personalities
has attracted attention. Studies on specific corpora (the letters of Alexander, of

1
Recent instances of this approach include Rosenmeyer 2001; Jenkins 2006; the papers in Morello
and Morrison 2007; Olson 2010. Studies on the ancient novel (in particular the seminal collection
edited by Holzberg 1994) paved the way for this approach.
2
Costa 2002 and Rosenmeyer 2006 focus on the literary and fictional; Trapp 2003 has a wider
scope.
vi Preface

Phalaris, and of Themistocles) and on the authenticity of royal or seemingly

official letters embedded in literary texts continue to appear, but the focus has
now shifted from the issue of authenticity to determining the function, organiza-
tion, and Sitz im Leben of these corpora.3
As for ‘real’ letters, new documentary material, and in particular letters on
papyri (or, for the Roman world, those written in ink on the extraordinary
wooden leaf tablets from Vindolanda), have made it possible to come close in
an unprecedented way to the everyday life of a woman in Graeco-Roman Egypt,
or to that led by a Roman soldier posted in a garrison on Hadrian’s Wall; a find
like that of Vindolanda has also made it clear how widespread literacy and
epistolary contacts were, even at the periphery of the Roman world. Thus some
of the recent scholarship on ancient letters has led to an extraordinary advance in
our understanding of levels of literacy, and more generally of life and gender-
relationships, in Roman Egypt, in Roman Britain, and indeed in the ancient world
tout court.4 The new material has also enabled further work on the typology and
phraseology of ancient letters.5
Other new documents, in the shape of Greek letters on lead or ceramic
published only in very recent years, now make it possible to attempt a more
detailed narrative of the development of letter writing, from its origins as the
transposition into writing of an oral message given to a messenger, to its becoming
the main instrument through which Hellenistic kings communicated with the
Greek poleis. Welles’s Royal Correspondence of the Hellenistic Period (1934) is still
the landmark study of the official letters preserved on stone; since then, with the
exceptions we shall look at presently, there have been important publications
concerning specific letters or groups of documents, but no general study of the use
of letters for official communication in ancient Greece. Attempts have, however,
been made at looking into the articulation between personal and official letter
writing, and more generally at exploring the question of the origin of the official
letter (a question that had already been posed by Welles).6
New ideas and impulsions in respect to the connotations of the letter as a
communicative medium have come from studies dealing with the diffusion of
news and information in the ancient world (e.g. Longo 1981, Lewis 1996), and
even more from some recent approaches to the royal correspondence in the
Hellenistic period. Jean-Marie Bertrand’s seminal article ‘Formes de discours
politique: décrets des cités grecques et correspondance des rois hellénistiques’,
first published in 1985, underlined the distinction between decrees of the city-
states and royal letters, two antithetical, yet interrelated forms of public written
discourse. John Ma has further shown how the Seleucid kings and the cities

3
On letter collections see e.g. Cortassa and Culasso-Gastaldi 1990 (Themistocles); Hinz 2001
(Phalaris); Merkelbach 1954 (Alexander). Royal letters in literary texts: Gauger 2000. General discus-
sion in Rosenmeyer (2001: 169–233) and Trapp (2003: 23–37).
4
Egypt: Bagnall and Cribiore 2006; Vindolanda: Bowman and Thomas 1983, 1994, 2003, now at
<http://vindolanda.csad.ox.ac.uk>; Bowman 1998; synthesis: Muir 2009.
5
Formal classification (mainly the prescript): Exler 1923; phraseology: Koskenniemi 1956. For
more recent work see the bibliography in Schmidt 1997: 775. On syntax and pragmatics of non-literary
Latin letters, Halla-Aho 2009.
6
Van den Hout 1949; Stirewalt 1993, an uneven work, with some excellent ideas; Porciani 1997.
 Preface vii

interacted, under the sign of a shared euergetic language, by exchanging letters

and decrees.7
The contrast between city-decree and royal letter appears as a significant
moment in a long process. It is this process I am interested in: how letter writing
comes to define itself in relationship to the other writing practices of the polis. In
terms of methodology, a historically grounded understanding of letter writing and
its evolution has to be framed within a discussion of the uses of literacy and orality
in the archaic and classical world, of the relation of writing to specific political
forms, and of the larger question of the communication between individuals,
poleis, and kings. Thus, at the centre of this book will be the distinctions between
‘letters’, or, to be more precise, ‘official letters’ and other genres of writing as well
as oral forms of communication.
Three issues are central in this. The first one concerns the choice of media: in
the world of the Greek poleis, a world very much based on face-to-face interaction,
where decisions were taken in the assembly, the choice between oral and written
messages matters, both from a formal point of view and in terms of impact on the
audience.8 Letters are ideally placed for this kind of questioning: because epistol-
ary communication is very close to oral communication, presenting in written
form most of the deictic marks of oral direct discourse, the person of the sender is
very markedly present in the epistolary communication. ‘The idea of parousia, the
projection of the official’s person, the sense of his felt presence, and the transmis-
sion of his authority is fundamental in official letter writing . . . The reiteration and
reanimation of the official’s authoritative voice authenticated the message and
effected the recipients’ response.’9
A remark of Philostratus concerning the now lost treatise in epistolary form
‘How to write letters’ by Philostratus of Lemnos, his son-in-law, shows that the
ancients were perfectly aware of the illocutionary tendency of letter writing.
According to Philostratus, his son-in-law’s letter had been written against Aspa-
sius, the ab epistulis (secretary) of the emperor (Caracalla or Severus Alexander),
who wrote letters in far too complex a style; what follows is Philostratus’ percep-
tion of the right tone of a royal letter.
ÆPŒæøæ ªaæ c › KØ  ººØ, P E K Łı Å ø P KØåØæÅ ø , Iººa
Å, P Æs I ÆçÆ, KØc  ı çŁ ªªÆØ, Æç ØÆ b æ Å f  ı.
(Philostr. VS II, 628 Kayser)
For an emperor when he writes a letter ought not to use rhetorical syllogisms or trains
of reasoning, but ought to express only his own will; nor again should he be obscure,
since he gives sound to laws, and lucidity is the interpreter of the law.
The bringing together of the sound of the emperor’s voice and of the law it
represents in Philostratus’ discussion of epistolary writing underlines the performa-
tive aspect of imperial utterances.10 The sound of a Hellenistic king’s voice will have

7
Bertrand 1985 (developing ideas sketched in Wörrle 1978); Ma 1999; survey of royal letter writing
in Virgilio 2011.
8
On face-to-face interaction as opposed to mediated communication see Bassi 1998.
9
Stirewalt 1993: 5.
10
Cf. Millar 1992: 637 (and 93 for short discussion of Philostratus’ passage); for a survey of the
imperial correspondence, see 213–28, as well as Millar 1967.
viii Preface

been very much the same, at least in his intention; but how will it have been heard in
the Greek poleis, as compared with the Roman empire? More generally, how was the
sound of the epistolary voice marked in the context of the Greek city? Did poleis send
letters, did they choose this way to make their voice heard elsewhere?
The second main issue concerns historical context and literary form: how did
letter writing evolve in ancient Greece? When was it first perceived as a distinct mode
of writing (and was it ever perceived as a genre)? It is generally accepted that until the
late fifth or early fourth century in Greece the letter is not codified as such: the extant
letters do not show traces of the standard introductory and closing epistolary
formulae, but appear to be simply the written version of an oral message, sent by
an individual who cannot use his voice. How did the letter reach its standard form,
attested from the second part of the fourth century bc, and comprising prescript,
formula valetudinis, body of the message, and final greeting? As for official letters,
was Welles right in his assumption that the official letter to an individual developed
out of the private letter, while the official letter to a community was based on the
prevailing form of communication between communities, the city decree?
The third issue, obviously linked to the two already raised, concerns ideology:
how does official letter writing differ from other forms of official writing or official
speech? Does it carry particular ideological connotations? Why did specific
individuals or groups choose (or not choose) this means of communication?
And what place does it occupy in the Greek polis, especially seeing that it is the
preferred means of royal long-distance communication? Probably because of the
relative scarcity of epistolary documents of the archaic and classical period, letters
have played a marginal role, if they have been at all considered, in discussions of
the respective roles of orality and literacy in Greek culture and society. These
discussions have tended to be confined to the archaic and classical period, on the
basis of the assumption that from the fourth century onwards literacy was anyway
widespread. That writing was available is undisputed; but was it always chosen?
And what kinds of writing (or of speech) were chosen as representative? Types of
writing—and types of speech—might have been felt to be linked to specific forms
of government and to specific discourses.11
An excellent critical reading of the debate surrounding the use of writing was
provided by Rosalind Thomas’s Literacy and Orality in Ancient Greece (1992).12
The ideological model has by now prevailed over the notion of technological
determinism: it is commonly admitted that society determines the uses it will
make of writing. The drawing of crude connections between writing and specific
forms of government is also generally avoided, and both the thesis that establishes
a link between writing and democracy, and its opposite, namely that there exists
an essential connection between writing and autocracy, have run their course.13

11
Terms such as ‘ancient perceptions’, ‘connotations’, or ‘representations’ of writing and letter
writing, as well as (less often) ‘discourse’ or ‘ideology’ will appear relatively often in what follows. I take
representation to mean the process by which social meaning is transmitted among individuals through
literary creations, visual images, and other media; the significance of representation lies in its role in
both the perpetuation and contestation of discourses. Discourses enable, reproduce, and diffuse the
power relations underlying material reality; ideologies misrepresent these relationships.
12
See now her reassessment: Thomas 2009.
13
See Thomas 1992: 144–7; Hedrick 1999 (with Sickinger 2009); Thomas 2009; and already Musti
1986. Holders of the first view include Meritt 1940; Detienne 1988b; Missiou 2011; the second has been
 Preface ix

This is certainly a good thing; but while statements of principle have been
relatively numerous, not enough detailed attention has been given to the specific
functions of different kinds of writing—and of speech.14 This is what I shall
attempt in the pages that follow.
The book is (somewhat artificially) divided into an introduction, and two main
parts: a narrative of beginnings, so far as we can reconstruct them, and as the
Greeks saw them; and a contextualization of letter writing in the polis.
After the introduction, which focuses on general aspects of letter writing
(ancient and modern definitions of letter writing; conveyance of letters in the
ancient world; and analysis of the terminology and its implications), the second
chapter discusses the evidence for the beginnings of alphabetic writing and its
impact, and more generally the sociology of early writing; the main point is the
existence of regional differentiations, something that will be stressed again and
again in the book. I then trace, against this background, the development of letter
writing as a genre, on the basis of a detailed analysis of the enunciation of the
earliest Greek letters (550–350 bc), and of a comparison of these early letters with
other communicative forms such as curses. My point is that until the mid-fourth
century bc there is no shared, stable epistolary format: letters appear to be the
relatively haphazard transcriptions of oral messages.
The third chapter focuses on representations, and in particular on the traditions
on the invention of writing and letter writing. These traditions are analysed for
what they can tell about the connotations attributed to writing (and specifically to
letter writing) by the ancients; whenever possible, I have tried to contextualize
these narratives within their chronological horizon. It emerges that writing,
throughout the archaic and classical periods, is firmly anchored among human
beings (the gods have very little to do with it); and that while letter writing may
have been actually used early on for long-distance communication, this—with the
notable exception of Bellerophontes’ letter—finds no reflections in art or litera-
ture. When narrating the origins of writing, sixth- and fifth-century Greeks seem
to have chosen to see in it a craft imported from the outside—without specifying
any use for it—or to have attributed its invention to some epichoric hero, in which
case, the function of writing is usually said to be the preservation of memory and
poetic creation. Three divergent accounts exist, however, that see in ‘writing’
‘epistolary writing’; I shall argue that these accounts are part of a fifth-century
debate on the uses of writing, and that in all three cases letter writing is negatively
marked. The chapter ends with a look at ancient discussions on the origins of the
epistolary greetings, and on its connection with oral greetings.
A study of the letters in the ancient historians (ch. 4) closes this first part, and
prepares for the next. The analysis of communication in Herodotus confirms the
conclusions of the preceding chapters on the closeness of oral and written
message. However, letter-writing characters are in Herodotus marked as oriental

notably upheld by Steiner 1994. Detailed discussion in Pébarthe 2006: 15–30 and, with focus on Athens,
Pébarthe 2005; but see Thomas 2010 for a slightly different evaluation.
14
An early exception: Stoddart and Whitley’s (1988) contrastive analysis of literacy in Crete, Attica,
and Etruria. On the necessity of a ‘qualitative model’ see Davies 2005; Liddel 2009; Osborne 2009; Day
2010: 30–1. Also interesting is the focus on ‘sacred words’ between orality and literacy in Lardinois,
Blok, and van der Poel 2011.
x Preface

or tyrannical, while letter writing itself can be often deceptive. This is no longer
the case in Thucydides, although letters still tend to prove dysfunctional when
viewed in the context of the polis. The references to letter writing in Xenophon do
not present any marked differences from those in Thucydides; things change with
Ctesias, to whom the first embedded love-letter is owed, and, in a different way,
with Anaximenes. I shall argue that this change reflects a more general change in
communicative practices in Greece, a change that we see actualized in Polybius’
work, as well as in the epigraphical letters studied in chapter 7.
The second part of the book considers the connotations of letter writing within
the polis. The fifth and sixth chapters concern Athens, and look at letter writing
through the lens of literary texts; the seventh chapter, based mainly on epigraph-
ical material, looks at the wider Greek world. In the fifth chapter, I analyse the
appearances of letters on the Athenian stage, both in drama and in comedy, to
tease out the connotations of epistolary writing. Tragedy clearly presents epistol-
ary communication as problematic; while it is difficult to correlate directly the
uses of letters in tragedy with those in Athenian political—or everyday—life, a
comic fragment of the fourth-century comic poet Antiphanes explicitly contrasts
the masculine and public speech of the rhetors with the feminine and private
writing of letters. As it happens, in Antiphanes both are marked negatively—that
is after all often the point in comedy—yet this contrast is an important one. In the
sixth chapter, I pursue the analysis of the contrast between public speech and
letter writing, as it plays itself out on the political and forensic scene, through a
discussion of the mention of letters in the speeches of the Attic orators; I end with
a look at the very first epistolary treatises, and with an attempt at teasing out the
reasons for the choice of this extraordinarily successful form, and its implications.
The seventh chapter moves outside Athens and into the Hellenistic period. I first
describe the formal contrast (in terms of both structure and terminology) between
decree and official letter. The latter is the instrument through which Hellenistic
kings communicate their decisions, at all levels; Greek poleis use letters for internal
communications, but do not usually inscribe them, and for internal decisions and
interstate communication they tend to use decrees. However, some Greek poleis
appear to have made use of letters for interstate communication, and some letters by
poleis are inscribed by other poleis. Thus, while some Greek cities seem to resist the
letter (or at any rate, resist its monumentalization in their landscape), other Greek
cities do not have problems with this instrument. Faced with this situation,
I advance the hypothesis that the letter, a very personal instrument of communi-
cation, may have been more acceptable in those poleis which were accustomed to an
oligarchic, personal way of dealing with power. Here, the point about regional
differentiation in the uses of writing, made in the first chapter, finds its counterpart.
Three appendices close the book, giving the text and translation of the Greek
documentary letters known to date, from the earliest one (c.550 bc) to the mid-
fourth century bc; a selection of ancient texts on the invention of writing; and a
list of official letters by Greek poleis preserved on stone, as well as a list of the
letters sent by Roman magistrates and inscribed on stone, from the earliest
attested until 1 bc.
 Acknowledgements

It is a pleasure to be able to thank the persons and the institutions that have made
this book possible. The project began its life in 1999, during my year as fellow at
the Center for Hellenic Studies in Washington; I am grateful to the directors,
Debbie Boedeker and Kurt Raaflaub, and to the other fellows (notably Ruth and
Victor Caston, Jonathan Hall, Franziska Lang, Astrid Moeller, David Rosen-
bloom, Hans van Wees, and Victoria Wohl) for the fantastic atmosphere and
the intellectual stimulation which enabled me to refine the guiding questions that
inform my investigation into Greek epistolography. After my return to L’Aquila
University, I had the opportunity of presenting part of my work in Tours, at the
second and third conferences ‘Epistulae antiquae. Le Genre épistolaire antique et
ses prolongements européens’, organized in 2000 and 2002 by Élizabeth Gavoille
and Léon Nadjo; the chapter on Antiphanes’ Sappho received a first airing there,
as did ideas on communication, oral and written, in Greece and the Near East.
Maurizio Giangiulio, Mauro Moggi, Mario Lombardo, and Leone Porciani offered
welcome feedback on letter writing in Herodotus’ Histories at the conference
‘Erodoto e il modello erodoteo’, held in Trento in 2003; Lucio Bertelli, Umberto
Bultrighini, and Mario Lombardo helped me define my ideas on the role of letter
writing in the polis, on the occasion of the conference on ‘Democrazia e anti-
democrazia nel mondo greco’, held in Chieti in 2003. A most welcome interrup-
tion to work on the project occurred in 2004, with the birth of Michael; I cannot
adequately express my gratitude to my colleagues at L’Aquila, and especially
Professors Franca Ela Consolino, Maria Grossmann, and Maria Carla Giammarco
Razzano, for their understanding and support, on that occasion and throughout
my time there. These memories are even more precious in the light of the fact that
five years later, after I had already left, L’Aquila was devastated by an earthquake,
and the lives of all my colleagues—and the life of the city—were changed forever.
Part of my maternity leave was spent in Cambridge, in the congenial setting of
Clare Hall: this was when the project finally began to take definitive shape (not
least because of the superb resources of the University Library), though it was
again slowed down by another change of place and circumstances: my move to
Durham, in September 2006, meant learning to cope with—as well as function
in—an academic world entirely new to me. If I managed to settle into the system
with relative ease, it is thanks to the help of my husband, Ingo Gildenhard, and the
understanding of my new colleagues. And the system had its rewards: a term of
institutional leave combined with a top-up grant from the AHRC allowed me to
spend most of the academic year 2008/9 on the manuscript. During this period
I had the opportunity of presenting four chapters of the book to audiences in Paris,
as a visiting professor at the École des Hautes Études en Sciences Sociales, which,
fittingly, brought me back in touch with my academic beginnings: I am grateful to
Claude Calame not only for his invitation and for the numerous comments he
and the other participants made, but also for his constant friendship, and for all
that I have learnt from him ever since the years I spent as ‘assistante en histoire
xii Acknowledgements

ancienne’ in Lausanne: his influence, and more generally the Lausanne years, have
been decisive in shaping my work.
The first version of the manuscript was finished in 2009, on the shores of the
Bodensee, at the University of Konstanz, where I spent the last leg of the AHRC
fellowship, at the kind invitation of Professor Ulrich Gotter: as in Cambridge,
I benefited from the superb resources and facilities of the University Library,
which contributed much to the shape of the first full draft of the book. That this
moment of closure was entirely preliminary is due to the incredible luck I had with
OUP’s anonymous readers: I owe to their careful reading more than I can say.
Their comments resulted in a fundamental reworking of the entire manuscript,
both at the level of overall organization and innumerable minute details. One
concern they raised, which I found impossible to address on my own, was English
idiom, and I am therefore very grateful to Dr Leofranc Holford-Strevens for
agreeing to proofread the revised draft, which resulted in significant improve-
ments, and not only on the formal level. One is always faced with the tension
between care for the details and the overall picture, a tension well encapsulated by
Aby Warburg’s statement ‘Der liebe Gott steckt im Detail’—a reformulation of the
old German proverb ‘Der Teufel steckt im Detail’. Finding the crucial details is
difficult enough as it is; but getting them always right well-nigh impossible. Here
help and feedback from colleagues have been vital. I owe much to audiences at
Somerville (Oxford) and the University of Toronto (in particular Andreas Bendlin
and Martin Revermann) in 2006, when the project was still at an early stage; to the
comments by Michele Faraguna, Edward Harris, and Mirko Canevaro on my
ideas on letter writing in the orators during the conference ‘The Letter: Epistolary
Formats in the Ancient World, 3000 bc—ad 53’, organized in Rome by Uri
Yiftach-Firanko; to the help of my Durham colleagues Ted Kaizer, Andrej Pet-
rovic, P. J. Rhodes, and David Thomas; to discussions with Marcello Carastro,
Silvia Milanezi, Claude Mossé, and Pauline Schmitt-Pantel in Paris; to discussions
with Manuela Mari on a number of issues, among which notably the Macedonian
diagramma and letter writing in Polybius; to feedback from Jaime Curbera on the
part dealing with defixiones; and to conversations with Ulrich Gotter during my
time at Lake Constance, especially on the interface of language and power. With
great generosity, Madalina Dana shared information on three recently published
letters on lead. Last but not least, I am profoundly grateful to my copyeditor,
David Pelteret, who invested an enormous amount of time and effort into getting
a difficult typescript ready for press.
But I doubt that this book, which I began while on my own, would ever have
seen the light of day without the family support I have enjoyed in its later stages:
I therefore dedicate this book to Ingo and Michi.
 Contents

List of Abbreviations xv
Map xx

1. Ancient Greek Letters: An Introduction 1

1.1. Sketching a Typology 2
1.2. A Modern Definition 8
1.3. Transmission 10
1.4. Ancient Greek Terminology 13

Part I Greek Beginnings: Writing and Letter Writing,

Evidence and Representation
2. Writing and Letter Writing: The Evidence 23
2.1. Writing in Archaic Greek Society: Who Writes What,
and Why? 27
2.2. The Earliest Letters 35
2.3. Conclusion 56
3. Writing and Letter Writing: Representations 59
3.1. Writing as a Craft Imported from Outside the Greek World 63
3.2. Writing as the Invention of a Greek Culture Hero 66
3.3. The Invention of Writing as Letter Writing 72
3.4. A Greek Narrative of the Beginnings of a Genre 89
3.5. Conclusion 98
4. When a Letter and Why? Narrative Strategies in the Ancient Historians 101
4.1. Mediated Long-distance Communication in Herodotus 103
4.2. Mediated Long-distance Communication in Thucydides 130
4.3. Into the Fourth Century: Continuity and Change 149
4.4. From the Fourth to the Second Century 160
4.5. Conclusion 178

Part II Letter Writing and the Polis

5. Writing and Letter Writing on the Athenian Dramatic Stage 183
5.1. Theatrical Letters: Tragedy and Satyr-play 183
5.2. Theatrical Letters: Comedy 240
5.3. Conclusion 258
6. Letters on the Legal and Political Stage 265
6.1. Antiphon 269
6.2. Andocides, Isaeus, Lysias 271
6.3. Demosthenes 272
6.4. Aeschines 280
xiv Contents

6.5. Isocrates and the Genre of Epistolary Treatises 286

6.6. Conclusion 292
7. Poleis and Kings, Letters and Decrees: Official Communication
in the Hellenistic Period 297
7.1. The Language of Royal Letters and Civic Decrees 298
7.2. Letters by Greek Poleis 311
7.3. Interpretation 327
Epilogue 331

Appendices
1. Archaic and Classical Documentary Letters 335
2. Ancient Traditions on the Invention of Writing 357
3. Official Letters Sent by Greek Poleis or Koina and Inscribed
on Stone, in Chronological Order 365
Bibliography 385
Index Locorum 415
Thematic Index (Names, Places, Topics) 432
 List of Abbreviations

Alberti I. B. Alberti, Thucydidis Historiae, i–iii (Rome, 1972–2000)

AVI H. R. Immerwahr and R. Wachter, Attic Vase Inscriptions <http://avi.
unibas.ch/home.html>
ARV2 J. D. Beazley, Attic Red-Figure Vase Painters, 2nd edn (Oxford, 1963)
BE J. and L. Robert et al., Bulletin Épigraphique (in Revue des Études
Grecques, 1938–)
Bernabé A. Bernabé, Poetae epici Graeci: Testimonia et fragmenta, i–ii (Leipzig
and Berlin, 1988–2007)
BGU Aegyptische Urkunden aus den Königlichen (later Staatlichen) Museen
zu Berlin, Griechische Urkunden (Berlin, 1895–)
BNJ I. Worthington (ed.), Brill’s New Jacoby: Online (Leiden, 2007–)
<http://referenceworks.brillonline.com/browse/brill-s-new-jacoby>
Campbell D. A. Campbell (ed. and trans.), Greek Lyric, i–v (Cambridge, MA,
and London, 1982–93)
CEG P. A. Hansen (ed.), Carmina epigraphica Graeca, i: Saeculorum VIII–
V a. Chr. n. (Berlin, 1983)
CEL P. Cugusi (ed.), Corpus Epistularum Latinarum Papyris Tabulis
Ostracis servatarum, i: Textus; ii: Commentarius (Florence, 1992)
CID F. Lefèvre (ed.), Corpus des inscriptions de Delphes, iv: Documents
amphictioniques (Paris, 2002)
CIRB Corpus inscriptionum regni Bosporani (Moscow and Leningrad, 1965)
CVA Corpus Vasorum Antiquorum
D.–K. H. Diehls and W. Kranz (eds.), Die Fragmente der Vorsokratiker, 6th
edn (Berlin, 1952)
DNP H. Cancik and H. Schneider (eds.), Der neue Pauly: Enzyklopädie der
Antike, 16 vols. (Stuttgart, 1996–2003)
DT Defixionum tabellae quotquot innotuerunt tam in Graecis Orientis
quam in totius Occidentis partibus praeter Atticas, collegit, digessit,
commentario instruxit . . . A. Audollent (Paris, 1904)
DTA R. Wünsch (ed.), ‘Defixionum tabellae Atticae’, in Inscriptiones
Atticae aetatis Romanae, iii: Appendix (Berlin, 1897)
EBGR A. Chaniotis (ed.), Epigraphical Bulletin for Greek Religion, published
in Kernos, 1991–
ETCSL J. A. Black, G. Cunningham, J. Ebeling, E. Flückiger-Hawker,
E. Robson, J. Taylor, and G. Zólyomi, The Electronic Text Corpus of
Sumerian Literature (Oxford, 1998–2006) <http://etcsl.orinst.ox.ac.uk/>
Fouilles de
Delphes III É. Bourguet, G. Colin, [et al.], Fouilles de Delphes, III: Épigraphie, i–vi
(Paris, 1909–76)
FGrH F. Jacoby (ed.), Die Fragmente der griechischen Historiker (Berlin and
Leiden, 1923–58)
xvi List of Abbreviations

FGrHCont G. Schepens (ed.), Die Fragmente der griechischen Historiker

Continued (Leiden, 1998–)
G.–P. B. Gentili and C. Prato (eds.), Poetae elegiaci: Testimonia et fragmenta,
2nd edn (Leipzig, 2002 [1988])
Hude C. Hude (ed.), Herodoti Historiae, 3rd edn (Oxford, 1927)
IC M. Guarducci (ed.), Inscriptiones Creticae, 4 vols. (Rome, 1935–50)
I. Cos M. Segre, Iscrizioni di Cos (Rome, 2004)
I. Délos iv P. Roussel and M. Launey (eds.), Inscriptions de Délos, iv: Décrets
postérieurs à 166 av. J.-C. (Nos. 1497–1524); dédicaces postérieures
à 166 av. J.-C. (Nos. 1525–2219) (Paris, 1937)
IG Inscriptiones Graecae (Berlin, 1913–)
I. Gonnoi B. Helly (ed.), Gonnoi, ii: Les Inscriptions (Amsterdam, 1973)
I. Iasos W. Blümel (ed.), Die Inschriften von Iasos, 2 vols. (Bonn, 1985)
I. Labraunda J. Crampa (ed.), Labraunda iii, The Greek inscriptions, 1–2 (Lund,
1969–72)
I. Lampsakos P. Frisch (ed.), Die Inschriften von Lampsakos (Bonn, 1978)
I. Magnesia O. Kern (ed.), Die Inschriften von Magnesia am Mäander (Berlin, 1900)
IPArk G. Thür and H. Taeuber, Prozessrechtliche Inschriften der griechischen
Poleis: Arkadien, Österreichische Akademie der Wissenschaften.
Philosophisch-historische Klasse, Sitzungsberichte 607 (Vienna, 1994)
I.v.Olympia W. Dittenberger and K. Purgold, Die Inschriften von Olympia (Berlin,
1896)
K.–A. R. Kassel and C. Austin (eds.), Poetae comici Graeci (Berlin and
New York, 1983–)
Kannicht R. Kannicht (ed.), Tragicorum Graecorum fragmenta, 2: Fragmente
ohne Autor, zusätzliche Fragmente und Indizes zu Band 1 (Göttingen,
1981); Tragicorum Graecorum fragmenta, 5. i–ii: Euripides
(Göttingen, 2004)
LCS A. D. Trendall, The Red-Figured Vases of Lucania, Campania and
Sicily (Oxford, 1967)
Legrand Ph.-E. Legrand (ed. and trans.), Hérodote: Histoires, 11 vols. (Paris,
1932–54)
LGPN P.M. Fraser and E. Matthews (eds.), A Lexicon of Greek personal
names (Oxford 1987– )
LIMC Lexicon Iconographicum Mythologiae Classicae, 8 vols.; Indices, 2 vols;
Supplements, 2 vols. (Zurich, 1981–2009)
LSAM F. Sokolowski, Lois sacrées de l’Asie Mineure (Paris, 1955)
LSJ H. G. Liddell, R. Scott, H. S. Jones, and R. Mackenzie, A Greek–English
Lexicon, 9th edn (Oxford, 1925–40); Suppl. by A. E. Barber et al.
(Oxford, 1968)
Masson, ICS O. Masson, Les Inscriptions chypriotes syllabiques, 2nd edn, with
addenda (Paris, 1983)
McCabe D. F. McCabe, ‘The Princeton Project on the Inscriptions of Anatolia’,
Institute for Advanced Study, Princeton (1991), Packard Humanities
Institute CD #7, 1996, now consultable at <http://epigraphy.packhum.
org/inscriptions/ >
 List of Abbreviations xvii

Milet i 3 A. Rehm (ed.), Die Inschriften, in Milet. Ergebnisse der Ausgrabungen

und Untersuchungen seit dem Jahre 1899, vol. iii, Das Delphinion in
Milet (Berlin, 1914)
M.–L. R. Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions to
the End of the Fifth Century bc (Oxford, 1969)
M.–W. R. Merkelbach and M. L. West (eds.), Fragmenta Hesiodea (Oxford,
1967)
OED Oxford English Dictionary, 3rd edn (Oxford, 2011–), published online
at <http://www.oed.com>
Pack2 R. A. Pack, The Greek and Latin Literary Texts from Greco-Roman
Egypt, 2nd edn (Ann Arbor, MI, 1965)
P.Bon. O. Montevecchi (ed.), Papyri Bononienses (Milan, 1953)
P.Cair.Zen. 3 C. C. Edgar (ed.), Zenon Papyri, Catalogue général des antiquités
égyptiennes du Musée du Caire, iii (Cairo, 1928)
Pf. R. Pfeiffer (ed.), Callimachus, i: Fragmenta (Oxford, 1949)
P.Hal. The Graeca Halensis (ed.), Dikaiomata: Auszüge aus alexandrinischen
Gesetzen und Verordnungen in einem Papyrus des Philologischen
Seminars der Universität Halle (Pap. Hal. 1) mit einem Anhang
weiterer Papyri derselben Sammlung (Berlin, 1913)
P.Hib. B. P. Grenfell and A. S. Hunt (eds.), The Hibeh Papyri, I (London,
1906)
P.Lond.Lit. H. J. M. Milne (ed.), Catalogue of the Literary Papyri in the British
Museum (London, 1927)
PMG D. L. Page (ed.), Poetae melici Graeci (Oxford, 1962)
P.Oxy. The Oxyrhynchus Papyri, Published by the Egypt Exploration Society
(London, 1898–)
P.Paris J. A. Letronne, W. Brunet de Presle, and E. Egger (eds.), Notices et
textes des papyrus du Musée du Louvre et de la Bibliothèque Impériale
(Paris, 1865) (Papyri nos. 1–71 republished in UPZ)
Radt S. Radt (ed.), Tragicorum Graecorum fragmenta 3: Aeschylus
(Göttingen, 1985; 2nd edn, 2008); Tragicorum Graecorum fragmenta
4: Sophocles (Göttingen, 1977; 2nd edn, 1999)
RC C. B. Welles, Royal Correspondence in the Hellenistic Period: A Study
in Greek Epigraphy (New Haven, 1934)
RE A. Pauly, G. Wissowa et al., Realencyclopädie der classischen
Altertumswissenschaft (Stuttgart, 1894–1978)
Rose V. Rose (ed.), Aristotelis qui ferebantur librorum fragmenta (Leipzig,
1886)
RVP A. D. Trendall, The Red-Figured Vases of Paestum (Rome, 1987)
RVAp A. D. Trendall and A. Cambitoglou, The Red-Figured Vases of Apulia,
3 vols. (Oxford, 1978–82)
SB Sammelbuch griechischer Urkunden aus Aegypten (Strassburg, etc.,
1915–)
SEG Supplementum epigraphicum Graecum, 1– (Leiden and Amsterdam,
1923–)
xviii List of Abbreviations

SGD D. R. Jordan, ‘A Survey of Greek Defixiones not included in the

Special Corpora’, GRBS 26 (1985), 151–97
S.-M. B. Snell and H. Maehler (eds.), Pindari carmina cum fragmentis,
8th edn (Leipzig, 1987)
Syll 3 W. Dittenberger, Sylloge inscriptionum Graecarum, 3rd edn, 4 vols.
(Leipzig, 1915–24)
Tod M. N. Tod, A Selection of Greek Historical Inscriptions, ii: From 403
to 323 B.C. (Oxford, 1948)
TRS H. de Genouillac, Textes religieux sumériens du Louvre, i–ii (Paris,
1930)
UPZ U. Wilcken (ed.), Urkunden der Ptolemäerzeit (ältere Funde), i: Papyri
aus Unterägypten (Berlin and Leipzig, 1927) (nos. 1–150); ii: Papyri
aus Oberägypten (Berlin, 1935–7) (nos. 151–229)
V. E. M. Voigt (ed.), Sappho et Alcaeus (Amsterdam, 1971)
W. M. L. West (ed.), Iambi et elegi Graeci, 2nd edn, 2 vols. (Oxford, 1992
[1971])
Ziebarth, Neue
Verfluchungstafeln E. Ziebarth, ‘Neue Verfluchungstafeln aus Attika, Boiotien und
Euboia’, Preussische Akademie der Wissenschaften, Sitzungsberichte
(1934), 1022–50
Map. Findspots of letters, c. 550 BC to c. 350 BC, on lead and ceramic.
Olbia, 7 letters; Berezan, 2; Panticapaeum, 2; Hermonassa, 2; Emporion, 3; Rosas, 2; Attica, 8; all other places, one letter each.
 1
Ancient Greek Letters
An Introduction

˜æø ZøØ åÆæØ.  E Ł E A Ø å æØ å , N ÆP  ªØÆØ

ŒÆd a º Ø Ø ŒÆa ºª  ªª . KææŁÆ b ŒÆd ÆP , ŒÆd ŒÆŁØ  Ø
ªæÆłÆ c A Æ K غØÆ Ø FÆØ ‹ ø i ÅŁd K åºBØ  f Ææa
F. ‰  i IÆ ºÅØ ªØÆø, Æ ØØ H Ææa F Iª æ ÆØ
ºØ ØŒ F Œ ºÅ: åæÆ ªaæ åø æe  f OçŁÆº f ŒÆa
æ ƪÆ  F Ł F. PåØ.
verso: ZøØ
Dromon to Zenon, greetings. I give thanks to all the gods if you are in good
health yourself and everything else has been satisfactory. I too am well, and in
accordance with what you wrote to me I am taking the utmost care that no one
troubles your people. When you are ready to sail up in good health, order one of
your people to buy a cotyla of Attic honey; for I require it for my eyes by order
of the god. Farewell.
verso: To Zenon.

The above is one of the many letters the sands of Egypt have preserved. It was sent
around 260–250 bc by a certain Dromon to Zenon, the private agent of Apollo-
nius, the dioiketes (financial minister) of Ptolemy II Philadelphus.1 The address-
ee’s name, inscribed on the verso of the papyrus sheet, appeared on the outer part
of the roll; inside was the text, opening with a standardized greeting, the so-called
epistolary prescript, giving the names of sender and addressee (› E fiH EØ
åÆæØ—the order of sender and addressee might vary depending on their social
importance and on the type of letter), followed by an inquiry about health, itself a
very frequent feature in the opening of letters, then by the request (the ‘body of the
letter’), and by a formula valedicendi (here PåØ, but ææø is also very
frequent). This letter may be taken as representative of the basic format of the
ancient letter, a format which, in its main lines, did not change in the long period
from the end of the fourth century bc to the close of antiquity.2 But letters could

1
P.Cair.Zen. 3. 59426 (SB 3. 6804); the translation given here follows closely that of Hunt and Edgar
1932: 273–4, no. 91.
2
Overview in Klauck 2006: 17–25. The conventional formulae used for opening and closing the
letter, for inquiries about health, etc. are laid out in Exler 1923, based on evidence from papyri dating
from c.300 bc to c.300 ad; see also Thraede 1970; White 1981–2; Tite 2010 on prescripts; Martin 2010
2 Introduction

(and can) take fairly different shapes, depending on the occasion and purpose for
which they were (and are) written. I shall thus begin with an overview of the
various types of letters that can be grouped under the umbrella appellation of
‘Greek epistolography’; this part will close on a modern definition of the letter.
From there, I shall move to conveyance: because of the obvious impact exercised
by the sociological, economic, and political contexts, it is important to look at the
modalities of the transmission of letters (and more generally of information) in
ancient Greece. A discussion of the ancient terminology for ‘letter’ will close this
chapter.

1.1 . S K E T C HIN G A T Y PO L O G Y

Sending a letter or a trusted messenger is, in a world lacking telephone or internet,

the only means of ensuring private communication between persons or groups
who cannot communicate directly. As Cicero puts it, there are ‘many types of
letters, but one thing is certain, that the thing itself was invented so that we could
inform those absent of anything that it might be important for them to know
either concerning ourselves or themselves’.3 Communication is thus foremost;
and yet, just as he underlines the primary communicative function of the letters,
Cicero recognizes that there are many types of letters. Beyond the transmission of
information, letters were used to create and maintain a social network of relation-
ships, as well as to entertain and amuse. Sending a letter rather than an oral
message meant moreover that the content of the exchange could be preserved and
stored in an archive. All this explains the importance of letters: writing and
reading letters was a central part of daily life in the ancient world. This applied
to monarchs, as is shown by the celebrated saying attributed to Seleucus, that ‘if
people only realized what a great task it was just to read and write so many letters,
they would not even pick up a diadem that had been discarded’ (Plut. Mor. 790 ab);
but from the Hellenistic period onwards, this was true of all levels of society.4
It is thus not surprising to find that the ancients did reflect on this activity. No
systematic ancient treatment of letter writing as a genre has survived, but, besides
occasional reflections in the letters of masters of the genre, there are extant some
theoretically oriented discussions of letter writing in the context of larger works, as
well as manuals for the instruction of would-be letter writers.5

on the ‘body’. This letter has actually one rather exceptional feature, in that it presents also, in addition
to the inquiry about health, a ‘thanksgiving formula’ (examples and discussion in Collins 2010).
3
Cic. fam. 2. 4. 1 (to Curio): Epistularum genera multa esse non ignoras, sed unum illum certissi-
mum, cuius causa inventa res ipsa est, ut certiores faceremus absentis, si quid esset, quod eos scire aut
nostra aut ipsorum interesset. Compare Cic. Q. fr. 1. 1. 37.
4
Diffusion at all levels: Trapp 2003: 34. On late antique epistolary networks, and on the character-
istics of late antique letter collections (letters as ‘calling cards’, or to create an image of the writer and
his network: the really important parts were transmitted orally, by a trusted messenger), see the essays
of Bradbury 2004 (Libanius), Salzman 2004 (Symmachus), Liebeschütz 2004 (Ambrose); on the re-
routing of such aristocratic networks into episcopal correspondence, Sotinel 2004.
5
See Koskenniemi 1956: 18–21; Cugusi 1983: 43; Malherbe 1988; Trapp 2003: 42–6; Klauck 2006:
183–8; Poster 2007 is also useful, but marred by a number of slips.
 Introduction 3

The earliest sustained reflection on letter writing is found in chapters 223–35 of

Demetrius’ De elocutione (On Style), a work probably composed around the
middle of the second century bc.6 Among other things, Demetrius is responsible
for citing the earliest formulation of what became a rather trite (as well as inexact)
commonplace: at the beginning of his discussion of the epistolary style
(K Ø  ºØŒe åÆæÆŒæ), he refers to one Artemon (known as the editor of the
letters of Aristotle, an activity confirmed by two late sources) for the opinion that
one must write dialogues and letters in the same style; for a letter is ‘the one part of
a dialogue’ (r ÆØ ªaæ c K Ø  ºc x  e æ  æ  F Øƺª ı).7 The
notion that a letter is a homilia, a meeting in conversation of two persons, will
remain central to letter writers throughout antiquity and beyond; famously Cicero
said that the letter is ‘a [private] conversation between absent friends’, amicorum
conloquia absentium (Phil. 2. 7). To the notion of interaction with friends through
epistolary conversation (homilia) is linked that of the parousia, the presence of the
correspondent, felt at the moment of reading, but also at the moment of writing a
letter: in quite a few instances the writer says that the very act of writing brought to
her/his eyes the image of the addressee.8 Demetrius admits that there is some truth
in this statement, but then proceeds to take issue with it, on the grounds that a
dialogue imitates oral improvisation, while a letter is written and sent as if it were a
gift, that is to say, requires elaboration of a different type. To show what kind of
writing should not be used in letters, Demetrius proceeds to cite a fairly elevated
passage from a letter of Aristotle to the Macedonian general Antipater, concluding
that this resembles more a declamation (epideixis) than plain, light conversation.9
There must be some irony going on here, as Demetrius uses for his refutation of
Artemon’s theory a letter of Aristotle, whose correspondence had been edited by
Artemon himself. The rhetorician then adds that, because of its written character, a
letter should not be full of interruptions, of breaks in sentences, and in general
should not be mimetic.10 This emphasis on writtenness is extremely interesting in
the light of the overlap (terminological, but also practical, as we shall see) between

6
Text and translation in Malherbe 1988; text, translation and commentary in Trapp 2003: 180–2.
Translation and discussion: Klauck 2006: 184–8. Date: Trapp 2003: 43. Larger discussion, including the
later theoreticians: Koskenniemi 1956: 21–53.
7
Demetr. Eloc. 223, with Trapp 2003: 317–19. On the identity of Artemon, also Koskenniemi 1956:
24–6.
8
The graphe is thus the image of the correspondent. Cf. the scene of letter reading in Plautus’
Pseudolus, esp. 35–6, 63–4, with Jenkins 2005: 364–9; Turpilius fr. 1 inc. fab. Rychlewska (= Jer. Ep. 8. 1,
discussed below, 250: sola res est, quae homines absentes praesentes faciat); Cic. Fam. 15. 16. 1, Fam. 2.
9. 2; instances of sexual closeness brought about by letter reading: Hodkinson forthcoming; see
Koskenniemi 1956: 172–80 for instances in letters from papyri; Klauck 2006: 191–2. In the final verses
of the poem by H. Barbusse ‘La Lettre’ (from Pleureuses (Paris, 1895)), writer and addressee become
confused in the mind of the writer: ‘Et mi-rêvant, je ne sais guère | Si c’est moi qui t’écris, ou toi . . . ’
9
Demetr. Eloc. 225: › ªaæ oø Øƺª K ØØŒıø fi  ØŒ Aºº , P ºÆº FØ. The key
terms here are the verbs K ØŒıØ and ºÆºø: the first refers to the practice of giving epideixeis,
public presentations, while the second indicates everyday, light conversation. ¸Æºø appears, however,
in riddles on writing (below), and marks feminine epistolary language, in contrast to male public
speech, in a fragment of a fourth-century comedy, Antiphanes’ Sappho (below, 245 and 254–6).
10
 ªaæ  ØÆÅ A Æ æÅÆ ŒÆd Å Ø  ŒæØB fi æ Ø Aºº , P ªæÆç ÆØ K Ø  ºÆE ,
‘All such imitative style is fitting for an actor, but not for communications couched in writing’ (Demetr.
Eloc. 226).
4 Introduction

oral and written message; Demetrius is clearly driven by his agenda (fixing generic
boundaries) and by the importance that writing by his time has acquired in
everyday life. Remarkably, it is at this point that Demetrius, marking a slight retreat
from the position first assumed, provides us with the second best-known common-
place of epistolary communication, the letter as a mirror of the soul:
—ºE   b Kåø e MŁØŒe  K Ø  º, u æ ŒÆd › Ø º ª · åe ªaæ NŒÆ
ŒÆ  B Æı F łıåB ªæ çØ c K Ø  º. ŒÆd  Ø b ŒÆd K ¼ºº ı ºª ı Æe
NE e qŁ  F ªæ ç  , K Pe b oø , ‰ K Ø  ºB . (Demetr. Eloc. 227)
The letter, like the dialogue, should be strong on characterization; for a person writes a
letter almost drawing an image of his own soul. In every other form of composition it
is possible to discern the writer’s character, but in none so clearly as in the letter.11
This is, again, an exquisitely literary worry that does not find any obvious
reflection in most of the letters we shall be looking at; but it explains the abundant
use made of letters in the novel and in epistolary romances. The remaining
paragraphs of On Style dedicated to letter writing are more oriented towards
practical aspects, such as length (to be avoided), structure (relatively free), choice
of topic (not all topics fit the format of a letter), type and tone of argumentation
(friendly, and making use of proverbs, i.e. of a common, shared property, and
popular as well). All this leads Demetrius to the conclusion that letters should be
written in a mixed style, a combination of the elegant and the plain. For this
ancient rhetorician, the letter is thus, up to a point, a conversation in writing, but
distinguished by a set of relatively fixed rules. Interestingly, the last example
Demetrius cites in his discussion of the style in which letters should be written
is that of a letter addressed to a polis or a king: here, a more elevated style might be
in order (Eloc. 234).
At any rate: if the Artemon mentioned by Demetrius is indeed the editor of
Aristotle’s correspondence, then a reflection on letters had begun already at the
turn of the fourth century bc; the very fact of editing a letter collection implies a
recognition that these documents are of interest as letters. If we can trust Lucian,
the interest in collecting and editing letters was not limited to philosophical
schools: the grammarian Dionysodorus of Troezen, a student of Aristarchus and
active in Alexandria in the second half of the second century bc, is said to have
edited the letters of Ptolemy I, and he too might have prefaced his collection with
some comments on the appropriate style.12 But for us, Demetrius’ treatise is the
earliest text giving a sense of how letter writing was perceived.
The main characteristics of the letter, and consequently the set of appropriate
rules fixed by Demetrius, remained much the same throughout antiquity: the
other ancient discussions of the theory of letter writing (the chapter 27 of Iulius
Victor’s Ars Rhetorica, dated to the fourth century ad, and the section de epistulis

11
Koskenniemi 1956: 40–1 and 175–80, with references to a number of similar passages that show
the diffusion of this topos; Trapp 2003: 39–40.
12
Interestingly, Lucian mentions the edition because Ptolemy in one of his letters allegedly inverted
the normal order, greeting Seleucus with ‘Health’, and closing the letter with ‘Rejoice’ (åÆæØ): ˚Æd
— ºÆE b › ¸ ª ı ºŒø fi K Ø ººø ÆçH I æł c  Ø K IæåB fi b B K Ø  ºB
ªØÆØ ÆPe æ Ø , K d ºØ b Id  F KææH ŁÆØ  ªæ łÆ e åÆæØ, ‰ ˜Ø ı øæ › a
K Ø  ºa ÆP F ıƪƪ çÅ Ø (Luc. Laps. 10).
 Introduction 5

of the excerpta rhetorica published by Halm) follow very much in Demetrius’

footsteps.13
Besides theoretically oriented works, manuals on letter writing also circulated.
Two of them have survived: the  Ø K Ø  ºØŒ  (Epistolary Types) of Pseudo-
Demetrius, a work of uncertain date (proposals range from the second century bc
to the third century ad); and another manual preserved in two versions, one,
referred to with æd K Ø  ºØÆ ı åÆæÆŒBæ (On the Epistolary Style), attrib-
uted to Proclus and going back probably to the fourth century ad, the other,
bearing the title ¯ Ø  ºØÆ Ø åÆæÆŒBæ (Epistolary Styles), attributed to Liba-
nius and dated to the mid-fifth century ad.14 Two lucky papyrological finds show
that many more manuals of this kind must have existed.15 The manuals attributed
to Demetrius and Proclus/Libanius are very basic, offering the novice writer
examples of types of letter (the first lists twenty-one examples, the second expand
the range to forty-one, but of course the categories often overlap), to be chosen
according to the occasion. These manuals, while not discussing epistolary writing
and its connotations, offer interesting, although at times contradictory, insights
into letter writing and its uses.
The opening of [Demetrius’] Epistolary Types seems to imply that the manual is
addressed mainly to persons working in the public administration, who would
have used the various examples of letters as ‘primers’:
Since according to the theory that governs epistolary types, Heraclides, letters may be
composed in a great number of styles, but are written in those which always fit the
particular circumstance, and since letters ought to be written as skilfully as possible,
but as it happens are written indifferently by those who undertake such services for men
in public office, I . . . have sketched a sample of the arrangement of each kind . . . 1. The
friendly type is the one which seems to be written by a friend to a friend. For frequently
those in a prominent position are expected by some to write in a friendly manner to
their inferiors and to others who are their equals, for example to military commanders,
viceroys, and governors. There are times, indeed, when they write to them without
knowing them personally.16

13
Text and translation of Julius Victor’s chapters on epistolarity in Malherbe 1988: 62–4; text,
translation, and commentary in Trapp 2003: 184–8. Discussion: Koskenniemi 1956: 27–33, 44–5. Note
also the De Epistulis by Philostratus of Lemnus (II 257, 29–258, 28 Kayser) and Gregory of Nazianzus,
Ep. 51 (text and translation in Malherbe 1988, resp. 42–3 and 58–61), with Poster 2007: 32–4.
14
Text and translation in Malherbe 1988: 30–41 and 66–81 respectively; French translation with
commentary: Malosse 2004. See also Koskenniemi 1956: 54–7; Trapp 2003: 44–6; Klauck 2006: 202–4.
15
Koskenniemi 1956: 57–9. The first text, P.Paris. 63, a collection of copies of everyday letters
(dated to 164/163 bc), clearly intended for schools, was published by Wilcken in UPZ 1. 110, 144, 145,
and 111; interestingly, while 110 and 111 seem to be copies of ‘real’ letters, 144 and 145 might be
rhetorical exercises in the form of letters. The second text, P.Bon. 5, is a very late (third or fourth
century ad) primer for the writing of Greek and Latin letters, containing in two columns fragments of
approximately ten types of letters: text and translation in Malherbe 1988: 44–57, and now also in
Cugusi, CEL i, 1. Discussion of both documents in Klauck 2006: 204–5.
16
Cf. Malherbe 1988: 30–2. The following are the twenty-one types identified in [Demetrius’]
proem: çغ، , ı ÆØŒ ,  ØŒ , OØØ ØŒ , ÆæÆıŁÅØŒ , K ØØÅØŒ ,  ıŁÅØŒ ,
I غÅØŒ , łŒØŒ , K ÆØØŒ , ı ıºıØŒ , IØøÆØŒ , KæøÅÆØŒ , I çÆØŒ ,
IººÅª æØŒ , ÆNØ º ªØŒ , ŒÆŪ æØŒ , I º ªÅØŒ , ıªåÆæÅØŒ , NæøØŒ , I ıåÆæØ ØŒ (cf.
Klauck 2006: 194–202).
6 Introduction

This programmatic statement is, however, not followed up in the rest of the
treatise. While in his opening and in the first example of a letter the author of the
Epistolary Types seems to imagine a context in which the problems and situations
to be addressed are those typical of the official letter, the definitions and the
examples of letter types actually given in the treatise are relatively restricted and
essentially of private character (friendly letter, blame letter, commendatory letter,
suitable, as pointed out by Welles, mainly for correspondence between members
of the same level of a bureaucracy);17 yet practical, everyday letters concerning
economic exchanges or invitations, that is, types of letters we know were very
much used, are missing too. Also completely unmentioned are standard parts of a
letter such as the prescript and the closing greeting: these are evidently taken for
granted. More disturbing is the fact that the definition of the type sometimes
clashes with the example, as is the case of the friendly letter (no. 1), supposedly an
instrument to be used by those in superior positions to obtain prompt acquies-
cence, but presenting in the example a case of lifelong friendship between sender
and addressee.
It is at any rate clear that for the ancient theorists, the genre ‘letter’ encom-
passed writings of various types, and that it was important to follow the rules
specific to each. As we saw, Demetrius had specified that letters written to cities or
kings should be composed in a more elevated tone than those written for private
use, but had also highlighted as bad instances of this practice, because veering
towards the treatise, the letters addressed by Aristotle to Alexander and by Plato to
the relatives and friends of Dion.18 Cicero in his Familiares sets out a clear
distinction between letters written for private use and letters written for a wider
public, aliter . . . scribimus quod eos solos quibus mittimus, aliter quod multos
lecturos putamus (Cic. fam. 15. 21. 4).19
However, if there is a clear sense that there are rules, these rules are not really
made explicit, nor do the definitions offered include any analysis of the formal
structure of a letter. Rules and formal structure must thus be deduced from the
actual instances of letters. Even a cursory glance at the epistolary material from
antiquity is enough to show that a wide number of fairly different types of letters
are attested which do not really fit the categories of the ancient manuals. This
disparate material may be organized in three main groups: personal (or private,
or documentary) letters, having a specific, practical purpose; literary letters
(a category covering a remarkably wide range, from philosophical treatises to

17
Welles 1934: xlii; Welles explained the noticeable absence of what he called the ‘letter-decree’
with the hypothesis that the treatise might come from Egypt (on the Egyptian origin of the Typoi see
also Klauck 2006: 195). Cf. Koskenniemi 1956: 47–51, 61–2; discussion below, ch. 7.
18
Eloc. 234: K d b ŒÆd º   ŒÆd Æ ØºF Ø ªæ ç ,  ø Æ  ØÆFÆØ [ƃ] K Ø  ºÆd
ØŒæe KÅæÆØ H .  åÆ   ªaæ ŒÆd  F æ  ı fiz ªæ çÆØ· KÅæÅ  Ø [ŒÆd] På u 
ªªæÆÆ r ÆØ I K Ø  ºB , u æ ƃ æØ  º ı æe ºÆæ , ŒÆd æe  f ˜ø
NŒ ı  —º ø . Interestingly, this is the only reference to official letters in the whole part of
the De elocutione concerning letters; the rest seems to concern mostly private correspondence.
19
Ancient texts mentioning the distinction between public and private letters, and the criteria
according to which this distinction should operate, are collected in Cugusi 1983: 30–41. Interestingly,
public letters are dealt with in two pages. The rest all concern private letters: these are the letters said to
be similar to a dialogue, close to a sermo, mirror of the soul, etc.
 Introduction 7

encomia to apologies to fictive letters, including those embedded in longer

narrative works); and official (or diplomatic, royal, imperial) letters.20
However, this seemingly neat distinction hides a much more complex reality.21
There are some truly personal, ‘private’ letters, such as those preserved on papyrus
or on lead, that owe their survival to chance. But many of these personal letters
dealt with economic affairs, and would have been archived as records of transac-
tions: we are already out of the purely personal. Similarly, the ‘private’ corres-
pondence of famous individuals also presents problems from the point of view of
categorization, as it is open to the suspicion of having been kept for publication, by
these same individuals or by someone else, because of its more general interest.
‘Literary’ letters present other types of problems. How does one define the
degree of ‘literariness’ of a letter? Fictional letters in a work of fiction will be
literary letters; but what of real letters embedded in a fictional context, and what of
letters written with diffusion and possibly publication in mind? The format of the
letter could be used for different kinds of public communications: open letters
with a propagandistic goal, letters with a philosophic or moral character, dedica-
tory letters, consolatory letters, treatises, encomia, pamphlets and apologies in the
form of a letter, and so on. Then, there are forged letters, attributed to famous
historical or mythological characters: fascinating texts, documenting a remarkable
attention towards epistolary writing, as well as its importance as a genre. These
were never meant as real letters; yet they play, in a greater or lesser measure, on
the illusion of a rupture of the epistolary privacy to attract the intended audience,
to give the readers the voyeuristic impression of intruding upon private affairs.22
Official letters might seem less difficult to categorize; but here too, strict
divisions are hard to trace. In the case of the correspondence of emperors, for
instance, the distinction between personal and official correspondence is often
blurred: the official letters were addressed to administrators, proconsuls, or
provincial governors, most of whom were personally known to the emperor, or
to cities and associations, which would have sought to be represented in Rome by
a friend of the emperor.23 The same applies to the relationship between the Greek
cities and the Hellenistic kings: whenever possible, the cities tried to secure the
mediation of a çº , a ‘friend’ of the king, that is, they tried to combine the
official request with a personal note. Moreover, the socio-political and cultural
context would exert an influence on the letters themselves, modifying or shaping
their message. A simple administrative letter, whose primary function seems to be

20
Sykutris 1931. See also Reed 1997 (with discussion of the connections between epistolary and
rhetorical practices); Schmidt 1997 (who divides into eight types: edicts (comparable to laws); official
correspondence; open letters, similar to speeches; private letters between individuals; didactic (literary)
letters; literary public letters, poetic or in prose; pseudepigraphic letters, aiming at fictional rhetorical
prosopoeia; and dedicatory letters); Görgemanns 1997a and 1997b; Klauck 2006: 67–70. Nicolai (2004:
118–20) proposes a division into four groups: practical messages; epistles combining a political message
and an elaborate literary form; fictive letters, embedded in all sort of literary works; and letters
composed as exercises in schools of rhetoric, devoid of any practical use.
21
Sykutris 1931: 186–8; Trapp 2003: 3–5, who concludes: ‘The letter, then, is clearly a diverse form
of writing . . . but mapping the diversity, and fixing labels on its constituents, is interestingly problem-
atic’; Gibson and Morrison 2007.
22
Rosenmeyer 2001: 198.
23
Millar 1992: 215. The letters of Pliny in book 10 are another case in point.
8 Introduction

that of solving a problem, transmitting a message, or imposing a decision, may

also, when monumentalized through its inscription on stone, fulfil propagandistic
purposes in favour of the authority from whom it emanated, or of the authority
that ordered its inscription; sometimes this last function may become the most
important one.24
Thus, it is best to avoid trying to impose excessively neat distinctions among the
various types of letters; on the contrary, accepting that there was fluidity between
the various sorts of letter is important to understand what were the really
important elements, those always present.25 In fact, the very overlap pointed out
in the discussion of [Demetrius’] Epistolary Types, where the first type of letter is
qualified as friendly (and the example indeed presents a case of lifelong friendship
between sender and addressee), while its aim is to obtain prompt acquiescence
from subordinates, explains the significance of the choice of a letter for communi-
cation: the letter aims at persuasion through argument, yes, but mainly through
relationship. As Poster puts it, ‘the letter functions performatively to establish the
friendly relationship it proclaims’;26 this explains also the use of generalized and
formulaic expressions of friendship within the letter. Far from being trite formu-
lae, they accomplish an important purpose, that of establishing a relationship,
within which the request or information conveyed is then framed.

1. 2 . A MO D E RN D E F I N I TI ON

After looking at ancient definitions, and attempting an empirical organization of

the material, we need to offer a modern definition that we can then map onto the
field of ancient epistolary writing.27 This is not meant to be a watertight definition:
Gibson and Morrison have argued convincingly for the impossibility (and indeed,
the futility) of such an attempt.28 Rather, my point of departure is the question
with which they close their essay, namely, what the definition of a letter is
supposed to accomplish. Whether ultimately any given text is to be categorized
as ‘letter’ or as ‘poetic epistle’ or ‘epistolary treatise’ is not a central issue; what is
important is to be clear on the markers that give some texts a ‘family resemblance’,
that make them ‘epistolary’ in character, and at times ‘letters’ stricto sensu; and

24
A point stressed by Funck 1996 (see also Ma 1999: 141 on the effect of recording on stone all the
steps in the transmission of the Seleucid administrative correspondence). The letter with which
Themistocles invites the Ionians to rebellion (Hdt. 8. 22, discussed below, 116–17) is a good example
of a text whose efficacy does not depend on the kind of message transmitted and on its acceptance by
the intended audience. Incidentally, very often there is more than one intended audience.
25
Rosenmeyer (2001: 5–11) offers a useful critical discussion of twentieth-century attempts at
classifying epistolary genres; I would, however, disagree with her conclusion that ‘letter-writing is
inherently fictional’ (11). Her definition dilutes the meaning of ‘fictional’: letter writing is inherently
fictional in the same way as any communicative act (including face-to-face speech) is ‘inherently
fictional’: we present ourselves to the other.
26
Poster 2007: 26–7.
27
i.e. after having looked at emic definitions (we shall presently also have a look at ancient Greek
terminology), we move to an etic definition.
28
Gibson and Morrison 2007 (esp. 15).
 Introduction 9

even more, to be clear on what this resemblance, this sense of epistolarity, does to
the texts themselves, how it inflects their meaning and their perception by the
recipient.
The following definition best encapsulates what I am looking for: a written
process of communication between two or more specific individuals or groups
(real or fictional) who find themselves in a situation of spatial distance, or more
precisely, who are not in direct, face-to-face contact. As a result of this spatial
distance, and of the time-lag necessary for the letter to arrive at its destination,
epistolary exchanges imply a temporal distance, which will find a reflection in the
temporal deixis adopted in the letter itself.
Sender and addressee may be represented in the text of the letter through
personal pronouns of the first, second, or third person singular, or also through
the first, second, and third person plural. Thus, in the writing of the message a
spatial and temporal deixis is activated that is specific to the communicative
situation created. The ways in which this happens are culture-bound and
meaningful.29
Most of the literature on epistolarity, ancient and modern, stresses the fact that
the ideological underpinnings of letter writing are philophronesis, parousia, and
homilia, and that the letter’s main function is to maintain friendly relations,
substitute for the sender’s presence, and continue a conversation. Thus, sending
a letter is seen as a way of making oneself present during one’s absence, of binding
together two distant persons, as in a dialogue.30 Yet, a minority of critics have
highlighted the monologic, rather than dialogic, character of the letter.31 While
this varies from situation to situation, it is important to stress that epistolary
communication, when imposed (by choice, or by the circumstances) as a substi-
tute for face-to-face communication, may have the effect of creating a distance.
This is an aspect keenly felt in the Greek tradition. More generally, the interplay of

29
The above is a modified version of the definition given by Sallaberger (1999: 9), in his model
study of Old-Babylonian letter writing; see also Trapp 2003: 1–2; Gibson and Morrison 2007.
Létoublon (2003) offers a definition that, while strongly culture-specific (i.e. closely linked to our
own, as well as to the ancient Greeks’, notion of a letter), highlights important points. Edmunds 2008
has an excellent discussion of deixis (the system of reference that, through deictic expressions such as
pronouns or verbal forms, points to the (usually) extralinguistic contextual situation) in Greek and
Latin literature.
30
Koskenniemi 1956: 34–47; 152; 92: ‘Die Aufgabe des Briefes ist ja eine doppelte: er dient nicht nur
als Benachrichtigungsmittel, sondern auch als ein persönliches, vereiningendes Band’ (‘The letter has a
double purpose: it serves not just as a means to inform, but also as a personal, connecting link’); cf.
Martin 2010: 189–91.
31
Cf. Sallaberger 1999: 9–19; Klauck 2006: 3, citing an extraordinary passage from one of Kafka’s
letters to Milena: ‘You know after all how I hate letters. All the misfortune of my life . . . derives, one
could say, from letters, or from the possibility of writing letters. People have hardly deceived me, but
letters always . . . The easy possibility of letter-writing must—seen merely theoretically—have brought
into the world a terrible disintegration of souls. It is, in fact, an intercourse with ghosts, and not only
with the ghost of the recipient but also with one’s own ghost which develops between the lines of the
letter one is writing and even more so in a series of letters when one letter corroborates the other and
can refer to it as a witness. How on earth did anyone get the idea that people can communicate with one
another by letter! Of a distant person one can think, and of a person that is near one can catch hold—all
else goes beyond human strength. Writing letters, however, means to denude oneself before the ghosts,
something for which they greedily wait.’ (F. Kafka, Letters to Milena, ed. W. Haas, trans. T. and J. Stern
(New York, 1953), 229).
10 Introduction

distance and closeness is to be viewed as a distinctive factor of epistolary commu-

nication that can be spun both ways and that leaves traces at the textual level.
‘Ordinary, face-to-face conversation works on the assumption that any uttered
sentence has as its deictic center (origo) the person who utters it and as its spatio-
temporal reference the moment and place of the utterance itself ’;32 with a letter,
however, no matter how close to ordinary conversation the exchange may be felt
to be, there is a clear separation between the time of composition (the Coding
Time), the time (or times) to which the letter itself may refer, and the time in
which the addressee receives the letter and reads it (the Receiving Time).33 It is
this situation that is so often exploited in narratives, for instance in the novel; in
the ancient world, the time needed for a letter to reach its destination meant that
this gap would actually often play an extremely important role.
Finally, specific formal elements permit the definition of a text as a letter;
among them, the most important is the introductory formula, naming sender
and addressee. The characteristics of this introductory formula, usually reflecting
the relationship existing between the sender and the addressee, are culturally
determined, and thus changing; but the existence of some kind of introductory
formula is necessary for a text to be defined as a letter. Other elements contribute
to determining the epistolary status of a text, such as formulae of transition to the
central part, wishes of good health, closing formulae with greetings; but their
importance for the definition of a letter as such is less central, while their
dependence on the socio-cultural context is heavier.34

1. 3. T R A NS M IS S I ON

The ease or difficulty of conveying a message to its recipient obviously shapes both
the perception of the message and the very choice of whether to rely on a message
for communication. A situation of easy transmission gives the impression (often
deceptive) that the message itself is reliable and transparent, that it indeed offers a
window onto the events; it also makes it easy for everyone to share in this system
of conveyance, so that the society as a whole lives in an interconnected world. In
the absence of easy ways of conveying letters, the very fact of sending one takes on
the connotation of an event, of something special and particularly meaningful;
moreover, the sender has to reckon with the likelihood of delays in the delivery,
and with the possibility of the loss of the letter itself—delays and loss of messages
play an important role in the Greek novel, but also earlier in Greek tragedy.35
However, it is important to distinguish, in the case of archaic and classical Greece,
between long-distance and short-distance communication (e.g. between two

32
D’Alessio 2004: 267–8.
33
Cf. Green 1992: 126–7. For such an analysis, applied to epigrams, see Day 2010: 112–20.
34
Importance of prescript: White 1981–2: 92. Interesting remarks on formal and rhetorical aspects
of the prescript and the body of letters in Tite 2010; Martin 2010. More on epistolary formulae below,
35–47.
35
e.g. Létoublon 2003. This happened also in reality: cf. the rich picture of the transmission of
letters in Hellenistic and Roman Egypt in Parsons 1980: 5–6; Head 2009a and 2009b.
 Introduction 11

cities, or within the chora of the same polis), as well as between private (personal)
and official communication. For personal correspondence one had to rely on the
goodwill of individual travellers (xenoi, merchants, theoroi) that might be going in
the direction needed; there was thus no guarantee that the message would
effectively reach the addressee. Since there never was a civilian postal service,
sending letters at a long distance, although not a rarity, must have been, at least
until the Hellenistic period, mainly limited to specific categories of messenger
such as merchants or soldiers on campaign. Messages could certainly also be sent
at a short distance through slaves or servants, or again through the agency of the
occasional passer-by or neighbour, and often the sender would write in the letter,
as a form of guarantee for the recipient, the name of the person to whom the
message had been given. For the archaic and classical period at any rate, our
documentation includes mainly letters of long-distance traders, or small notes of
limited importance meant for local recipients.36
Official communication between poleis, or between magistrates and poleis,
fared only marginally better. There were ¼ªªº Ø (messengers) and ŒæıŒ
(public heralds);37 for longer distances that needed to be covered quickly, some
cities had couriers, the so-called æ æ Ø (day-runners) or æ  ŒæıŒ
(couriers), such as the Athenian Pheidippides, who according to Herodotus
covered in two days the distance from Athens to Sparta (Hdt. 6. 105–6).38 In
time of war, a general would usually have at his disposal some particularly fast and
trustworthy runners: Herodotus mentions the Argives sending a hemerodromos
‘after finding the best of their runners’ to warn Mardonius of the arrival of the
Spartans with Pausanias (Hdt. 9. 12); the wording (  ı Ø ŒæıŒÆ H
æ æø Iıæ e ¼æØ   K c ØŒ) makes it clear that the
Argives disposed of a pool of day-runners for this type of purpose, from which
they chose the best. That most poleis will have had some fast long-distance
runners among their heralds is also made likely by the casual tone with which
Socrates compares the exertion involved in entering into a discussion with
Protagoras on Protagoras’ terms to competing in the stadion, performing in the
long-distance race (ºØå ), or keeping up with an hemerodromos (Plat. Prot.
335e). Similarly, Aeschines mentions as a matter of fact that the Phocian tyrant
Phalaecus had couriers (æ  ŒæıŒ ) at his disposal, who brought back news of
how matters were in Athens. And Aeneas Tacticus in his Poliorcetica suggests that
‘the trumpeter and the runners should be based and spend their time around the
general’s tent, so that if there is any need of signalling or bringing a message they
will be ready’.39 War, as one might have expected, is one of the contexts in which

36
Overall discussion in Longo 1981: 27–30; Klauck 2006: 63–5; Llewelyn 1994: 26–57 focuses on the
Hellenistic and Roman period; see also Head 2009a on named letter carriers in the Oxyrhynchus
Papyri, and Head 2009b for fascinating material on the role of letter carriers in Jewish letters.
37
Longo 1981: 30–42, who stresses their subordinate role: their duty was ‘to repeat’.
38
˜æ  ŒBæı, ‘runner herald’, is attested in the lexica of Hesychius (æ  ŒBæı· › K d ıB
  a K ØŒÅæıŒÆ Ø Æ ŁÆØ, æ æ ) and Harpocration, s.v., as well as in Photius
and other lexica, but also in Aeschin. 2. 130 (the dromokerykes of the tyrant Phalaecus), Polyaenus, Strat.
5. 26. 1 and Aen. Tact. 22. 3. Discussion in Matthews 1974. On Pheidippides, see below, 113 and 117.
39
Aen. Tact. 22. 3: —æd b e æÆªØ  ŒÅ F ŒÆd ØƺE Id e ƺ تŒc ŒÆd  f
æ  ŒæıŒÆ , <¥ ’>, K  Ø fiÅ ÅBÆØ j ÆæƪªEºÆØ, K   ı  æåø Ø; another reference to
dromokerykes in Aen. Tact. 22. 22.
12 Introduction

hemerodromoi/dromokerykes are most often found.40 However, these messengers

might have carried indifferently a written or an oral message; the difference would
not have been felt, since one and the same messenger had to cover the distance
from sender to recipient. And for a diplomatic exchange, a polis would have rather
sent ambassadors, chosen ad hoc for their appropriateness to the mission and
their ability in speaking; they would have spoken in support of the civic decree,
with far more effective results than simply handing over a letter. Messages might
even be given to heralds orally rather than in writing: if captured, they would
probably not be tortured to force them to deliver information—heralds in theory
were sacred—while a letter, if intercepted, would instantly betray whatever plans
were being made.41
The combination of elements that might have led to the development of an
organized system of communication and transport (size of the territory con-
trolled, political power, strongly centralized administrative structure, and pres-
ence of disruptive forces such as strong enemies on the frontiers, ethnic diversity
within the state, or power of local officials) was simply not present, or not for long
enough, in Classical Greece.42 A postal service of sorts was developed only in the
Hellenistic kingdoms, through adapting the one that had been in place under the
Achaemenids and, even before, in Assyria. The Persian system had been described
by Herodotus and later by Xenophon; it finds its most striking expression in a
passage of De mundo, in which Asia is described as entirely covered by day-
runners, spies or scouts, message-bearers, and especially overseers of beacon-
towers (æ æ Ø  ŒÆd Œ d ŒÆd IªªºØÆçæ Ø çæıŒøæØH  K Bæ ),
the last working ŒÆa ØÆ å , ‘in relays’, so that the king can know the same day
what is happening over all the land he controls.43 Diodorus narrates that in 302 bc
Antigonus created a network of غØÆçæ Ø (‘letter carriers’), adapting what had
been the Achaemenid relay system to the necessities of what would become the
Seleucid kingdom;44 the Ptolemies further developed the system, establishing a

40
Philostratus (Gymn. 4) links the invention of the ºØå , the long run of 24 stades at Olympia, to
the fact that the Arcadian dromokerykes had to go to Greece for matters related to war, and were not
allowed to use horses: frequently running the dolichos proved a good training. (The passage is of course
to be taken as significant of how long distance running was seen rather than as an explanation of how it
came to be: ‘Greece’ is impossibly vague).
41
Lewis 1996: 148.
42
Llewelyn 1994: 4. Thus, the Suda’s definition (Å305 Adler) of hemerodromoi links them directly to
royal activity: ‘Day-runner: the sun. Those who serve the royal orders most quickly are also called so.
The hemerodromoi are young men, slightly older than ephebes, near to their first beards, who serve
such needs, equipped with bows and arrows, javelins and slings, and nothing more. For these things are
useful to them for travel by road’ ( Hæ æ
› lºØ . ºª ÆØ b ŒÆd ƒ ÆE Æ ØºØŒÆE ØÆ  Ø
ÆåÆÆ ØÆŒ   Ø. ƒ æ æ Ø  Ø ’ N , Kçø Oºª  æ æ Ø, æø ªø Kªª ,
ÆE  ØÆÆØ  Åæ  Ø åæÆØ
ø ŒÆd ºH, IŒ ø ŒÆd ıæ ºø ºŁø K Øçæ Ø
º  P
ÆFÆ ªaæ ÆP E æe c › Ø æÆ ªÆØ åæ ØÆ.)
43
[Arist.] De mundo 398a30–5. Cf. Hdt. 8. 8. 1–2; Xen. Cyr. 8. 6. 17–18; Llewelyn 1994: 2–5; Klauck
2006: 60–3.
44
D.S. 19. 57. 5: ÆPe b A Æ c  Æ, w q ŒæØ , ØºÆ ıæ E ŒÆd ıºØÆçæ Ø , Ø’ z
Oø Xºº  ÅæE ŁÆØ Æ. See also, for the period of the Successors, D.S. 19. 13. 5, 7; 19. 14. 4;
19. 85. 5; 19. 100. 3. غØÆçæ (or ıºØÆçæ ) is regularly used in Diodorus for a messenger
transmitting a written message: so for those sent by the Selinuntioi to Gelon and by the Athenians to
the Spartans at the time of the Persian wars, 11. 21. 4 and 11. 28. 5; for the couriers between Pausanias
and the King, 11. 45. 2; bibliaphoroi are sent by the Selinuntioi to the Syracusans, 13. 54. 3; in 14. 101. 2
 Introduction 13

regular postal service for official communication, with relays for riders throughout
the kingdom, and using camels for the transportation of heavier goods;45 from the
Hellenistic period onwards the term ªæÆÆ çæ , ‘letter carrier’, is attested
(first in Polybius, although Phylarchus may have already used it);46 finally Strabo,
writing in the time of Augustus, equates hemerodromoi and grammatophoroi,
giving the impression that a day-runner is someone who essentially carries
dispatches—as was probably by then indeed the case.47 The royal letters that the
Seleucids sent to their administrators reveal the multiple relays required for the
transmission of royal orders; but it is important to keep in mind that this structure
covered only the territories internal to the kingdom, and would thus not have been
of any use in international diplomacy.

1.4 . A N C I E N T G R E E K T E R M I N OL OG Y

Let us now turn to the ancient terminology for letter writing: for the variety of
epistolary types might find a reflection in the terminology, or the terminology
might give us a clue as to how to view the various types of letters.
English has two main words for a written message sent by one person to
another (or to others): ‘epistle’, and ‘letter’.48 Both terms have their root in the
classical past: epistle goes back to Greek K Ø  º, via the Latin borrowing
epistula, while ‘letter’ derives from Latin littera (the Greek uses ªæ Æ or
 ØåEÆ to cover the same field of meanings). In modern-day English, epistle is
chiefly used ironically of very long or pretentious letters; for formal or particularly
elaborate letters, ranking as literary productions; for letters having a public
character; or for letters addressed to a body of persons, such as the New Testament
ones.49 ‘Letter’ today has three main senses. The term is most frequently used to

we find bibliaphoroi of the Thourioi; the Macedonian kings send bibliaphoroi (Eurydice, 19. 1. 4), as
does Agatocles (20. 18. 1). Hence, the occurrence of the term in a passage where Diodorus relies on
Ctesias (D.S. 2. 26. 8 = FGrH 688 F 1b, concerning Sardanapallus) cannot automatically be ascribed to
Ctesias.
45
Muir 2009: 11–12, quoting some fascinating documents (a receipt from the association of
couriers based in Oxyrhynchus, BGU 1232, and a part of the day-book for one of the postal offices,
dated to 255 bc, from P.Hib. 110; a longer extract from the same papyrus is in Austin 2006: 546–8,
no. 309); P.Oxy iv 710, 1–4 (111 bc) mentions 44 bybliaphoroi in the Oxyrhynchite nome. The detailed
discussion of the official ancient postal systems by Llewelyn 1994: 1–25, moves from the Persian empire
to Ptolemaic Egypt, and then focuses on the Roman period.
46
ªæÆÆ çæ ı appears twice in Phylarchus FGrH 81 F 55; but the fragment is preserved in
Polybius (2. 61. 1–5), in a passage in which he criticizes Phylarchus, and the precise wording may well
be Polybius’ own. See below, 168–70.
47
Strabo 5. 4. 13, speaking of the Picentes: Id b æÆÆ æ æ E ŒÆd ªæÆÆ ç æE
I åŁÅ Æ· ‘And instead of doing military service, they were appointed to serve as couriers and
letter carriers’ (the reason being that they had sided with Hannibal; it is slightly surprising to find
communication entrusted to suspect populations). Interestingly, Augustus went back to the principle of
one messenger covering all the distance (but using relays of horses), because the messenger could thus
also convey the message orally, and add further information (Suet. Aug. 49. 3).
48
One could add ‘missive’, i.e. something that is sent (from Latin mittere), generally a written
document, and so often a letter; it is a relatively generic term.
49
See OED, s.v. ‘Epistle’, for a detailed list of the various meanings and uses of the term.
14 Introduction

denote a conventional symbol (an alphabetic character) representing a phoneme

or a group of phonemes. It can also be the precise terms of a statement (the letter
as opposed to the spirit of a text). Finally, it denotes a communication in writing,
addressed to a person or body of persons; it can be either private, or else public,
official, and authoritative in character.50 The last sense is the one which will
mainly occupy us here; but let us meanwhile note that the same richness and
fluidity that we encountered in the English terminology was already present in
Greek and Latin usage.
Latin littera corresponds very precisely to Greek ªæ Æ. In both Latin and
Greek, the same term, littera/litterae and ªæ Æ/ªæ ÆÆ, is used to indicate
the graphic signs with which language is noted, as well as the ‘letter’ qua written
message. This is rather striking, and needs some further attention. Prima facie, the
fact that a term used to indicate a very basic notion, the graphic sign, central to
any form of writing, is also used to denote written communication might appear
to speak to the great importance of the latter. And yet this kind of reasoning can
cut both ways; before we understand the implication of this sharing of names, we
need to fully grasp the meaning of ªæ Æ (littera) ‘letter as alphabetic sign’.
Moreover, just as in ancient Greek other terms besides grammata could be used to
indicate the alphabetic letters (stoicheion for instance), so other more specific
terms for ‘letter’ existed. A detailed analysis of the terminology reveals that the
various terms imply specific connotations and different ways of looking at the
object in question.
The distinction between grammata and stoicheia (‘elements’) has been recently
discussed by Jesper Svenbro; on the basis of a number of passages from the scholia
to the Techne of Dionysius Thrax, he concludes that gramma/grammata denotes
the individual alphabetic letter, considered on its own and deprived of a mean-
ingful context, while stoicheion/stoicheia indicates the letter as part of an intelli-
gible sequence (taxis);51 as such, the stoicheion is always on the side of the sound.
Gramma appears thus as the unmarked, more general term, while stoicheion is the
marked, specialized one.52 This basic distinction, if extended to the entire range of

50
Detailed list of the various meanings and uses of the term in OED, s.v. ‘Letter’ (n) 1. Italian
‘lettera’ and French ‘lettre’ present the same richness in meanings.
51
Svenbro 2008. For an allusion to the distinction between grammata and stoicheia, see Dion.
Thrax, Techne 9: ‘There are 24 grammata, from Æ to ø. They are named grammata because they are
impressed through drawing and scratching (Øa e ªæÆÆE ŒÆd ı ÆE ı F ŁÆØ); for among the
ancients grapsai meant to scratch, as in Homer ‘Now you boast of having scratched the flat of my foot’
(Il. 11. 388). They are also called stoicheia because they are in file and order (a b ÆPa ŒÆd  ØåEÆ
ŒÆºEÆØ Øa e åØ  Eå ØÆ ŒÆd  Ø)’. On this passage see Lallot 1998: 45 and 95–6; Svenbro
2008.
52
Svenbro 2008, and further Lallot 1998: 96, 98. The clearest distinction between the two is in schol.
Dion. Thr. 197. 17–23: ‘And for this reason he says that they are stoicheia, because they have a file and
order, the ones in respect to the others (Øa e åØ ÆPa <  Eå ØÆ ŒÆd>  Ø æe ¼ººÅºÆ); for in
this case they are stoicheia. When they are not written in order (c ŒÆa  Ø ªæ çøÆØ), they are
called grammata and not stoicheia. As for instance pros, it is written correctly and kata taxin; for this
reason, they are called stoicheia, for as stoicheia, they produce the pros. If, however, I change their order,
writing rpos, they are called grammata and not stoicheia—because they are not anymore kata taxin.’
( x  æ , N f ŒÆƺººø ŒÆd ŒÆa  Ø ªªæÆ ÆØ, Øa  F ŒÆd  ØåEÆ ºªÆØ, I º F Ø ªaæ
‰  ØåEÆ c æ · Ka b Kƺºa ªæ łø æ , ªæ ÆÆ b ºªÆØ, PŒØ b  ØåEÆ, Øa e c
ŒÆa  Ø ÆPa r ÆØ.)
 Introduction 15

meanings comprised by gramma, implies that grammata is used for a written

message, a ‘letter’, simply as a result of the unmarked nature of the term: it denotes
an ensemble of visual signs, not yet deciphered, not yet provided with sound. This
may be taken further to imply that the choice of grammata, when opposed to the
other possibilities for indicating ‘letter as written message’ (e.g. K Ø  º), marks
its writtenness, rather than aspects related to content, conveyance, and meaning.
A number of other terms can designate long-distance written communica-
tion;53 they may be organized in three groups, depending on whether the focus
is on the support of writing, on the writtenness of the message, or on the message
itself.
The first group, based on a metonymic usage, includes those terms referring to
the material support on which the document is written, such as papyrus, leather,
lead, or ceramic, or to its shape (tablets).
The most important in this set is ıº  or غ , deriving from º ,
papyrus. Byblion may have in Herodotus the generic meaning of ‘papyrus-roll’
(compare Hdt. 5. 58.3), but it usually indicates what is unequivocally a letter
(3. 40, 43, 128, here eight times). Throughout classical antiquity, the term was
used for book, papyrus roll, document in general, and also specifically letter; in
Polybius and Diodorus Siculus for instance, ıºØÆçæ means ‘letter carrier’.54
But in general byblion will over time be used more and more with the specific
meaning of ‘book’, ‘book-roll’.
—Æ (‘tablet’), attested in the Odyssey with the basic meaning of ‘wooden
plank’ (12. 67) or ‘dish’ (14. 141, a meaning that would last), is used in the Iliad (6.
169) for the folded support of a written message; it is found in connection with
various kinds of writing also later, to indicate, for example, votive tablets (Aesch.
Suppl. 463), lists (Aesch. fr. 281a Radt), geographical maps (as in Hdt. 5. 49),
paintings (Simon. fr. 178 PMG), and catalogues, such as those compiled by
Callimachus (Diog. Laert. 8. 86, Athen. 6. 244 a; cf. Callim. frr. 429–53 Pf.).55
In its meanings ‘writing tablet’ and ‘letter’, Æ competes with º (and its
diminutive º ), a term of Semitic origin that occurs very frequently in fifth-
century drama to indicate writing tablets, metaphorical or real, but also letters.56
The term was never reserved for a specific type of document.
The same strategy of naming the document from the writing support is used by
the writer of one of the earliest Greek letters, who refers to his  ºØ  (‘small

53
What follows builds on Stirewalt 1993: 67–87.
54
Plb. 4. 22. 2, D.S. 2. 26, and P.Hal. 7. 6 (232 bc), where ıºØÆçæ is found. The ıåÆE ºø
in Aesch. Suppl. 947 may be ‘folds of books’, as in Weir Smyth’s translation, or also, as is more likely,
folded letters (sealed and conveying a decision): see discussion below, 194–7. The meaning of ‘book’ is
attested already in Aristophanes, Av. 974; cf. the discussion below, 242.
55
Chantraine 1999, s.v. In view of some of the later meanings (geographical map or votive tablet) it
is perhaps not accidental that pinax is chosen in the Iliad for a letter marked with ÆÆ ºıªæ ,
terrible, significant signs, but not necessarily alphabetical. In Aesch. fr. 281a Radt. (the so-called ‘Dike’
fragment) the variation between deltos and pinax to indicate the same object may point to the catalogic
aspect of Dike’s reading: the goddess writes down the offences of the mortals on a deltos, l. 21, but when
the moment comes she recites the catalogue of names from the pinax, l. 22.
56
Drama: see below, chs. 3. 3. 1 and 5. 1. Herodotus uses deltos in 7. 239 for the letter sent
by Demaratus to Sparta; the Second Platonic Letter presents itself as a deltos (2. 312d). See Chantraine
1999, s.v.
16 Introduction

piece of lead’), most likely because of the lack of any specific term for letter.57
 ºØ  too is not specific: it is used for instance for the accounts of the temple
of Nemesis in Rhamnous;58 and Pausanias (9. 31. 4) states that the Boeotians
living around the Helicon showed him a ºı  damaged by time, close to the
spring on which Hesiod’s Erga were inscribed.
Similarly, Herodotus affirms that the Ionians used the term ØçŁæÆ (‘skin’) for
the papyrus sheet, because in ancient times, owing to the rarity of papyrus, they
had used the skin of goats and sheep, a statement that has now been confirmed by
one of the earliest Greek letters on lead, in which diphthera is used for commercial
records.59 All these terms of course are not specific to letters; they may serve to
indicate almost any type of written documents, including letters. This kind of use,
where the document takes its name from the support, and where the name does
not give any hints as to the nature of the written document itself, seems to have
been more widespread in the sixth and fifth century bc; it clearly reflects a
situation in which a specific generic terminology does not yet exist (one is
reminded of the similar case of the Solonian axones and kyrbeis). From the fifth
century onwards, however, the vocabulary tends towards a greater specificity, and
documents are named on the basis of their content or their function.
The second group is also non-specific, and maintains this non-specificity; it is
formed by derivatives of ªæ çø.60 ˆæ ÆÆ is of course the most important term
among them: Herodotus uses ªæ ÆÆ for all types of writings, and in particular
for inscriptions; but the term is found for letters in 1. 124 and 5. 14.61 Other terms
of this family can be used to denote long-distance communication. Thus, ªæÆç,
piece of writing, is used in the plural in Euripides (IT 735: æŁ Ø ªæÆç );
the same letter will be referred to with ªæ ÆÆ ten verses later (IT 745).
Thucydides uses ªæÆç for the letter of Pausanias to Xerxes (1. 128. 7); the term
recurs in 1. 129. 1, where it is juxtaposed to K Ø  º. Here the juxtaposition, in
the sentence  ÆFÆ b  ªæÆçc Kº ı, ˛æÅ b l ŁÅ  B fi K Ø  ºBfi (‘So
much the writing revealed, and Xerxes was pleased by the message’) shows well
the distinction between the two terms, one highlighting the visual aspect and
pointing at the letters out of which the message is composed, the other indicating
the import of the message itself, its content. ªæÆç  is used for the copy of an

57
The letter of Achillodoros from Berezan, dated to 550/500 bc, discussed below, 38–9; also
appendix 1, no. 1.
58
Lead sheet, dated to c.500–480 bc (e åæÆ Iº ÆØ e hØæ Øe e K ^Ø  ºı Ø, SEG 38.
13 l. 4): Dubois 1996: 51.
59
Hdt. 5. 58. 3: ŒÆd a º ı ØçŁæÆ ŒÆº ı Ø I e  F ƺÆØ F ƒ  ”ø ; see below, 28 and
n. 23; appendix 1, no. 5; and Chantraine 1999, s.v. ‘ØçŁæÆ’. Note also Eur. fr. 627 K., N d ª æ, N d
ØçŁæÆØ ºªªæÆçE ººH ª ı ÆØ ¸  ı ªÅæı ø (discussed below, 217–18).
60
Overall description in Chantraine 1999, s.v.
61
The Lexicon to Herodotus of Powell lists the following senses of ªæ Æ (34 times altogether):
‘letter of the alphabet’ (14 times); ‘piece of writing’, usually inscription (18 times); in the plural, ‘an
epistle’ (twice: 1. 124. 1 and 5. 14). ˆæÆç is used for ‘painting’ (5 times: 3. 24. 2; 4. 36. 2; 2. 78; 2. 86. 2;
2. 182. 1, the last three instances referring to imitations of a painting) or ‘a picture’ (3 times: 1. 164. 3; 2.
73. 1; 2. 73. 2). ˆæÆÆØ  appears 8 times, as ‘scribe’ (7. 100. 1; 8. 90. 4), ‘secretary’ (3. 123. 1, and of
Persian satraps 3. 128. 3 bis, 3. 128. 5), and ‘treasurer’ (2. 28. 1; 2. 28. 5). ªæ çø refers 7 times to the
author’s own writing; 5 times it means ‘depict’; 18 times ‘write’, ‘write down’, on all sort of supports;
4 times ‘cause to be written’. Instances of IÆ-, I -, Kª-, K Ø-, ŒÆÆ-, æØ-, æ ª-, ıªªæ çø have
been discussed where relevant—they need not detain us here.
 Introduction 17

original writing that may or may not have been a letter, while ªæÆ Ø  (found
for instance in Antiphon) indicates a small note (again, this may or may not be a
letter, depending on the situations).
The connotations of this group change in time, as is perhaps best shown by the
fact that in Herodotus ªæ ÆÆ mainly speak (inscriptions and other written
documents are very often introduced with a formula of the type a b ªæ ÆÆ
ºª ÆFÆ), while in Thucydides they tend to show (ź F Æ); in Euripides, it is
rather the tablet that speaks.62 Overall, the principal connotation of this group seems
to be the focus on the visual and sonorous (once vocalized) aspect of the message, as
well as on the notion of the stability and permanency of written communication.
As for the third group of terms used to denote a ‘letter’, it is organized around
the verb K Ø ººø. The verb, and with it the noun K Ø  º, are attested in
Aeschylus and Sophocles with the meaning of ‘oral order, injunction’ (there is no
certain mention of a letter in the extant works of these authors). In Euripides
however epistolai in the plural is used both for an order transmitted orally (IT
1446, Ba. 442) and for a written communication (IT 589).63 A passage of the
Iphigenia among the Taurians illustrates well the difference between an epistole
and a deltos: Iphigenia says,
Kªg b ÆæH c I   Æ åŁ e | ŁBÆØ Ææ Pb a Ka K Ø  ºa | › 
ººø º  N @æª çæØ (Eur. IT 731–3).
I fear that, once he has left this land, the person who should bring to Argos this tablet
will make no case of my letter/request.
The tablet is the physical object, the stage prop designated with the deictic 
(this tablet here); epistole is, once again, the content of the message (Iphigenia is
afraid that the messenger, once he has left, may simply throw away the tablet with
her message rather than delivering it), but also more generally the request, as
made both through words and in writing. This interpretation is reinforced by the
similar distinction (involving, however, now also grammata) made slightly later in
the same play (Eur. IT 582–94): there, deltos is again the tablet/letter as object;
epistolai in the plural refers clearly to the letter-as-content; while the letter-as-
writing (non-pertinent content, because viewed from the point of view of Pylades,
who simply has to ‘carry its light weight’) is termed grammata.64
A passage from an unknown play of Cratinus (fr. 316 K.–A.) confirms that at
least until the third quarter of the fifth century epistole does not refer to a letter,
but rather to an order: the fragment (¼Œ ı F ŒÆd  c K Ø  º) is cited in
Zonaras’ lexicon (p. 804) in order to support the claim that K Ø  ºc ŒÆd c
K ºc ºª ı Ø, ‘they also call an order epistole’.65

62
The shift in the terminology possibly mirrors a shift from reading aloud (or public performance)
to silent, individual reading.
63
On the semantic evolution of K Ø  º in tragedy see Stéfanis 1997: 169–91. The use of letters in
the Iphigenia plays of Euripides is discussed below, 224–35.
64
The passage is discussed in detail below, 225.
65
Intriguing ancient discussion in the Lexicon of Orus, fr. A 42 (a fragment partly preserved in
Zonaras’ Lexicon, s.v.  ¯ Ø  º): K Ø  ºc ŒÆd c K ºc ºª ı Ø. ˚æÆE · ¼Œ ı F ŒÆd 
c K Ø  º (fr. 316 K.–A.). ŒÆd › —º ø (Tim. 71 d)· Å Ø B  F Ææe K Ø  ºB ƒ
ı  Æ A . In the apparatus, Alpers points to the Synagoge lexeon chresimon (746 Cunningham
18 Introduction

A situation similar to that of tragedy is found in historiography: in Herodotus

K Ø ººø may be used for an oral or a written message; but K Ø  º is used only
of oral injunctions.66 In Thucydides, besides a few instances in which it is
impossible to decide, there are some very clear cases in which K Ø  º (or the
plural K Ø  ºÆ) refers to (the content of) a written message (e.g. 1. 129, discussed
above).67 Only with Xenophon does K Ø  º systematically mean ‘letter’, with
the singular referring to one document only, and the plural to an ensemble of
documents.68 From the fourth century onwards, epistole will be the normal term
to indicate a letter (and manuals on how to write letters will bear names such as
 Ø K Ø  ºØŒ ).
The examination of the terminology allows three conclusions. First, until the
fourth century there is no specific term for ‘letter’: throughout the fifth century,
the most specific terms are grammata and epistole. The first refers to letters and
other types of writing (Herodotus, for instance, uses grammata of inscriptions);
the second to written orders as well as oral injunctions. When a writer wants to
make it clear that he is talking of a letter, the solution that tends to be adopted, in
the case of deltos and pinax, but also with grammata, is to draw attention to the
closed, ‘folded’ aspect of the missive, for instance via the use of the adjective
ıŒ , ‘folded’, or a periphrasis such as º ı ıåÆ, ‘folds of the tablet’; the
seal functions as an indicator, stressing, at the same time, the secretive or
deceptive connotations of this mode of communication.
Second, it is possible to distinguish further between the two terms grammata
and epistole: the former emphasizes the visual, external aspect of the actual
writing; the latter puts the stress on the content. In turn, when the material
support of the written message plays a role, authors employ the metonymic
group of terms (deltos, pinax). Consequently, the distinction between ‘letter’ and
‘epistle’ argued for by some modern scholars, who wish to define the former as the
direct expression of the writer’s feeling and the latter as an elaborate, literary
document, has no basis in the ancient terminology; more generally, such a
distinction does not seem to fit the way the ancients conceived of epistolary
communication.

= Bachmann Anecd. 232. 16) = Suda  2632 = schol. Plat. epist. 337 e: K Ø  º· K º, K  ŒÅłØ ; sch.
Vet. Soph. Aj. 781: K Ø  º · K º (and note the fascinating schol. rec. Soph. Aj. 781ab:  ’
K Ø  º · ªæ çÆØ ŒÆd K Ø º . I e B K d æ Ł ø ŒÆd  F Kºº ÆØ. b. › b " OÅæ K Ø ºc
ºªØ, ƒ b æ Ø K Ø  º , K º , ÅÆÆ); schol. Soph. OC 1601: K Ø  ºa · K º , æ Ø ;
Hesych.  5255 K Ø  ºÆ· K ºÆ, K ØƪÆ. `N åº —æ ÅŁE. Øƪæ ÆÆ. ŒÆd K Ø  º Ø ; Eust.
in Il., 29. 15; 312. 6; Thom. Ecloga nominum et verborum Atticorum,  121. 1 (K Ø EºÆØ b P   e
Øa ªæÆ ø, Iººa ŒÆd e ÆP æ  ø ŒºF ÆØ. ŒÆd K Ø  ºc ‰ Æø .) See also below, 241 and
n. 192.
66
Powell has two instances where K Ø ººø means ‘send a letter’ (3. 40. 1 and 7. 239. 4); 5 where
it means ‘command’ (6. 3bis; 4. 131. 2; 6. 97. 2; 7. 223. 1); K Ø  º always means ‘injunction’ (4. 10. 1;
6. 50. 3).
67
The Index Thucydideum of Bétant lists nine instances of K Ø  º, and four of K Ø  ºÆ in the
plural; discussion below, 136–50.
68
Cf. for K Ø  º: Xen. Hell. 1. 4. 3; 1. 7. 4; 7. 1. 39; Anab. 1. 6. 3 ter; 3. 1. 5; Cyr. 2. 2. 9 ter; 2. 2. 10; 4.
5. 26; 4. 5. 34; 5. 5. 4; 8. 2. 16; Ages. 8. 3; for K Ø  ºÆ: Cyr. 8. 2. 17; Ages. 8. 3. See discussion below,
150–5.
 Introduction 19

Third, the fact that the term gramma is used to indicate ‘letters’ cannot be taken
to indicate a strong correlation between writing and epistolarity: gramma is
simply the unmarked term.69 The terminology with which a narrative or the
very text of a letter describe the letter are thus important, as they can give pointers
to the specific aspect being highlighted in the ‘reading’ of the document.
Two ancient riddles on writing will provide a fitting closure to this discussion.
They do not concern epistolary writing; but they give important clues as to
the shifting and controversial place that writing—and written communication—
occupies in the Greek imaginary. Both come from the fourteenth book of the
Anthologia Palatina, a book specifically on riddles. The answer to the first one
(Anth. Pal. 14. 45) is ‘wax’: ‘I am black, white, yellow, dry, and wet; and when you
spread me on the plains of wood, by Ares and by the hand, I emit a voice, although
not speaking’.70 The second one (Anth. Pal. 14. 60) concerns ‘writing tablets’, and
is built on much the same images as the preceding one: ‘Wood gave birth to me
and iron reformed me; | I am the mystic receptacle of the Muses; | when shut, I am
silent, but I speak when you unfold me; | Ares alone is the confidant of my
conversations.’71 In both riddles, Ares points to the use of iron for the stylus, and
to its ‘violent’ action in incising the wax; the notion that writing speaks while
being at the same time silent is present in both, although exploited in different
ways (note the presence of the same verb, ºÆºø); the second riddle adds the
Muses, hinting at the possibility of preserving memories offered by writing.
‘Mystic’ (ı ØŒ) strikes here an odd, slightly dissonant chord: it may refer to
the initiation needed to learn writing, but references to the difficulty of writing are
something very rare in Greek literature: writing was a craft among others, not
particularly connected with initiation into the Mysteries.72 A better alternative is
to take ‘mystic’ in connection with the Muses, and to refer it to poetic initiation, or
also to understand the term as marking the separateness, the privacy, of the
written text itself, once the tablets are folded.

69
The distinction between litterae and epistula in Latin seems to have run on almost opposite lines:
according to Gavoille 2002, epistula refers mainly to the letter as an object, or to its specific form
(author, addressee, style, length); litterae is the result of an enunciation, it refers mainly to the ‘words’
(the content). Nadjo 2004 offers a study of other Latin terms for ‘letter’.
70
¯Nd ºÆ , ºıŒ , ÆŁe Åæ  ŒÆd ªæ · | s b  ıæÆø ø o æ KÆ fi Å , |
@æœ ŒÆd ƺ fiÅ çŁªª ÆØ P ºÆºø. Interestingly, a hexameter followed by an elegiac distich (an
unusual structure), for a text built on opposites.
71
 #ºÅ   Œ, ŒÆØ æªÅ  b Åæ · | Nd b  ı ø ı ØŒe KŒåØ · | ŒºØ Å
تH· ºÆºø ’, ‹Æ KŒ  fi Å , | Œ Øøe e @æÅ  F  å ı Æ ºªø. Note also the difficult
riddle Anth. Pal. 14. 24, concerning probably two tablets (the equivalent of the two matrices men-
tioned) on which the stylus (the father who presides over memory) has written (‘generated’) a word
(‘panther’) divided over the two tablets. Another such riddle (P. Lond. Lit. 63 = Pack2 1765) is discussed
by Di Marco 2009.
72
There is however in the fifth- and fourth-century Athenian context a marked emphasis on
writing in the context of mystery cults: cf. Scodel 2011.
 Part I
Greek Beginnings
Writing and Letter Writing, Evidence
and Representations

ÆPaæ › fiÅ |  ¯ººØ çø Æ ŒÆd çæ Æ HæÆ Œ Çø | ªºÅ

ZæªÆÆ  F  › Łæ Æ, ı çı  b | ±æ Å  ØåÅe K ¼ÇıªÆ ÇıªÆ
Æ | ªæÆe Iت  Ø    æÆ تB.
Nonn. Dionys. 4. 259–64
But Cadmus, bringing gifts of voice and thought for all Hellas, fashioned tools
to echo the sounds of the tongue, and mingling sonant and consonant in one
order of connected harmony, he rounded off a graven model of unsilenceable
silence.
 2
Writing and Letter Writing
The Evidence

At the beginning was the letter: it is often assumed that writing came into being as
the substitute, the transcription, of an oral message that had to be sent at a
distance.1 The earliest Mesopotamian account of the invention of writing, a
composition known as Enmerkar and the Lord of Aratta, part of a group of
epics concerning the beginnings of Sumer and revolving around the conflict
between two cities and their rulers, is a good example of this assumption.2 The
reason for the invention is in this story the fact that the messenger travelling
between the city of Unug and Aratta, carrying to and fro the challenges posed by
Enmerkar and the counter-requests sent in form of riddles by the Lord of Aratta,
could not repeat one of the messages, because it was substantial and extensive, or
too deep and difficult. Because of the impossibility of the tongue of the messenger
to repeat such a message, Enmerkar ‘patted some clay and wrote the message as if
on a tablet. Formerly, the writing of messages on clay was not established. Now,
under that sun and on that day, it was indeed so. The lord of Kulaba inscribed the
message like a tablet.’3 This account of the invention of writing will be further
expanded upon by the story of the invention of the envelope by Sargon, also an
early story; while in the first millennium, in Berossus’ account, writing will be
presented as having been given to mankind, together with all the other crafts ‘and
those things conducive to a settled and civilized life’, by a monster sent by the

1
An instance of this ‘conventional wisdom’ is the story ‘How the First Letter Was Written’, in
R. Kipling, Just So Stories (London, 1902), discussed in Harris 1986: 1–3. For the distance between
spoken and written language, see Glassner 2003: 223–6; Cooper 2004; Robertson 2004 (esp. 19).
2
Versions of these epics (not however of the one concerning the invention of writing) are attested
already in texts of the twenty-sixth century bc from Abu Salabikh, while one text from Ebla of the
twenty-fourth century bc shows that these narratives enjoyed a wide geographical diffusion: Glassner
2003: 11. Scholars agree on the fact that even if the tablets themselves date for the most part from the
Isin-Larsa period (2017–1763), the Ur III period (2112–2004 bc) must have represented an important
moment in the formation and evolution of the cycle: Vanstiphout 2004: 1, 13–14; Black 1998: 23–4 and
n. 57.
3
Enmerkar and the Lord of Aratta, 501–6. On the story, see Vanstiphout 2004; Glassner 2003:
15–17; Klein 2000; Ceccarelli 2002. The reception of the message by the Lord of Aratta (‘The lord of
Aratta looked at the tablet. The transmitted message was just nails, and his brow expressed anger. The
lord of Aratta looked at his kiln-fired tablet’, in the ETCSL translation) is also extremely interesting,
since it points at the ability of writing (and in particular cuneiform writing) to transmit more than just
one meaning.
24 Greek Beginnings

gods, Oannes, at a time preceding the flood.4 In the new version, writing is no
longer connected with communication at a distance, but rather linked to memor-
ialization, something that ties in well with the changed historical and social
circumstances;5 as we shall see, an oscillation between commemorative and
communicative functions is also present in Greek traditions.
Yet, if we move from the Near Eastern representations of the invention of
writing to the actual documents, the first clay tablets in cuneiform writing are lists
and economic or administrative documents.6 Nor are letters present among the
first literary texts, which begin to appear some four hundred years after these
documents. The first dated letters come from Ebla, and were written more than
eight hundred years after the invention of writing.7 Unsurprisingly, there is a gap
between representation (or discourse) and reality.
And in Greece? Starting with the last quarter of the eight century bc, graffiti
attest to a personal and widespread use of writing in the ancient Greek world;
but alphabetic writing seems to have served initially mainly for dedications or
funerary inscriptions, and then, slighly later, for recording monumental texts,
that is, for public texts, or at any rate, texts circulating openly. This corres-
ponds rather neatly to the way in which most Greek traditions on the invention
of writing characterize the new craft: although a letter fleetingly and ominously
figures in the Iliad, and although the earliest reference in Greek literature to
something written thus concerns letter writing, the document is there presented
as a (rather enigmatic) given, not worthy of discussion. Narratives of the
discovery of writing, from the late sixth century throughout the fifth and fourth
centuries, sharply differ from ‘conventional wisdom’ in presenting the new craft
either as something that simply ‘arrives’ from the outside, not earmarked for any
specific use, or also as something that is invented to allow organization (hence the
importance of numbers), to enable recording, and to facilitate poetic composition.
Letter writing enters the field of representations only in the second half of the fifth
century, and in a rather negative way: harking back to the Iliadic occurrence,
letters (and more generally communication at a distance) are marked as fraught
with dangers (Palamedes), oriental, tyrannical, feminine (Atossa, Deioces). In
this, Greek narratives differ remarkably from the Near Eastern accounts of the
invention of writing: the communicative aspect is much less in evidence.8 The
late entrance and negative connotation of letter writing in Greece has most
likely to do with the very different socio-cultural context. Let us look at this in
more detail.
Writing is already attested in the second millennium in the pre-Greek civiliza-
tion of Minoan Crete, with the still undeciphered Cretan hieroglyphs and the
slightly later Linear A, which is also not yet fully understood. The two types of

4
So in the account by Berossus, FGrH 680 F 1 = Verbrugghe and Wickersham 1996, F 1; see
Glassner 2003: 10.
5
Michalowski 1999.
6
For the dynamic relationship between the Sumerian bureaucratic organization and the tool it
produced, writing, see Green 1981; Englund 1998; Glassner 2003: 1–2; Cooper 2004; Veldhuis 2006.
Robson 2007 introduces numeracy into the discussion.
7
See Michalowski 1993: 1–7, for an introduction to, and a selection of, early letters from
Mesopotamia.
8
See also below, 63–71.
 Writing and Letter Writing—The Evidence 25

writing overlap in part: the first is attested for 2000–1600 bc, the second for 1850–
1450 bc. Most documents appear to concern transactions. Sealstones and ac-
counting tablets seem to be the main types of documents in Cretan hieroglyphics.
Linear A too appears to have been used for accounts; it is attested mainly on Crete,
but also in some places on the Greek mainland and on other islands.9 As for letter
writing: if one of the most recent attempts at deciphering the Phaistos disc were to
prove correct, the fired clay disc would contain a Luwian letter to Nestor.10 But the
decipherments on offer are now legion. Moreover, if interpreted as a letter, the
main interest of this document for a study of epistolography in ancient Greece
would lie in the fact that it would show familiarity—a contact—with the epistolary
genre at an early time, in an area later inhabited by Greeks; but such a familiarity
can be inferred anyway from the well-attested contacts between the Minoans and
later the Mycenaean Greeks on the one side, and the Hittites and Egyptians on the
other.
Based on Linear A, a syllabic script called Linear B was developed to notate
Mycenaean Greek.11 Documents have been found both on Crete and in numerous
localities of mainland Greece; on Crete they cover the period from c.1400 to 1250
bc, while on the mainland tablets written in Linear B are found until c.1200 bc.
The documents we have concern mainly administration (accounts, catalogues,
and lists are frequent), as well as religious or military issues; importantly, they are
all temporary records, preserved only because of the accidental destruction of the
archive by fire; no legal documents such as contracts, registrations of ownership,
or loans have as yet come to light, nor are there any documents attesting to a
broader use of writing. As Palaima states, ‘there are no Mycenaean literary,
judicial, historical-propagandistic documents, no “private” economic records, no
personal or official letters, and no documents relating to the scribal training or
intended for scribal reference such as syllabic abecedaria, lexical lists, or sample
forms of documents’.12 There is some controversy as to the possibility of the
existence of writing on non-durable materials. The discovery of remains inter-
pretable as wooden folding tablets in the palace of Pylos adds some weight to
the argument for the existence of writing on parchment—based among other
things on the cursive form of the script, more suited to writing with a pen or

9
The most detailed survey of writing, from the Cretan pictographic to the Greek alphabet, is
Heubeck 1979; although slightly outdated in respect to the numerous recent finds, it is still very much
worth consulting. Uchitel 2004 gives an informed overview of Linear A documents; Schoep 2007
recognizes the mainly economic scope of the tablets and the administrative aspect of the use of
sealstones, but prefers to stress the symbolic and ideological connotation of writing as expression of
a legitimation strategy by the elite.
10
Achterberg et al. 2004 (letter of the Great King Tarhundaradus of Arzawa in Anatolia to Nestor
king of Pylos: ‘At Mesara is Phaistos. To Nestor, to the great man in Ahhiyawa . . . ’. Others however
(e.g. Timm 2005) interpret the text in connection with Linear A. General comments on attempts at
decipherment in Duhoux 2000. Almost nothing is certain about the disc: not the date, which oscillates
between 1850 and 1600 (if the stratigraphic report of the excavator, Pernier, is to be trusted) to as low as
1400 bc, nor even its authenticity, according to Eisenberg 2008.
11
Bennet 2008 offers a masterful survey of the problems linked to the transition from Cretan
hieroglyphic and, more importantly, Linear A to Linear B, and provides a synthetic statistical overview
of the function for which Linear A and B writings were used.
12
Palaima 2004: 154.
26 Greek Beginnings

brush and ink or paint than with a stylus on clay; but such writing might have
served simply as a further stage of administration.13 There is also uncertainty as
to the status of the scribes. Some of them were probably members of the elite,
and it is possible to think that in some cases at least they were in charge of
those activities they registered: they would have learnt the craft of writing as a
consequence of their administrative and social position. But when the activ-
ities documented by one specific scribe concern sectors too diverse for it to
be plausible that one person only was in charge of them, then those scribes
will have been trained functionaries, providing administrative support to some-
one else.14
With the collapse of Mycenaean civilization in the twelfth century, writing
seems to have disappeared. Cyprus forms a notable exception: here, by the middle
of the eleventh century bc a syllabic script is attested, modelled on the Cypro-
Minoan scripts present on the island since at least the sixteenth century bc. This
script was used throughout the history of the island, until the fourth century bc,
when it was replaced by the Greek alphabet. Elsewhere, however, no writing on
durable material has been found, although some form of recording on perishable
materials may have continued for some time.
When we again find written documents, they are written in an entirely
different script and present very different characteristics, in content, typology,
and diffusion. Just as had been the case for both the Linear B and the Cypriot
syllabary, which were modifications of the Linear A script, the Greek alphabet
derived from a non-Greek writing system, the North Semitic Phoenician
script. The new technique spread quickly, and by the end of the eighth century
most areas inhabited by Greeks used some variant of the alphabet.15 Local
variants persisted for a relatively long time, and the Greek traditions concern-
ing further additions and refinements brought to the alphabet by various
culture heroes (such as Palamedes, Epicharmus, and Simonides, who was,
interestingly enough, also known for having invented mnemotechnics!) are
clearly a reflection of this diversity. But what was writing used for? Why was it
adopted, what needs did it fulfil? What were its social implications? And what
of letter writing?

13
Tablets: Mylonas Shear 1998; further administrative synthesis on perishable materials, after
which the clay tablets would have been reemployed: Palaima 2004: 170–2. Contra: Bennet 2001.
14
Scribes as members of the elite and performers of the administration: Bennet 2001; as simple
accountants: Palaima 2004: 175–7. The comparison with the much better known status of the scribes at
Ugarit (Rainey 1968) is instructive.
15
Excellent survey of scripts and writing around the Mediterranean between 1200 bc and 800
bc in Amadasi 1998. For Greece, Jeffery 1990 (with Johnston’s supplement) remains invaluable;
very detailed survey of preceding literature in Heubeck 1979: 73–105. See also Bartonek and
Buchner 1995; Wachter 1996; Lazzarini 1998; Wirbelauer 2004; J.-P. Wilson 2009; Woodard 2010
(as well as his earlier publications). On Kalapodi, Palme-Koufa 1996; on Kommos, Csapo et al.
2000, and Johnston 2005: 385–6. The finds from the Eretrian colony of Methone (191 graffiti, 25
of which are alphabetical, dating between 730 and 700 bc) have been published in Bessios
Tzifopoulos and Kotsonas 2012. For the variants between scripts as a form of expression of
local identities, see Luraghi 2010.
 Writing and Letter Writing—The Evidence 27

2 . 1 . W R I T I N G IN A R C H A I C G R E E K S O C I E T Y :
WHO WRITES WHAT, AND WHY?

Why alphabetic writing took off so quickly is still an open question. Letters are
not, however, among the earliest preserved documents from the Greek world. The
first letters to have come down to us can be dated to c.550 bc, that is, more than
two hundred years after the date of the first preserved written documents; for the
time being the first extant written texts, apart from single letters which may have
been numerals, trademarks, or signs conferring on the object a quasi-magical
power, consist of names and ownership inscriptions, dedications, and verses. The
emphasis on ‘extant’ is important: the possibility of written texts on perishable
materials such as papyrus or leather cannot be excluded. Still, while keeping in
mind that documents on perishable materials might profoundly change the way
we think about the early uses of literacy, any sensible discussion has to start from
the preserved documents.
The first epigraphical documents (dating to the eighth century) are spontan-
eous short inscriptions on vases: painted inscriptions, or graffiti, incised after the
firing of the vase. The first graffiti on rock appear slightly later. The vast majority
of these earliest inscriptions consists in statements of ownership, including names
of deities and offerings to a god. But there are also abecedaries, individual names,
curses, acclamations, insults, or short texts expressing pride in the writing. (The
signature of the person who made or painted the pot is a variant of this: Pithecusae
offers an instance dating to the end of the eighth century.) There are also some
longer texts in verse: thus, the famous Cup of Nestor from Pithecusae, possibly an
object to be used in a sympotic context, although it was found in the tomb of a
young boy; the similar (albeit very fragmentary) hexametric text on three lines
inscribed on a Rhodian cup of the same type, from Eretria in Euboea; the inscrip-
tion on the Dipylon oinochoe from Athens, a prize in a dancing context; and an
extremely fragmentary poetic text on a geometric oinochoe from Ithaca.16 There is
in all this no suggestion of a centralized use of writing. The contexts of provenance
bear this out: these early documents come mainly from necropoleis and sanctuar-
ies, and more generally from public areas.17
Based on this evidence, two theories concerning the invention of the alphabet
have been advanced. Wade-Gery long ago proposed that alphabetic writing (as
opposed to consonantal writing) had been invented in order to write hexameter
verse, and more specifically the Homeric poems.18 Such a view relies on the fact
that a fairly high proportion of the earliest Greek inscriptions consists of metrical
texts; indeed, the argument involves the hypothesis that it was the complexity of

16
Respectively Jeffery 1990: pl. 47 n. 1 (CEG 454); Johnston and Andreiomenou 1989; Jeffery 1990:
pl. 1 n. 1 (CEG 432); and Jeffery 1990: pl. 45 n. 1 (CEG 453). Comprehensive survey of Greek early
inscriptions, from the beginnings to c.650 bc, in Powell 1991: 119–80.
17
Survey of the evidence in Powell 1991: 181–6 (with the corrections of Wirbelauer 2004: 193–4); a
thoughtful overview in Thomas 1992: 52–65. See also Jeffery 1990; Lazzarini 1998; Davies 2005;
J.-P. Wilson 2009: 549–56. Wirbelauer’s interpretation (that these early inscriptions have to do with
mobility, because of the mobile character of the support, and so with communication at distance) does
not seem to make much sense.
18
Wade-Gery 1952: 11–14; Robb 1978; Powell 1989; 1991; 2002.
28 Greek Beginnings

Greek song, with its metrical patterns, that brought the adapter to realize the
advantage of a move from a consonantal script to a fully alphabetic one, which
also noted the vowels. The alternative theory stresses the fact that the contact
between Greeks and Phoenicians must have taken place in a context of exchange
and commerce. According to this hypothesis, the Greeks who took up the script
from the Phoenicians would have used it, at least in the beginning, in the same
way as the Phoenicians did, for commercial purposes.19 The difficulty here is that
there are no extant early commercial documents or letters; the inscriptions
making statements of ownership offer only a weak argument in favour of the
theory, since they are found on very simple pots and never on more significant,
decorated vases, where claiming ownership might have been relevant.20 More
compelling is the evidence of graffiti on commercial and storage amphorae,
securely attested at Pithecusae (twenty out of a total of forty-six inscriptions):
here writing seems to have been used (in a severely abbreviated form) to mark the
type of product, or the quantity.21
Thus in both cases the argument rests, at least in part, on the hypothetical
existence of texts now lost. And indeed, some documents point to the existence of
other documents written on perishable material. An inscription such as the one on
the so-called ‘Nestor’s cup’, dated to c.725 bc, seems to presuppose the influence
of a book script, both because of the alignment of the verses (their mise en page)
and because of the signs separating metrical units.22 Herodotus (5. 58. 3) affirms
that the Ionians ‘from ancient times’ call the papyrus sheets ‘skins’ (a º ı
ØçŁæÆ ŒÆº ıØ), since due to the lack of papyrus they formerly used the skins
of sheep and goats. This statement is confirmed by a lead letter of c.500 bc,
mentioning records on skins, as well as by the proverb, preserved in the Suda and
in Diogenianus, ‘you say things more old-fashioned than skin’ (IæåÆØ æÆ B
ØçŁæÆ ºª Ø). Finally, two Hesychian glosses point to the same direction,
although specifically linked only to Cyprus;23 a term transcribed ØçŁ æƺ ØçH
is attested in a Cypriot syllabic inscription, with the sense of ‘skin-writer’.24
Bearing all this in mind, what can we say about the type of use that was made of
writing, and by whom? In order to recover the uses to which writing was put, we
have to rely on the extant documents, but we must also contextualize them in the

19
Heubeck 1979: 150–2; Lombardo 1988; Harris 1989: 60–1; Millett 1991: 259–60.
20
Buchner, in Bartonek and Buchner 1995: 139.
21
See Buchner, in Bartonek and Buchner 1995: 140–1, 144–5, as well as the discussion of the
specific examples by Bartonek at 165–76; J.-P. Wilson 2009: 548–9.
22
Heubeck 1979: 115; Cassio 1998: 67–8, with further bibliography.
23
Ample discussion in Heubeck 1979: 152–4. The letter: Johnston, in Jeffery (1990: 430); Wilson
1997/98 (SEG 48, 1012); Dana 2004; below, app. 1, no. 5. In the Suda Æ 4076, the proverb is followed by
the following comment:  ªaæ ØçŁæÆ Æ ºÆØ , K fi w  Œ E › Z f I ªæç ŁÆØ a ª  Æ
(compare Diogenianus, v. 1, cent. 3, 2, in Leutsch–Schneidewin, Corpus paroemiographorum Grae-
corum; slightly different version in Diogenianus v. 1, cent. 4, 95a). Hesychian glosses: Hesych. Æ 2842
Iº Ø æØ · ªæÆç E . ˚æØ Ø (‘paint-brush: pencil, the Cypriots’);  1992 ØçŁ æƺ Øç·
ªæÆ Æ ØŒÆº  Ææa ˚ıæ Ø (‘anointer of skins: teacher, among the Cypriots’).
24
For ØçŁ æƺ ØçH see Masson, ICS, no. 143 (epitaph of a schoolmaster, fifth or fourth century
bc); and for the second element see the bronze tablet of Idalion dated c.470–50 bc, Masson, ICS, no.
217. 26 Nb a() º  () a ϝØjÆ  NƺƺØ Æ<() (‘and this tablet, written with these
words’), with p. 243 (my thanks to L. Holford-Strevens for this reference).
 Writing and Letter Writing—The Evidence 29

society that used them to surmise what their possible functions were.25 This
implies accepting the possibility of a fairly high differentiation in the use of
writing, depending on the social and socio-economic structures in place in the
various areas of the Greek world; it also implies distinguishing between the type of
use the adapter may have had in mind, and the uses to which the Greeks
immediately put the new craft; it is the latter that interests us. In this context, it
is worth bearing in mind that the dichotomy which scholars have posited between
‘merchants’ on the one hand and ‘aristocrats’ interested in poetry on the other
may have been pushed too far. It is likely that the adapter of the alphabet was
involved in trade—he must have been in close contact with Phoenicians, and
presumably for some time, in order to realize what could be done with their script;
it is therefore indeed likely that he may have wanted to put it to the same use as the
Phoenicians. As Willi has recently pointed out, the Phoenicians appear to have
used writing mainly for dedications in sanctuaries, and in a lesser measure for
ownership inscriptions and funerary inscriptions; the situation in the Greek world
is very similar (the prominence of dedications is even more striking if we consider
that many ownership-inscriptions can be interpreted as—or double up as—votive
inscriptions).26 Writing will have increased the prestige of dedications to the gods,
made initially by those traders who had entered into contact with the Phoenicians,
and then by others. But long-distance traders were part of what we might call an
‘aristocracy’; when not travelling, they will have spent some time dining together
in what would later be called a symposion, or dancing, and they may have used
their recently learned skill in such contexts to inscribe jocose texts such as Nestor’s
cup, while also inscribing marks on trade- and storage-amphoras, making dedi-
cations in local or Panhellenic sanctuaries, and possibly registering items on
perishable materials.27
Let us now extend the chronological range and pay attention to location. The
epigraphical material of the period from the eighth to the early fifth century offers
a rich and varied picture: the media change, in the sense that besides inscriptions
on ceramic there are now numerous texts inscribed on stone, on the walls of
temples for instance, or on statues. Modifications in the relation between writing
and the spoken word are also observable. Early writing appears to have been used
in service of the spoken word: most of the early texts are statements meant to be
read aloud.28 However, from the mid-sixth century at the latest, the forms taken

25
This is why, although it is possible that a complete copy of Homer was written down, and
although it is possible that the desire to write poetry may have exerted an influence on the change from
a consonantal to a fully alphabetic script, it nonetheless does not make sense to assume that writing
down the epos was the driving force behind the diffusion of writing: it is difficult to figure out the uses
made of it (a point forcefully argued by Cassio 1998: 75–82).
26
Willi 2005, giving as instances of how the Greeks might have got in contact with this practice the
dedications of a bronze horse blinker in the Heraion of Samos and of a horse frontlet in the
Daphnephorion of Eretria, both bearing the same Aramaic inscription; and stressing that the first
Greek writers cannot have expected other fellow-Greeks to understand their writing, a situation that
makes of gods the ideal readers. See also Hall 2002: 94–6.
27
This ‘soft’ version of the commercial origin of the alphabet is argued in detail in Lombardo 1988.
28
See Thomas 1992: 62 and 73; Pucci 1988; Svenbro 1993: 26–43 and 44–63, with the remarks of
Porciani 1997: 160–2; Calame 1993; Magini 1998: 23–4; and for a somewhat different take on the
‘speaking objects’, Carraro 2007.
30 Greek Beginnings

by funerary and dedicatory inscriptions change: the frequency of first-person,

‘egocentric’ inscriptions (‘I am the cup of Nestor’) diminishes, while third-person
inscriptions become more and more frequent, implying possibly that by now
writing can be used in an impersonal way, as a self-standing instrument, a ‘third
person notice of information’.29 Much of this early writing may have had a
symbolic power, rather than an informative purpose; it has been argued that
this remained the case in a large measure also in the classical and Hellenistic
period.30 Finally, the contexts change as well, including now the everyday private,
the sacred, and the monumental/public. Thus the period between the early
seventh century to the early fifth century sees, in various degrees of intensity
and in different places, the monumental inscription of laws and decrees (at
Gortyna or Dreros in Crete, at Tiryns in the Argolid, in Chios); archival uses of
writing (the ‘contract’ of Spensithios in Crete, stating that he shall be responsible
for recording for the polis); commercial uses of writing (the maritime loans, if that
is what they are, from Corcyra; contracts and letters on lead by emporoi); uses
related to the prestige of writing (under this heading go most vase-inscriptions,
whether signatures of artists or jocose mottoes, descriptive labels, signs alluding to
writing, all having at the same time a decorative function that should not be
forgotten); dedications representing monumental elite statements, such as the
inscriptions on the bases of kouroi and korai, usually in verse, as well as small
dedications, on which writing itself may be the distinctive element; funerary
inscriptions, again mainly in verse (widespread throughout the Greek world);
and, from the mid-fifth century, lists and catalogues.31
In a ground-breaking paper, Stoddart and Whitley argued that it was possible to
identify two models for the contexts and uses of writing in the archaic Greek
world.32 In the first, best exemplified by Attica, personal inscriptions, such as
onomastic graffiti, dipinti on pots, dedications, and inscriptions on tombstones by
far outweigh official inscriptions such as laws, regulations, or decrees. This would
suggest a relatively widespread literacy. The other model is the Cretan one: there,
the earliest inscriptions are mainly official documents on stone, such as laws, while
personal inscriptions are comparatively rare.33 Such a situation they interpreted as

29
Thomas 1992: 63–4. An instance of writing taking centre-stage is offered by a law from Olympia
dated to c.525–500 bc concerning the behaviour of the theoroi (Minon 2007, no. 4 = van Effenterre and
Ruzé 1994–5, no. 109 = I.v.0. 7): the regulations are followed by a clause, ‘but if anyone judges against
the writing, the judgement should be void’ (ÆN  Ø aæ e ªæç  ØŒ Ø, I º Œ Y ± ŒÆ, l. 2):
now that it has been written down, the regulation is referred to as ‘the writing’; in the next paragraph,
the possibility of making changes in the regulations is discussed, and the regulations are referred to as
writings (^  ŒÆ ªæÆç  . . . , l. 3). The change is also examined, in another perspective, by Day
2010: 112–16; Stehle 1997: 311–16 offers a splendid discussion, focused on the communicative
situation.
30
Linders 1992.
31
The searchable site Poinikastas <http://poinikastas.csad.ox.ac.uk/>, based on the work of Anne
Jeffery, makes it possible to visualize the differences in document type, provenance, type of support,
and archaeological context. Focus on location and support is at the centre of Davies 2005.
32
Stoddart and Whitley 1988. See further Whitley 1997 and, for a wider contextualization, Whitley
2001: 128–33.
33
Note, however, that in their calculations, Stoddard and Whitley 1988 (as also Whitley 1997)
omitted to look for Cretans abroad: the graffiti of Cretans mercenaries c.450 bc on the Memnonion at
Abydos are evidence of personal literacy. See Chaniotis, in SEG 47, 1377.
 Writing and Letter Writing—The Evidence 31

pointing to a rarefied literacy, and to an elite use of writing, made to impress

symbolically, rather than to offer a democratic means of control. Most regions of
the Greek world, and in particular Euboea, Corinth, the Argolid, Laconia, and
possibly Boeotia seemed to follow the Attic pattern, while the situation in Elis
appeared to be closer to the Cretan one.34 New discoveries have modified the
picture: the finds from Kommos for instance have more than trebled the number
of graffiti from Crete.35 But the more detailed analysis proposed ten years later by
Whitley, taking into account the new finds from Crete (then still unpublished),
reached very similar conclusions.36 According to Whitley, in Attica writing would
have been relatively widespread, used by craftsmen, but also by aristocrats com-
peting for prestige, Ø ; in Crete, writing remained the preserve of a small,
restricted group. An alternative explanation of the difference between the two
areas, based not on differences in the degree of literacy, but rather on ‘epigraphic
habit’, the type of use made of writing, has been advanced by Chaniotis: the
Cretans tended to make a public use of writing, while the epigraphic habit of
archaic Athens would have been predominantly private and individual.37 Further
refinements should be brought to bear on the model: for instance, it has recently
been shown that the legal texts publicy displayed on temple walls in Crete rely
(exception made for the earliest group of texts) on archival documents; the early
presence of archives need not however be interpreted as proof of widespread
literacy, but rather as pointing to the existence of a ‘legal literacy’.38 The two
explanations actually can be meaningfully combined; the important point here is
that even within Greece, different groups made very different uses of writing: this
applies also, as we shall see, to letter writing.
Besides the epigraphical evidence, it is important to survey the literary evidence:
not so much for what it explicitly says about the uses of writing, as for what it
betrays in respect to the diffusion and the uses of literacy. As suggested above, the
orderly arrangement of the letters on the ‘Cup of Nestor’ seems to imply the
existence of a book-script; the first redactions of the epic poems cannot have been
much later than the seventh century, and a written support has also been
hypothesized for the composition of Greek melic poetry. Similarly, we hear that
around the mid-sixth century Heraclitus wrote a book in prose, which he

34
Hall 1997: 145.
35
Thus the tables in Stoddart and Whitley 1988: 763–4 need adjustments, as do also the otherwise
more detailed tables offered in Whitley 1997: 641 (Attica); but even with the adjustments, the difference
between the two areas remains evident. See now also J.-P. Wilson 2009: 559–61.
36
Statistics: for at the period 750–480, Whitley listed for Attica 758 dipinti, 240 graffiti, 82 inscribed
gravestones, 395 dedications, but only 8 legal texts, all relatively late in date; for the same period, Sparta
offers 79 dedications, no graffiti, no dipinti, 1 inscribed tombstone, and no laws; Crete has 19
dedications (of which 13 on armours), 13 graffiti, no dipinti, 5 tombstones, but 38 laws! The more
than 50 graffiti presumably by made by shepherds around 500 bc, found by Langdon on a hill in the
area of Vari in Attica (e.g. Langdon 2005), and comprising an abecedarium, but also short texts, further
reinforce the picture.
37
Chaniotis, in SEG 47, 1997, no. 1377, for whom the internal references in Cretan laws to the
written text imply that the texts were meant for reference, rather than as monuments.
38
Faraguna 2011 (an important discussion, touching on other regions of Greece, besides Crete);
distinction between the earliest group of Cretan texts and the others in Faraguna 2011: 9 and 12, argued
also on the basis of the layout (in a long continuous line for the earliest group, in columns recalling the
layout of a book-roll for the later texts—see on this also above, n. 22).
32 Greek Beginnings

deposited in the temple of Artemis in Ephesus, thus dedicating it to the goddess

(Diog. Laert. 9. 6); and ancient tradition dates the first libraries, owned by the
tyrants Pisistratus and Polycrates, to the second half of the sixth century (Athen.
1. 3a). In all these instances, writing functions mainly as a means of preserving
records and gaining prestige by specialized groups, such as rhapsodic guilds, or
exceptional individuals, such as the tyrants; it may also have served as an aid to
composition, for instance in the case of lyric poetry; but it still has no role, or only
a very restricted one, in the communication of the work itself.39
Can this situation be used to support the idea of the existence of ‘letters’, of
messages addressed to an absent person, and written on a perishable support?
While letters are not securely attested until c.550 bc, Rösler has argued that traces
of poetic epistolography can be found already in Archilochus, and that such a
form of poetic communication was well established by the time of Alcaeus.40 For
Archilochus, Rösler pointed at the famous epode 185 W. (Kæø Ø’ o Ø Ær  , t
˚ÅæıŒÅ, Iåı ÅØ ŒıºÅØ): the poem opens on an act of communication
(‘I shall tell you a story’), followed by an apostrophe to a ‘son of heralds’ (Kerykides,
a comic patronymic), and by a problematic ‘sorrowful message-stick’.41 Whatever
specific interpretation we may want to give of the opening and of the ainos that
followed it, clearly there is here a play on official communication. However, it is also
evident that this need not be an official written communication (actually is hardly
likely to be, unless we insist that the skytale necessarily implies a written mes-
sage)—and the poem was certainly meant for oral communication. Much less
clear seems to me the case of Archilochus fr. 89 W., also considered by Rösler as
an instance of poetic epistle, because of the distance between the ‘speaking I’ and
the addressee of l. 29, Erxies.42 The pragmatic situation in the poem is indeed that
of long-distance communication; but to move from this to a conception of the
poem as a real letter sent (in writing, as supposed by Lasserre-Bonnard, or also
orally) to Erxies in a situation of danger seems an unwarranted leap.
As for Alcaeus, two poems, fr. 401B V. (the ‘shield’ poem) and fr. 130b
V. (a poem composed from the exile) have been interpreted as ‘poetic epistles’.43
A passage of Herodotus (5. 95) shows that at any rate in the fifth century bc fr.
401B V. was indeed considered a message, written and then sent by Alcaeus to
Mytilene, to inform Alcaeus’ hetairos Melanippus of the events in Sigeum; Strabo
(13. 1. 38), after explaining that in the poem Alcaeus addressed a herald, asking
him to bring the news to those at home, proceeds to quote two badly corrupted
verses: ‘Alcaeus is safe (@ºŒÆ   ), but the Athenians hung up (his armour? the
shield that was his protection?) in the holy temple of grey-eyed Athena’ (trans.
Campbell, fr. 428a). Indeed, Alcaeus is here spoken of in the third person, as

39 40
Rösler 1980: 45–56. Rösler 1980: 272.
41
Problematic, for both textual and semantic reasons. Textual: we may have here a dative, or a
nominative: if a dative, as West, then we might imagine the speaker holding a (metaphorically sad)
baton and delivering his message; if a nominative, it may refer to the Kerykides, and qualify the types of
messages the ‘son of heralds’ tends to spread, or to the speaking voice, which would be presenting itself
as a sorrowful messenger. Semantic: we do not know what exactly skytale means: baton giving the
authority of official messenger to its carrier; baton bearing incisions, and functioning as aide-memoire
for the messenger; message (at any rate not encrypted) written on leather and rolled around a stick, as
in later Spartan use. See West 1988 for discussion; and below, 322, n. 66.
42 43
Rösler 1980: 274; cf. Lasserre-Bonnard 1958: 26. Rösler 1980: 272–85.
 Writing and Letter Writing—The Evidence 33

would be the case with a message. As for fr. 130b V. (= 130b Campbell), it opens
with a first-person speaker who is in exile, and is addressing a friend named
Agesilaidas. The most likely scenario is to think of a mixture of ocular and
imaginary deixis, with either a present ‘Alcaeus’ addressing an imaginary Agesi-
laidas, or an absent ‘Alcaeus’ (to be imagined by the audience) addressing,
through the speaker who trasmits the message, an Agesilaidas, present and visible
to the members of the hetairia. In this second case, it would be possible to speak of
an ‘epistle’.44 Even if this were accepted, however (and it is not clear that the
apostrophe and the situation described in Alcaeus fr. 130b V. can be interpreted
along these lines), these ‘poetic epistles’ need not imply actual letter writing; they
rather reflect the obvious fact that individuals who for some reason had to spend
time away from their home will, through travellers, merchants, or soldiers, have
sent messages home telling of their situation.45
Pursuing a different approach, Stehle has made a case for an interpretation of
some of Sappho’s poems (mainly Sappho 31 V., but also 96 V., 22 V., and 94 V.) as
detached from the performance context, and consciously exploiting writing as a
means to preserve communication in the face of separation. For Stehle, the
separation of speaker and singer in these poems means that the poems were
written to be sung by another person, the addressee, to herself, thus setting up a
communication between two otherwise separated figures; writing offered Sappho
a way of ‘creating a fictional ‘I’ not recuperable by the singer and using it to
simulate communication’.46 This is very different from what happens in Archilo-
chus’ and Alcaeus’ poems: in their poems, meant for performance, markers of
epistolarity are present; but in Sappho, no such markers are to be found, so that
independently of whether one is inclined to accept Stehle’s interpretation, Sap-
pho’s poems cannot be taken to imply anything in connection with letter writing.
The notion of apposing a sphragis, a seal, to one’s songs—or writings—is also an
important development, and in this case its reflection on letter writing is evident: a
letter is a written message by a specific individual, aimed at another specific
individual or group and at no other persons, conveying the exact words of the
sender, without any alterations introduced by the messenger. But for the recipient
to be certain that the message indeed comes from the stated sender, in a situation
where the handwriting cannot give a real clue, a sign is needed; and indeed, either
documentary letters were sealed with an imprint recognizable by the addressee, or
a ‘sign’ was written into the text, a reference to something known only to sender
and addressee.47 A sphragis is what gives authority to the transmitted message.
The apposition of a ‘seal’ in a poetic text is first explicitly stated in the poetry of

44
Rösler 1980: 273–85. Of course in later (re)performances the deictic referents would have become
imaginary—see Edmunds 2008: 82–4 for the continuum along which ‘historical I’ and ‘generic I’ are
situated.
45
D’Alessio 2004: 279–80 stresses that the deictic results of mediated communication may find
expression equally in oral and written mediated communication; a similar point had been made by
Calame 1995: 52–6, and 74 n. 35. See the discussion of the Pindaric ‘poetic epistles’ (Pythian 2 and 3,
and Isthmian 2) in Gibson and Morrison 2007: 4–7; and for a detailed discussion of Pythian 3, Young
1983 (esp. 31–4).
46
Stehle 1997: 288–311 (311 for the quote), and 314–15 for ‘epigraphic’ communication (discussed
below, 57 and n.128).
47
See Parsons 1980 for instances in papyrus letters.
34 Greek Beginnings

Theognis;48 but the sphragis may have already existed in the seventh century, as
one of the constituent parts of the citharodic nomos.49 The next occasion when the
practice is recognizable is in the sayings of Phocylides and Demodocus, and in the
elegiacs inscribed on herms by the Athenian tyrant Hipparchus: all three inserted
their own name just before the saying (Phocylides and Demodocus right at the
beginning of the elegiac distich; in the case of the texts on the herms, the hexameter
was probably used to indicate the geographical distance, while the name of Hippar-
chus and the proverb were located in the pentameter).50
On the whole, the notion of a ‘seal’ affixed to poetic or prose texts may be seen
as closely connected to the early uses of writing in epigraphical contexts: for even a
cursory look at early dedications (or epitaphs) or at the earliest potters’ inscrip-
tions on the eighth-century sherds from Pithecusae shows that the mark of the
individual affirming ownership or (extraordinary) craftsmanship was a consti-
tutive part of the first uses of writing. The practice by potters of signing their vases
or by sculptors of asserting their craftmanship is also well attested in the seventh
and sixth century.51 Note also the extraordinary seal on an island gem, a scarab
from Aegina bearing the inscription ‘I am the sign of Thersis, do not open
me’ (¨æØ j K Ø A Æ j  ¼ j ت ).52 What makes this text special is the
injunction ‘not to open’ whatever it was that was sealed with it: sealing here is not
just a statement of ownership, it also conveys privacy. All these examples show
that by the sixth century the notion of permanency in connection with writing had
developed in such a way as to render appealing the notion of expressing owner-
ship and stability within poetic compositions through the use of a sphragis. Letter
writing may have played a role in this: after all, letters are folded—and ‘sealed’. But
its role (and the role of writing itself) need not have been particularly important,
and for more than one reason: first, the appending of one’s own name (and, in
commerce, the adding of marks indicating content and ownership, as well as of a
seal, for instance on amphorae) had been a feature of writing from the earliest
beginnings; second, the seal functions perfectly well in an oral context, where,
besides the authority, it emphasizes (as in the transmission of an oral message) the
distance of the author.53

48
Theogn. 19–38, probably forming a unit, opens with: ‘Cyrnus, let a seal lie upon these verses for
me as I communicate poetic wisdom—they will never be stolen without detection, nor will anyone
accept a worse line when a good one is at hand. But everyone will speak thus: “These are the verses of
Theognis of Megara”’ (vv. 19–23: ˚æÅ,  çØÇ ø fi b K d çæŪd KØŒ Łø |  E  Ø—
º  Ø  h  Œº  Æ, | P Ø Iºº Ø ŒŒØ   PŁº F Ææ  , | z b A Ø Kæ E·
¨ FªØ KØ Å |  F  ªÆæø). The discussion on what exactly that seal might have been,
and what it may have meant, is still open; cf. Calame 1995: 25–6, 49–50; Porciani 1997: 51; Fain 2006;
Hubbard 2007, whose translation I cite.
49
See Kranz 1961: 16–17 and 27–8, mentioning the sphragis at the end of Timotheos’ nomos, but
reporting the practice back to the songs of Alcman; note also the use of the sphragis in Critias, fr. 3
G.–P. = 5 W.2, and Noussia-Fantuzzi 2010: 478–9 on Solon fr. 30 G.-P. = 36.20 W.2 (20, ªæÆłÆ).
50
Kranz 1961: 23; Hubbard 2007: 203–4. Instances and discussion of the early use of a sphragis in
prose works (the proem of Alcmaeon of Croton, D.–K. 24 B1, the reconstructed ones of Heraclitus and
Ion of Chios, that of Democritus, D.–K. 68 B 299, and others) in Porciani 1997: 47–53.
51
References in Kranz 1961: 20.
52
Agate scarab from Aegina, now in Breslau, Boardman 1968: 73–4 and fig. 2, no. 176 (IG iv 179;
SEG 40, 301).
53
For the ‘entrance’ of commercial notions into the poetic vocabulary of the sixth century cf.
Svenbro 1984: 155–79; but also the more sceptical view of Ford 2002: 132–9. For the distance
 Writing and Letter Writing—The Evidence 35

It thus seems fairly certain that messages will have been sent when needed,
sometimes in writing, but mainly orally through messengers (someone would
anyway have had to transport the writing); and that some of these messages, both
the oral and the written ones, will have been ‘sealed’, that is, marked with a clearly
recognizable sign of origin; the notion of communicating at a distance was
certainly not unknown in archaic Greece. However, there is no basis either in
the epigraphical material or in the literary documentation for seeing in long-
distance communication the earliest or the main use of writing, no matter whether
that communication was by individuals or by communities.

2.2. THE EARLIEST LETTERS

Let us now focus on the earliest written messages. In what follows, I propose to
trace the formation of a Greek epistolary format, from the earliest written
messages to the form defined by the epistolary treatises as standard, and current
from the end of the fourth century bc onwards. I hope to show how messages that
were initially informal communications couched in writing evolved towards a
fixed form with stable rules, perceived as such.54 For this it is necessary to trace the
changes in the enunciation, as well as in the technical vocabulary, within the
corpus formed by the early letters; but in order to follow the letters’ progress
towards an ‘epistolary specificity’, I shall also attempt to situate their enunciation
and language against that of a related set of documents, the defixiones, or curses.
These too may take the form of a message to the gods of the Underworld (some
curses explicitly present themselves as letters); moreover, the material support on
which they are written is also the same as that of most early letters.

2.2.1. Epistolary Formulae and Deictics: The Development

of the Epistolary Genre

From the fourth century onwards, a letter is a piece of writing introduced by a

prescript containing a greeting formula (the most common being ›  E  fiH  EØ
åÆæ Ø), followed by a message in the first person, and ending with a concluding
wish, such as erroso/errosthe (‘Farewell’) or similar.55 The fact that the greeting
formula remained in use almost unchanged for some 700 years, even if already by

introduced by the seal, and the fact that its presence in writing is simply a reflection of its functionality
in an oral context, Porciani 1997: 55. For oral messages, see below, ch. 4.
54
On the relationship of oral and written text see Day 2010: 22–3, discussing Bakker 1993: 16.
Letters are indeed substitutes for speech, but are mainly concerned with the transmission of the
message rather than with its preservation (as in the case of epigrams); still, the reading of a letter is a
reactivation of the oral message (see also Head 2009).
55
In its basic structure, the letter can be compared to a personal encounter: the prescript corres-
ponds to the initial greeting, the body of the letter is the conversation, the concluding wish marks the
separation (Koskenniemi 1956: 155). Cugusi 1983: 28–30 offers a collection of passages of ancient
authors listing the rules concerning the structure of the letter.
36 Greek Beginnings

the first century ad the construction of the greeting with the infinitive åÆæ Ø was
not understood and was felt to be problematic by the grammarians, shows how
strongly marked by tradition letter writing was.56 But for the earlier period, things
are not so simple. Some archaic and early classical literary texts, such as the poem
of Alcaeus concerning his shield or some Pindaric epinicians, were understood (or
could be understood and conceptualized) as letters. These texts may have contrib-
uted to shape what will become the accepted format of a letter; but they are far
from representing ‘standard’ letters.
Besides literary texts presenting themselves as messages, or containing embed-
ded letters, there are quite a few documents of the archaic and early classical
period that clearly constitute letters, attempts at written communication over a
long distance. In this area our knowledge has changed dramatically in the past
thirty years. Up to 1976, the year in which the most famous of these letters, found
at Berezan in the Black Sea, was published, only a handful of documentary letters
dating to before 350 bc were known.57 Studies of the form of the ancient Greek
letter had thus to rely on letters embedded in literary works; on the so-called ‘letter
of Darius to Gadatas’, a Greek version of a royal letter (originally written in
Aramaic?) sent by Darius I to his servant (the Greek has  Fº , ‘slave’) Gadatas,
an important regional administrator (possibly a satrap, or the steward of a satrapal
or royal paradeisos); and on an abundant papyrological documentation that,
however, did not go back in time before the third century bc. Embedded letters
will be discussed in the next chapters; as for the ‘letter to Gadatas’, the document
we have was actually engraved in the second century ad, and while it may well be a
copy of an earlier document, not enough is known of its transmission to justify
making use of it for a study of the formal evolution of Greek letter writing.58 But
since 1976 the situation has changed: the number of documentary letters, on lead
or on ceramic sherds, has more than trebled. The corpus of archaic and classical
documentary letters now reaches forty-two, and it is likely that more will come to
light. These letters come from all over the Greek world: eight are from Attica (four
on ceramic sherds, four on lead) and two (on lead) from Chalkidike, while the
remaining letters have been found in peripheral areas of the Greek world:
the Black Sea region, Sicily, and the area around the Gulf of Massalia. Nine letters

56
Koskenniemi 1956: 157; see Gerhard 1904 and Koskenniemi 1956: 41–2 for the problems with
the prescript felt by Dionysius and Apollonius Dyscolus; translation and detailed commentary of
Apollonius On Syntax 3. 63–77 (329–41 Uhlig) in Lallot 1997: i, 229–33 and ii, nn. 151–86.
57
Parsons 1980: 13 counted three (nos. 1, 12, and 39 of the list given in appendix 1); and indeed, the
other instances then known (25; 26; 30; 31; the four graffiti from the Athenian agora, 35–8; 40) were all
either not fully published, or fragmentary, or not typical.
58
The literature on the document (I. Magnesia 115; M.–L., no. 12) is extensive. Most scholars
consider it authentic: see e.g. Porciani 1997: 25–33, Fried 2004: 108–119, Lane-Fox 2006, Tuplin 2009,
Muir 2009: 84 (‘The earliest genuine piece of state correspondence we have from the Greek world’),
Lombardi 2010, as well as Briant 1996); but a strong case against the authenticity has been made by
Briant (2003), who argues that the personnel of the temple of Apollo at Magnesia forged the document;
see also Gauger 2000: 205-11; Grabbe 2004: 116–17. Importantly, even those who argue for the
authenticity add qualifications. Thus, Porciani 1997: 32 considers it as a Greek witness of Persian
epistolography; Fried 2004: 116 concludes: ‘Rather than assuming that it is a literal translation of an
Aramaic or Persian original, I think that it may represent the Apollo priests’ tradition of Darius’s letter
to Gadatas inscribed “as it must have been”’; Lane-Fox 2006 leaves open the possibility of insertions
and changes; and Tuplin 2009 notes that the concept of ‘authenticity’ must in this case be elastic.
 Writing and Letter Writing—The Evidence 37

(seven on lead, two on ceramic fragments) come from the latter area, while the Black
Sea region has yielded twenty-two documents, fourteen on lead, the rest incised on
ceramic sherds.59 Most of these documents refer to commercial transactions, some
others concern smaller exchanges, and a few are personal communications.60
In the following discussion I shall not distinguish sharply between letters
written on lead and letters written on potsherds. It is true that the differences
between the two types of support have an impact on the message itself and its
meaning. Thus, lead may be more or less available, depending on the area one
lives in;61 it can be easily stored, but it can also be reused;62 it can be rolled up,
making it possible to send a ‘closed’, sealed message, with the names of sender and
addressee usually written on the outside; and it offers a better surface than a
potsherd for a longish or complex message meant for long-distance transmission.
Conversely, pottery is cheap (almost any sherd could do) and can be found almost
anywhere; but it is not that easy to store, so it would probably have been thrown
away after it had served its purpose; it is heavier and more cumbersome, thus less
convenient for sending a message over a longish distance; more importantly, a
graffito or dipinto on ceramic cannot be sealed. These differences affect not so
much the writing in itself as the type of messages sent and the distance at which
they are sent: for communication with one’s own slaves or dependants, or from
the polis to the chora, a ceramic fragment would probably serve most needs; lead
seems to have been employed for longer distances, for complex affairs, or in
specific situations (such as in the case of the letter of Lesis from Athens, no. 41: he
was working in a forge and so lead may have been easier to obtain than ceramic;
and he probably would not have wanted the content of his message to reach his
master). Apart from these differences, from the point of view of the enunciative
markers the two sets are comparable; tackling them together will give us a more
rounded picture of the uses of writing for personal communication in the archaic
and classical period.

59
List in appendix 1, with main bibliography on each document; surveys in Jordan 2003b, and most
recently in Eidinow and Taylor 2010, who view letters mainly as responses to situations of crisis. It
should be noted that some of the documents classified as letters by Dana 2007 (and before her by
Vinogradov 1998) stretch the limit of what one is willing to accept as letter (the ‘letter’ from Sicily,
appendix 1, no. 32, is also problematic in this respect); moreover, some of the documents included in
Dana 2007 have not yet been published—they may turn out not to be letters. When in doubt I have
tended to be inclusive.
60
Excellent discussion of the use of letters, and writing more generally, in the world of long-trading
merchants in the archaic and classical period, in Lombardo 1988. I know of only one official letter on
lead, dated to Hellenistic or imperial times, from Acra (listed at the end of appendix 1).
61
See Kroll 1977: 94–5 n. 29 for references to ancient and modern sources on the use of lead, and for
indications as to the price of lead in various areas of Greece at different times; a cost of 2 to 7 drachmas
per talent (c.25 kg) of lead implies a very low cost for the single sheet of lead (Kroll estimates that the
650 third-century tablets from the Agora and Kerameikos may as a group, in terms of the material
used, have cost between 1 and 2 drachmas).
62
This means that we mostly have letters that did not reach their addressee—for otherwise, after
having served its purpose, the lead would have been reused: Jordan 2000: 92–3 and n. 6, citing
Vinogradov 1971: 95–6. The remark of Bravo 1974: 113–16, that we do not find lead letters after the
fourth century bc because by then papyrus had become more affordable while being at the same time
easier to handle, does not really affect Vinogradov’s argument, notwithstanding what Jordan seems to
imply. On the reuse and storage of lead documents see also Kroll 1977: 94–5.
38 Greek Beginnings

Let us start with the letters from the Black Sea. The most famous and probably
the most ancient Greek documentary letter is a text inscribed on a piece of lead
coming from Berezan (appendix 1, no. 1), dated between 550 and 500 bc. It has on
the outside the names of both sender and addressees (a main addressee, and a
second one to whom the first one should communicate the message), as well as a
reference to the object itself, designed as ‘molibdion’, piece of lead: åغº æ - e
º- | Ø  Ææa e ÆEÆ | ŒIÆƪæÅ), ‘Achillodoros’ piece of lead, to his
son and Anaxagoras’.63 The use of ºØ  may mean that Achillodoros did not
know of a word for ‘letter’; but it should be noted that even when a specific term
became available, letters could be referred to with words denoting the type of
material used.64 As for the message itself, it opens with an address in the vocative,
followed by a text in which the writer speaks of himself constantly in the third
person (t —æøƪæÅ, › Æ æ  Ø Kغº -. IØŒ Ð ÆØ | e Æı , Œº.); the
second person singular is used for the addressee, already in the first sentence,
which fulfils the role of a prescript.65 Trapp aptly remarks that ‘this perhaps
reflects a feeling that sending a message by letter is like sending one via a living
messenger, who would naturally report the sender’s wishes thus’; he also advances
an alternative, in his opinion less likely explanation, namely that the hidden first-
person speaker may be the scribe who wrote the message for Achillodoros, and
who would naturally speak of Achillodoros in the third person.66 A further
possibility would be to imagine that the letter speaks by itself, like the archaic
inscriptions on ‘speaking objects’ studied by Svenbro.67 On the whole, the first
interpretation (words of the messenger) seems the most likely. The letter reports a
complicated affair concerning the seizure of goods—in fact, Achillodoros, the
sender, is at risk of being himself seized as a slave by a certain Matasys in a
procedure known as ıºA, by which Matasys, whose property (slaves and houses)
has been illegally taken by Anaxagoras, seizes in compensation Achillodoros
himself and the goods, both supposedly belonging to Anaxagoras. Kept through-
out in the present tense, that is, written from Achillodoros’ temporal perspective
(later, letter writers will commonly use the so-called ‘epistolary past’, i.e. they will

63
Bravo 1974: 175 and Trapp 2003: 198 both note that the initial predicative position of the genitive
possessive conveys ownership rather than authorship; the point is reinforced by the syntax (Ææa and
not e Ææa) and by the lack of sandhi (the non–conversion of the final  of ºØ  into , expected
before a ), which implies a fractional pause between the two terms. The piece of lead was to remain in
the family, for further reuse.
64
Trapp 2003: 198; see above, 15–16. This applies also to other categories of written documents: for
instance, in the very few instances in which the text of a curse refers to itself, it does so as ‘lead’, Eidinow
and Taylor 2010: 43 n. 55. The text of a decree ‘is’ the stele on which it is written.
65
Note the use of Kغºø (l. 1 and 9) in a letter, to mean ‘enjoin’. It is interesting to contrast the
address in vocative here (and in some other documentary letters) with Sykutris’ statement (1931: 188)
that in a literary text addressed to a person one finds the addressee’s name in the vocative in the first
sentence (he cites Isocrates’ Philip as an instance), ‘but that is not done in a letter’ (‘das findet sich aber
in einem Brief nicht’). Bravo 1974: 175–6 offers a close analysis of the syntax of the letter, which he sees
as poised between direct and indirect speech.
66
Trapp 2003: 196–7. The suggested comparison with the letters in Herodotus, however, does not
work, because after the introductory formula in the third person the Herodotean letters present an
enunciative shift to the first person: see below.
67
Wilson 1997–8: 37; Svenbro 1993. The question of ‘who wrote’ (as distinct from ‘who is sending’
and ‘who is speaking’) will be addressed at the end of the survey.
 Writing and Letter Writing—The Evidence 39

write the message from the point of view of the addressee), the letter offers an
interesting example of dramatization, with a ‘speech within the speech’: for in
explaining the situation, Achillodoros quotes the very words with which Matasys
justifies his act of appropriation. There is no closing formula.
Another letter from Berezan (appendix 1, no. 2, dated to c.540–535 bc) begins
Ææa[ - - ]: the names of either sender or addressee must have followed directly.
This is a fascinating document, but extremely fragmentary: items should be sent
and given back, a slave-girl has to be sent to the addressee by a certain Melas,
speed is important, but it is difficult to put all this together. While the wording the
prescript is uncertain (a verbum dicendi is present, but its interpretation is
disputed), it is at least fairly clear that the body of the letter was expressed by
the sender in the first person (cf. l. 2, Łº Ø Ø); the end is lost. One further
interesting feature of this letter is that, unlike the previous one, it is written
boustrophedon. In general, there does not seem to be much uniformity in the
writing styles of the lead letters from this area, something that would point to
individuals incising their message themselves (or asking someone else, but not
necessarily a professional, to do so). In this specific case, the editor, Vinogradov,
comments that the letter is written carefully and in beautiful letters, by a hand he
considers as accustomed to writing on stone.68
An important change in the enunciative posture appears in a long letter also
coming from the Black Sea area (appendix 1, no. 5), and dated to c.500 bc, which
bears on topics very close to Achillodoros’ letter. The letter has on the outside
the names of sender (nominative) and addressee (dative), while inside the order
is reversed (addressee first and then sender, ‘To Leanax, Apatorios’). Neither
outside nor inside is there any trace of a verbum dicendi or of a greeting. This
form of address presupposes a discourse in the third person, but except for the
prescript the letter is entirely written in the first person and remarkably articu-
lated. We are here once again (as with Achillodoros and Matasys) in the world of
long-distance traders; here too, Apatorios writes to Leanax because his goods
have been seized (غŠÆØ) by Herakleides son of Eotheris; a certain Menon,
who must be somehow related to Leanax (possibly one of his agents), has,
however, come in his help; Menon now says that if Leanax could send to
Herakleides and Thathaies (or possibly Thathaie) the records (ØçŁæØÆ), his
(Apatorios’) goods could be saved. Clearly all these people know each other; and
clearly not only do they keep in touch, whenever necessary, by sending lead
letters, they also have registers. Even more interesting is the fact that here the
written record seems to count for more than the oral evidence of a witness.69
Here too, there is no closing formula.
Another early letter on lead, dated to 530–510 bc and coming from ancient
Phanagoria (appendix 1, no. 4), shows that often the written text served only
to support the authority of a message transmitted orally: neither sender nor
addressee are mentioned, although a first person plural towards the end of the

68
Vinogradov 1998: 154. Bravo 2007: 58–66 now suggests that this is not a letter, but an Orphic text
meant for the funeral of a female initiate; Chaniotis, SEG 58, 988, notes that the meagre remains of the
text do not allow certainty.
69
Noted by Nieddu 2004: 35 n. 101, who refers further to Burkert for the interpretation of
diphtheria as ‘master-books’ in contrast with single sheets of lead serving as letters, molubdion.
40 Greek Beginnings

text shows that this is indeed a letter.70 The text opens with a deictic indication:
‘This slave here has been sold out of (or ‘in’) Borysthenes, his name is Phaylles’; it
then continues with the statement ‘we wish that everything . . . be returned’. One
way to make sense of this text is to imagine that the document was given to the
slave himself (or to a special messenger, as suggested by Vinogradov), and that the
slave was supposed to show it to the addressee, possibly a new owner. According
to this scenario, the slave had been bought to be given to the addressee, so that
(in exchange) the senders might recover all they wanted returned (or settle their
debt).71
A similarly perplexing instance, where the status and purpose of the written text
is unclear, is offered by a later document, this time incised on ceramic, and coming
from the chora of Olbia (appendix 1, no. 11). The text is as follows: ‘Nikophanas
son of Adrastos gave Zopyrion a horse; let it be sent to me in the city, and let the
document be given to him.’ Here, although we have names, it is unclear who the
sender and the addressee are. It is usually assumed that both Nikophanas and
Zopyrion are in the city, and that Nikophanas is sending the message to a member
of his family or an employee, whose name is omitted, to ask them to send the
horse, and to ask for a written acknowledgement.72 However, in this scenario the
usefulness of the written acknowledgement is unclear; it might make more sense
to assume that Zopyrion is the writer, and that he is sending his receipt to a
member of Nikophanas’ family in the chora (Kozyrka, where the letter has been
found) to get the horse, in exchange of which, the receipt will be given back to
Nikophanas. The point is that whatever the details of the affair, the enunciative
shifts within the text make it very difficult to know who is who; supplemental oral
information would have had to be conveyed for the exchange to work.
Smoother is the text of a letter dated to the fifth century (probably to its end),
coming from the small city of Kerkinitis near Olbia, and written not on lead, but
on an amphora fragment (appendix 1, no. 8). After the name of the sender and
that of the addressee in the dative, we have the expected shift to the first person.
This document concerns local affairs (something that ties in well with the use of
the cheaper—and non-sealable—ceramic fragment): the sender, Apatorios, asks
Noumenios (most likely a dependent of his) to bring in the dry fish and some
planks; the addressee is moreover required not to take any initiative, but to take
care of the oxen, and to control something (taxes possibly) concerning the
Scythians. There is no closing greeting formula.
From the fourth century onwards, letters from this region begin to present the
typical epistolary prescript and final farewell clause. A farewell clause is certainly
present on a lead letter from Panticapaeum, dated to c.400–350 bc (appendix 1,
no. 10); the beginning of the text is fragmentary, so that it is impossible to know
whether the name of the sender was followed by that of the addressee only, or also

70
The absence of sender and addressee was noticed and commented upon by Vinogradov 1998:
162, who compared with the letter on an ostracon from Ligurian Olbia (no. 31) and with the ostraka
from the chora of Olbia (no. 11) and from Gorgippia (no. 13).
71
It may also have worked as a kind of receipt, although it is difficult to see how it could be of much
use without the names of the owners, sender, or addressee.
72
Cf. Dana 2007: 79–81.
 Writing and Letter Writing—The Evidence 41

by a greeting åÆæ Ø; the latter seems more likely, given the valediction ææø .
But for the prescript, the letter is throughout written in the first person.
A standard opening appears also in a lead letter from Olbia dated to the mid-
fourth century bc, sent by a certain Artikon to the people back home (æØŒH
 E K YŒøØ åÆæ Ø, appendix 1, no. 12). After the greeting, the letter turns
directly to practical issues: the sender fears that his family may be ejected from the
house, and suggests a few alternative places they might move to, adding at the end
an instruction concerning a share of wool. A certain unease with the letter format
may be seen in the fact that the expected enunciative shift to first person speech
after the greeting is nowhere to be found (any deictic references to the sender are
avoided); the letter consists in injunctions to an addressee in the second person
plural. Moreover, the letter still has no final greeting clause. The same is true of a
slightly later letter from Nikonion (no. 20), which otherwise features a greeting
formula followed by shift to the first person, statements concerning the health of
the writer, his son, and a friend, and some specific requests.
The other documents from the Black Sea area are either too fragmentary, or do not
clearly belong to the category of letters, or both; but the texts examined illustrate
sufficiently well the slow move from a message formed by an address and a body
entirely written in the third person of the sender to a message comprising prescript
and clause, and a body expressed in the first person of the sender. The addressee is
always in the second person, with one possible exception (appendix 1, no. 7, where
the odd third person could be the addressee). Remarkably, even by the end of the
fourth century there is no stable format. What is particularly interesting, however, is
that the early messages (the two letters concerning the procedure of sulan) are very
elaborate, complex texts: if they do not present the structure of letters, it is evidently
because such a structure did not exist. The later messages are less elaborate, and seem
to concern mainly local affairs; yet by then the use of a specific formulary has
imposed itself, and so the writers make use of it.
At this point, however, we might want to ask whether sender and writer are the
same person, and if so, whether it was the sender, or the writer, who made use of
an epistolary format. This question, if put in general terms, is almost impossible to
answer. Some criteria may give clues: for instance, well-formed letters, a flowing
ductus, and the absence of word-division at line-ends within the body of the text
are usually interpreted as indicating the work of a professional. Of the letters from
the Black Sea, appendix 1, nos. 1, 4, 8, 12, and 18 do not have word-division at the
end of the line and nos. 10, 13, 20, and 21 tend to it; nos. 2, 5, 7, 9, 11, 15, and 17
have word-division at line-ends; nos. 3, 6, and 14 are too fragmentary for it to be
possible to assess line-ends, while nos. 16, 19, and 22 are still unpublished.
However, interpunction marks in the form of a semicolon placed more or
less regularly are found both on letters that do not have word-division at the
line-end (nos. 4, 12, and 18), and on letters with word-division at line-end (nos. 2,
3, and 5). The case of the complex letters of Achillodoros (no. 1) and Apatorios
(no. 5), two letters belonging to the same world, may be used to further test these
criteria. The first letter is difficult to follow because of the constant use of the third
person for the sender; it does not have word-division at line-ends. Imagining that
someone else wrote the letter for Apatorios would explain the syntax neatly; and
the absence of word-division might speak for the hypothesis of a scribe. The
second letter is much clearer than the first because of its use of first person
pronouns for the sender—and it has word-division at the line-ends; if a
42 Greek Beginnings

professional scribe was used here, he is invisible. However, first persons appear also
in letters which do not present word-division at line-ends (nos. 4 and 8); more
importantly, Achillodoros and Apatorios both belong to a world that made use of
registers (the diphtheria mentioned in no. 5): most of these long-distance traders
probably could write, but with varying levels of speed and proficiency, and if they
could not write themselves, they will have had on board someone who could, again
with differing degrees of proficiency. In general, the character and import of most
of these texts are such that it would not have made sense to look for someone to
scratch them. Thus, on the whole, I should think that in most instances the sender,
or someone close to him, and not necessarily having an elevated position because of
this ability, did write the text.73 One last point should be made: the answer to this
question of writer vs. sender is important only for what concerns the spread of
literacy. From the point of view of the definition of an epistolary style, the question
of who wrote is less important: because through receiving letters and through the
contact with people who wrote, even someone who could not write himself could
have absorbed the format and repeated it when dictating a message.
Let us now move to the letters from the West.74 They are all very fragmentary,
and fairly divergent restorations have been proposed; yet on the whole the picture
they present is comparable to that offered by the letters from the Black Sea region.
The oldest document is a letter on lead from Emporion, dated to the very end of
the sixth century bc, or to the beginning of the fifth (appendix 1, no. 23). It is a
long letter (fourteen lines of text remain), involving a number of parties (the
addressee, who is to go to Saiganthe; the Emporitai; someone called Basped-, a
non-Greek name; and possibly another person, unless it is Basped- who is referred
to with a pronoun) and complex deals (the addressee has to buy twenty of
something, as well as wine; Basped- has to deal with buying a ship; the cargo
has to be transported; the ship may have to be towed; something may be shared).
The beginning is lost, but the end is preserved, and it presents in the last line a
verb in the first person singular of the perfect tense, ‘I have ordered’, followed by a
greeting (Œ Œº ıŒÆ. åÆEæ ). The presence at the end of the message of a greeting
with a form of åÆæ Ø is extremely interesting: in the later formal codification
åÆæ Ø will appear in the prescript, while a form of Þı Ø, such as ææø or
ææøŁ , will be reserved for the farewell. Here the position at the end, at the place
which will be occupied by the imperative ææø , and just after the strongly
assertive first person ‘I have ordered’, may help explain the second person impera-
tive; åÆEæ is also typically found at the end of the Homeric hymns. As for the
presence of a first person perfect, Œ Œº ıŒÆ, it does not allow any certain
conclusions as to whether in the body of the letter the sender referred to himself

73
Differently Jordan 2000: 93; but his corpus comprised only nine letters, and most did not have
word-division at line-ends within the body of the text. The situation appears different now. Turner’s
argument, quoted by Jordan 2000: 93 n. 8, that ‘The mark of the public letter-writer, making articulate
the fears and only half-expressed thoughts of his clients, is to be seen everywhere in the platitudes and
clichés of which so many private [papyrus] letters are composed’, if it characterizes reasonably well the
situation from the third century bc onwards, definitely does not apply to our corpus (a little more
standardization would help interpretation enormously!). Old and new finds of fragments of alphabet
sherds from the Black Sea region show that basic writing skills were relatively widespread: Dana 2009.
74
On writing (Greek, Iberian, Etruscan, Gallic) in this area of the ancient world from the sixth to
the second century bc, the contributions published in Revue archéologique de Narbonnaise 21 (1988),
remain fundamental; see especially the remarkable survey in Bats 1988; see also Eidinow and Taylor
2010: 37 and n. 31, 41 and n. 47.
 Writing and Letter Writing—The Evidence 43

in the first or in the third person. Œ Œº ıŒÆ might be part of a text all organized in
the first person, but it might also be there, as Slings has suggested, to round off an
order expressed previously in the third person.75 The only other clue is the verb
used at l. 8, Œ º ı ; this can be taken as an imperative (Œº ı , ‘tell’), but a verb in
the third person (Œ º  -, ‘he tells’) is just as likely—in which case, the body of the
letter itself would have been composed in the third person.76
Two other letters from the same area are too fragmentary to be worth discuss-
ing in detail (appendix 1, nos. 24 and 26); they probably also concerned trade,
although for the second one the hypothesis of a defixio has also been advanced.
Also from the area comes an early letter on clay (appendix 1, no. 25, dated to c.500
bc) that presents interesting, rather odd characteristics: in particular, a åÆEæ
(second person imperative) right at the beginning, followed by an address to
‘Energos brother’; then, information about what is and what is not needed
(probably for a business in pottery), in the first person singular of the sender
(cf. ŒŒÆıŒÆ at l. 5); the written document is referred to as an epistole (ll. 13–14).77
The first document that conforms to generic expectations is a lead letter of a
ship-owner to a captain (appendix 1, no. 29), found in the harbour of Marseilles
and dated to the third century bc. On the outside it has simply the name of the
addressee in the dative; the letter itself opens with ‘Megistes to Leukon, greetings’,
followed by a standard formula valetudinis ( N ªØÆ Ø, ŒÆºH  E· ªØÆ 
b Œ[Æ]d  E), and by the message, which is about _ moving the anchor of the ship.
_
The letter closes with the clause På Ø, ‘fare well’. This letter was certainly never
opened: the lead, when it was found, did not present any traces of unrolling. It is
interesting to find a letter on lead still in the third century bc—but then papyrus
may have always been more expensive in the West. An official letter on lead from
the Black Sea, of late Hellenistic or possibly even imperial period, shows at any
rate that lead continued to be used for letters.78
Let us now move from the margins towards the centre. Two lead letters from
the Chalkidike have so far been discovered, both of them fourth-century, only one
of which has been published (appendix 1, no. 33). The letter presents the normal
epistolary prescript (‘-tos to Tegeas, greetings’); it is composed in the first person
singular (the  E that appears twice may imply associates), while the recipient is

75
Slings 1994: 115: ‘As in the Berezan letter, the writer may be referring to himself in the 3rd person;
if so, with Œ º  - he introduces an order which he rounds off with Œ Œº ıŒÆ in l. 14. This would be an
argument against first person forms in the body of the texts’.
76
Œ º  had been initially proposed by Sanmartí and Santiago (1987); in two later contributions
(both published in 1988) Sanmartí and Santiago suggested reading Kæ]Æ ŒÆd Œº ı (second person
imperatives); Santiago 2003 now goes back to the third-person reading: ‘the writer gives instructions
referring to himself in the third person; the verb is characteristic for this type of document recording
instructions (keleumata) of an employer to his employees’. We may add that the  following the verbs
is much easier to understand if we assume a third person.
77
Prudence is of the order, as the letter is known only from a bad copy with a ‘most unusual
alphabet’ (Johnston, in Jeffery 1990: 464 A): a fake? See bibliography in appendix 1, no. 25.
78
See the letter published by Saprykin and Fedoseev 2008, appendix 1, C. Lead letters must have
remained in use: one of Parthenius’ narratives has the heroine, Polykrite, communicate with her brother
by inserting a lead letter ( ºıÅ KØ ºc) in a cake: Parthen. 9. 5 (the story is of course older: the
short notice, or ‘manchette’, preceding most of the tales composing Parthenius’ Love sufferings and
indicating the source of the story, mentions in this case Andriskos of Naxos—FGrH 500 F 1, probably
active c.300—and Theophrastus; a slightly different version is also in Plutarch, De mul. vir. 17, 254 d).
44 Greek Beginnings

addressed in the second person singular; the sender is trying to buy wood, and in
large quantities (no less than seven talents, and the purchase has to be completed
in seven days, otherwise the agreement will be cancelled); an [ ææ]ø?[ ], ‘Be
well’, closes the letter.
Finally Attica. Here the greater amount of data again allows a diachronic
sketch, similar to that possible for the letters from the Black Sea region and
from Spain. The earliest letters from Attica are graffiti on ceramic sherds (appen-
dix 1, nos. 35–8).79 They range in date from the mid-sixth century to the late fifth
century, and not all of them may indeed have been letters (‘notes’ seems a more
appropriate term); they differ widely in format, but they all seem to concern small
everyday communications, over short distances, between people who know each
other well. None of these documents presents anything even remotely close to a
prescript, nor do any of them contain a final greeting clause. It is difficult not to
agree with Harris’s rather dismissive evaluation of these texts.80 The one that is
closest to a letter is a graffito of the last quarter of the fifth century, on a black
figure skyphos (no. 38), containing the following five-line text:    |
K º | ˆºÆŒ Ø | K ¼ı |   (), ‘Sosineos sent to Glaukos in the city a
_ (the form of) a scroll to Glaukos in town’,
bundle’ (or, with Jordan, ‘sent a letter in
reading the final part as a dative, K   ^[Ø]; or also, with Immerwahr, ‘Sosineos
sends to Glaukos: (send) a bundle into the _ city’).81 Whatever the exact interpret-
ation, the use in this graffito of the verb Kغºø in the past is intriguing, and
points to epistolary language—but we are in the last quarter of the fifth century.
As for the letters on lead, they are later, appearing only at the end of the fifth
century, and they are different in form: we are not anymore presented with
hurried messages, but with real letters. One of them, from the Pnyx (no. 40),
seems to concern a commercial transaction; it opens with the invocation to the
gods (Ł ) which is typical of official documents, but it then has a prescript of
sorts, with åÆæ Ø ŒÆd [ , after which the restoration ªØÆ Ø suggested by Louis
Robert seems to impose itself. Thus, we may be witnessing a conflation of official
and private formats, similar to what happens, at the same moment, in some
Attic curses. 82 However, other restorations are possible that eliminate the initial

79
The first and oldest, no. 35, is in fact written with Megarian letter forms, and thus was probably
written by a Megarian; the content does not lend itself to the idea that this is a long-distance letter (the writer
asks another person to put a saw under the garden-door), so it is best to think of a Megarian in Athens.
80
Harris 1989: 89. Pébarthe 2006: 81–2 uses these texts to show that there was a certain volume of
correspondence, even in relatively low social circles. I think these texts show that literacy was diffused
at all levels, and in this sense they are important; but they do not speak for an important use of writing
for correspondence.
81
Resp. Immerwahr AVI no. 272; Jordan 1978 (SEG 28, 41). However, the comparison proposed by
Jordan for    used of ‘a scroll-like parcel’, with the defixio IG iii, App. (DTA) 45, 2–3: ¯hÆæ 
[Œ]ÆÆ- | H K   [HØ] - | ºı[]øØ ŒÆd ․˝․․, does not really work: the ‘lead desmos’ is only
metaphorically the scroll, and it can be so because of the context, a katadesmos, where the notion of
‘bondage’ (esp. after ŒÆÆø, to bind) is part of the horizon of expectations (see below, n. 92).
Oikonomides (SEG 36, 121) restores the last two lines as K   | [ æ Ø] and interprets it as a letter
sent to a man in jail; but apparently the sherd was broken before being_ inscribed, and there is not room
for much else.
82
Discussed below; see also Eidinow and Taylor 2010: 44–5 and nn. 62–3, and the problematic SEG
51, 984 (at the end of appendix 1), with invocation to ‘good fortune’ (but it is unlikely that this is a
letter).
 Writing and Letter Writing—The Evidence 45

invocation; the content of the message is impossible to retrieve; and the end of the
text is lost.
The three other documents from Athens are more interesting. One, the letter of
Mnesiergos (appendix 1, no. 39), dated to the end of the fifth or the beginning of
the fourth century bc, has been known to the scholarly world since the end of the
nineteenth century: it is one of the very first lead letters to have been found. On
the outside, the lead sheet details the address and persons to whom the letter
should be delivered. The letter itself, on the other side of the lead sheet, opens
with: Å æª  K غ  E YŒ Ø åÆæ  ŒÆd ªØÆ · (‘Mnesiergos sent those
at home greetings and be well’). We are here at the exact transition between a
narrative opening and a formal prescript: the somewhat intrusive K غ points
to a narrative rather than to a letter; remarkably, it is in the past, making it clear
that the letter is composed from the point of view of the addressee. This is one of
the earliest instances of the so-called epistolary aorist; another was on one of the
sherds from the Athenian agora (no. 38:    K º ˆºÆŒ Ø . . . ); com-
pare instead the letter of Achillodoros (no. 1), or the letter of Lesis (no. 41).83 The
letter continues in the third person for one more sentence, then it shifts very
slowly towards the first person (I ø, in the first person future, is the last
word of the letter). This letter was clearly written by someone else: the çÆ[Œ] of
l. 4 (ŒÆd ÆPe oø çÆ[Œ] [ å ], ‘he said that he too was doing well’), again in
the epistolary past, is very difficult to account for otherwise. The construction
shows that Mnesiergos explained what he wanted to say to the writer, who
reproduced it in the third person, save for the last sentence. It is, however,
interesting to notice that this sentence is extremely close to the later formula
valetudinis N ææøÆØ, s i å Ø· ŒIªg b ªØÆø, ‘if you are in good health, it
would be good; I myself am also in good health’. One may suspect that the writer
(the person who wrote the letter for Mnesiergos) had some notions of how to
write a letter; but because he was here reporting Mnesiergos’ words, he adapted
the formula to the occasion. Alternatively, if Mnesiergos was writing, the weight of
the beginning in the third person, and the inclusion of the wishes for health in the
prescript, may have rendered difficult a quick switch to the first person.
Another lead letter from Attica (appendix 1, no. 41), dated to the early fourth
century bc, and sent by a young boy, Lesis, presents a very similar picture. Here too,
there was probably an address on the outside, but only traces are now visible. The
letter itself is traced with beautiful, regular letters and almost impeccable spelling;
from the signs left by the instrument used to incise the lead, Jordan (2000: 93)
concludes that a reed stylus was used, of the type normally used to write on
papyrus. The text opens with the names of sender and addressees, without any
greetings (‘Lesis sends to Xenocles and his mother’, ¸BØ {Ø} Kغº Ø ˛  Œº E
ŒÆd BØ Åæd); until the third line, it is composed in the third person. Then there is

83
And cf. the Œ Œº ıŒÆ of the letter from Emporion no. 23 discussed above (n. 75). Thuc. 1. 129. 3
(the letter of Xerxes to Pausanias) is the first example of the use of an epistolary aorist in a literary letter:
the sending of Artabazus to carry the letter (something that happens after the letter is finished) is
described with an aorist, thus being seen from the point of view of the addressee. Koskenniemi 1956:
186–200 explains this usage as a consequence of the derivation of the letter from an oral message: a
messenger, after he had arrived, would have used past tenses to describe the requests of the sender. The
analysis of oral messages in Herodotus would not seem to support this hypothesis.
46 Greek Beginnings

a shift to the first person. This could be explained by assuming that someone else
wrote for Lesis, and that after the initial sentences, instead of summarizing what
he was being told, the writer transcribed the boy’s words as spoken; but the shift to
the first person may also have been favoured by the urgency of the situation. The
letter does not have a closing formula.84
The last document to be considered was published as a letter of the Athenian
banker Pasion, its date being fixed at before 370/369. Whether the Pasion who is
writing is the banker or not is immaterial here.85 But the way the letter is
formulated, as well as its content, makes it unique. In the text as presented by
the first editor, the sender, who speaks from the very beginning in the first person,
giving his own and his father’s name, Dikaiarchos, simply states that he is sending
someone called Satyrion to punish and prosecute other persons who have
wronged him: ‘(I), Pasion son of Dikaiarchos, am sending Satyrion to punish
and to prosecute both Nikostratos, Deinon’s brother, and Arethousios, because
they are wronging me . . . ’.86 One cannot but agree with Sosin (2008) that this
reading makes for an unlikely letter: there is no message in it, no addressee, and
while the second ήd just before the name of Nikostratos can be explained as
building up a parallelism with the ήd preceding Arethousios, it still makes the text
difficult. Moreover, in a letter one does not expect the name of the writer’s father.
Sosin is surely right that in a message we need at the opening the names of a
sender and an addressee; hence his proposal to see in Pasion the sender and in
Dikaiarchos the addressee, who is asked to punish Satyrion and the others.87 Even
with this interpretation, some oddities of language remain, such as the very strong
initial first person (name + Kغºø) that finds no parallel in our corpus.88
These oddities might, however, be explained by the assumption that the letter was
written personally by the sender, someone possessing enough letters to be able to
write a short letter himself, but not particularly interested (or versed) in epistolary
conventions. The comparison with the letter sent by the slave Lesis from the
foundry is instructive: there, no words were split over end-lines; here, word-
division is endemic.
What can we conclude from this survey? Clearly, already in the sixth century,
Greek traders from the Black Sea, but also landowners, did use writing for long-
(and not so long) distance communication. The letters from Spain and the

84
On the context of this letter, see, besides Jordan 2000, Harris 2004 (who argues from the use of
‘despotes’ made by the boy to denote his master that Lesis was a slave), and Harvey 2007. The
possibility should be taken into account that Lesis is purposefully describing his situation (dire enough)
as that of a slave. L. Holford-Strevens points out to me that the mother here might be Xenokles’ (too?),
a possibility that has not been given due attention in the interpretation of the document.
85
Banker: Jordan 2003a. Doubts: Millett 1991: 267 n. 7; Gauthier, in B¯ 2004: 140; Sosin 2008: 106 n. 4.
86
So Jordan 2003a. Text and translation in appendix 1, no. 42.
87
Sosin 2008, an interpretation endorsed by Eidinow and Taylor 2010: 34. This requires one to
assume a second error in the initial line, besides the missing ˜ in Dikaiarchos’ name, namely an
instead of øØ for the dative: as pointed out per litteras by L. Holford-Strevens, it seems rather early for
reduction of øØ to ø, and ø is found elsewhere in the document.
88
It is paralleled by the very strong language used in the letter, a language not typical of legal
proceeding, but pointing rather to extra-judicial pressure: cf. Sosin 2008: 108. This might be a curse:
‘I, Pasion son of Dikaiarchos, am asking that Satyrion be prosecuted and punished’. The authors of a
curse do not usually name themselves (see below, 48 and n. 93, 57 and n. 130); however, in the case of
judicial prayer, the principal may state his or her name: Versnel 2009: 279.
 Writing and Letter Writing—The Evidence 47

Chalkidike fit this pattern too. Some of the ‘letters’ are ‘crisis-messages’, as

Thomas puts it, messages sent in rather special circumstances (seizure of goods
and/or of a person, as with Achillodoros and Apatorios);89 but quite a number of
them (especially those on ceramic sherds) are rather trivial communications.
Attica is different: compared with the very high number of extant graffiti and
curses of the sixth to fourth century, letters appear to be relatively rare, somewhat
late, and fairly idiosyncratic. Writing was thus used in Attica for private purposes,
but evidently not so much for letters.
In all cases it is very difficult to know whether the writer was the same as the
sender; while there are a few instances where the language used in the letter makes
it fairly certain that the letter was written by a person different from the sender,
overall there is certainly here scope for regional variations, and even for variations
from case to case. Moreover, in quite a few instances it is evident that the written
message alone would not have been enough; oral instructions would have been
required for the exchange to work. In those examples, one may suspect that the
written message functioned more as a receipt than as a letter.
As a result, it seems evident that for the fifth century and even for part of the
fourth it is impossible to speak of an epistolary style codified and perceived as
such.90 The comparison with other types of texts that use a very similar enuncia-
tive stance will confirm this.

2.2.2. Curses, Letter-curses, Letters to Gods

The material on which the earliest letters are written raises the question of the
possible connection with curses, which were also mainly written on lead. Of
course, in the case of the letters, the impression that lead was the main material
may be misleading, since documents written on lead (and ceramic) survived, but
any written on perishable materials will have been lost. Conversely, one might
point out that although the great majority of curses are written on lead, there are
some instances in which wooden tablets or other materials are used.91 But aside
from the writing surface, there is also a certain similarity in the phraseology, that
is, in the way the message in the two types of document is formulated; moreover,
some curses refer to themselves as a ‘letter’.
The earliest instances of the practice of inscribing curses on lead are later than
the earliest letters: they date to the fifth century, and come from Olbia on the Black
Sea, Sicily, and Attica; but the practice spread quickly and widely.92 The aim of

89
Thomas 2009; see also Eidinow and Taylor 2010.
90
For Sykutris (1931: 189) this applies to the fourth century too, and not just to documentary letters
but also to literary ones, such as those of Isocrates.
91
See Graf 1997: 132–4; Curbera 1999: 161; also Eidinow 2007: 201 for curses; Eidinow and Taylor
2010: 31 and n. 4 for both letters and curses. Appendix 1, no. 7 is an instance of a letter on ceramic
reused for a curse; see also the comments of Chaniotis in EBGR 2001, no. 26.
92
Faraone 1991: 3. Faraone suggests (1991: 5) that the defixio was originally a purely verbal/gestural
curse. The earliest terms for curses do not allude to writing: På (‘prayer’) is attested in Sicily,
ŒÆ   (‘binding spell’) is also frequently found, and of itself does not imply a material support;
Faraone compares the sung o    Ø  (‘binding song’) of the Erynies in Aeschylus’ Eumenides
( Ø  is used at 306 and 322). From the fourth century onwards, ŒÆƪæçø appears as the term
48 Greek Beginnings

curses is to immobilize, to magically ‘bind’, the other (this is the case of judicial,
competitive, or erotic ŒÆ  Ø, literally ‘binding curses’), or also to obtain
redress or justice, for example for a theft or for abuse, in which case the curse took
the form of a prayer.93 The most basic scheme for the binding ŒÆ  Ø is that
of giving lists of names that follow each other sequentially without line breaks. But
some early curse tablets are modelled on formats typical of public communi-
cation; more precisely, some early fourth-century ŒÆ  Ø from Attica (and
some earlier ones from Sicily) imitate public inscriptions, listing names with
patronymics in vertical parallel columns, and, in two cases, even adding the typical
heading of decrees, Ł · IªÆŁB fi åfi Å (‘Gods; with good fortune’).94 In other
instances, the curse tablet adopted the model of a private communication, making
use of the epistolary format, something that must have appeared almost natural,
since letters were also often written on lead.95 In two instances, again from Attica,
the term KØ º is used; in other curses, it is the presence of an address on the
outside of the lead sheet, or the use of verbs of sending, that orients one towards
an epistolary interpretation. Finally, two ŒÆ  Ø probably from Arcadia (or
Megara?) and one from Athens imply that the dead might read the ªæ ÆÆ,
meaning the tablet itself, which is clearly seen as a letter, with its potential for
communication. On the basis of this type of document, a category of ‘Unterwelts-
briefe’, mainly addressed to Persephone and Hermes, was tentatively defined by
Preisendanz, who, however, even while proposing the term (in which he was
following the lead of Wilhelm), pointed out at the scarcity of attestations.96 Let us
see some examples.

with which one ‘inscribes’ or ‘registers’ the victim of the curse in the presence of one of the infernal
gods. Overview in Graf 1997: 118–74; Eidinow 2007: 139–55. Superb contextual presentation of Sicilian
defixiones in Curbera 1999.
93
See Versnel 1991; Versnel 2009: 278–82 and 326–7, where he suggests that while in the defixiones
(or katadesmoi) the defigens does not give his name but the target is named, in the prayers the principal
names himself but the target is unknown (‘a thief ’); a more nuanced view in Eidinow 2007: 143–4 and
n. 16.
94
Faraone 1991: 4, who gives as examples of the first practice DTA 55 and SGD 48 (= NM 14470,
re-edited by Jordan and Curbera 2008, who date it now on prosopographical grounds to 345–335 bc),
while the heading is found in SGD 18 and DTA 158. On the difference between horizontal and vertical
lists, see Gordon 1999; on lists in Sicilian defixiones, Curbera 1999: 166–7; on Attic defixiones, Parker
2005: 121–33.
95
Eidinow 2007: 145–6 relates this to the language of business transactions. Gager 1992: 201
remarks that epistolary communication with spirits and deities is not uncommon in spells, and refers
to Egyptian parallels; at 248 n. 10, in the course of a comparison with the letter carried by Beller-
ophontes (Hom. Il. 6. 168–70), he stresses that the earliest Greek defixiones take the form of letters to
the underworld. While the parallels (and esp. the rapprochement with Bellerophontes) are interesting,
I would be sceptical of seeing here any influence from the East; see Johnston 1999: 90–5.
96
Faraone 1991: 4; Preisendanz 1972: ‘Wenn auch alle schriftlich geäußerten katadesmoi, eins-
chließlich der nur aus Namen bestehenden, die Aufgabe von Unterweltsbriefe zu erfüllen hatten und so
mit Wilhelm tachydromeia heißen konnen ( . . . ), haben sich doch bisher nur diese beiden epistolai
gefunden’ (‘even if all inscribed curses, including those consisting simply of a name, were meant as
letters to the Underworld and so could be called, with Wilhelm, tachydromeia, . . . to date only two have
appeared that are explicitly defined epistolai’: he meant DTA 102 and 103). ‘Epistole’ appears also in
SGD 109, discussed below. For ªæ ÆÆ to the dead see Audollent DT 43 and 44, discussed below,
n. 108. Cf. also Graf 1997: 130–1; Versnel 2002: 60–1.
 Writing and Letter Writing—The Evidence 49

The first curse to present itself as a letter is a text from Kotana in Attica, dated
to the fourth century bc:
KØ {}ºc |  ø | []Æ <Ø> | ŒÆd !æ  ç<fi Å> {} | 5 Œ Æ | "ØØÆ |
c # ØæÅ | c K < > IØŒ (F)Æ | ŁıªÆ<æÆ> | 10 ¼æÆ | ŒÆd æÆ <>ÆØÆ |
KŒ Å |  Ł º Æ ŒÆd £ ¼ææ ·| —ƪŒæÅ Æ<Æ> | 15 ˜ØçÆ  |  ƪÅ.
(IG iii, App. (DTA) 102)
I am sending a letter to the daimones and Persephone, and delivering (to them)
Tibitis, (daughter of) Choirine, who did me wrong, her daughter, husband, and three
children, two female and one male, Pankrates, Mantias, Diophantos, and Metagenes.
(Trans. Gager 1992)97
The explicit mention of an offence puts this curse in the category of ‘judicial
prayers’: in texts of this type, the sender feels that he has received an injustice, and
sends a letter to the gods of the Underworld (here Persephone and the daimones),
asking for rectification.
The other example of a curse in which the term KØ º appears also belongs
to the group of judicial prayers, as the invocation to Dike shows; it comes from
Attica as well:
a.  ¯æ [B
fi ] ŒÆd ! æ ç[][fi Å]   KØ [º]c I - |  [ø· ]  ÆFÆ N
IŁæ (ı) IÆç[Æ] [] | ÆP (). ˜ŒÅ, ıå E º (ı) ŒÅ. | ˚ƺºØŒæÅ :
ÆØŒæ ı : ¯PÆ[Œ]  | 5  Oºı Øøæ  [ . . . ¨] çØ[º  . . . ]øæ  _ |
Z[]ıæ  : —Æø[] #ÆæE  ˚ƺºØŒ  ˚Ø Æ | [ ºº]øæ  [ . . . Ł] 
!غ ŒºB _[˜Å] çغ  | ŒÆd ØŒ Ø ŒÆd Y Ø ¼ºº  | [ - - - - - - ]  | 10 ÆææØ . ? :
˜Å Œæ[Å] e  æd B ŒÅ[] | ØŒÆÇ[]   : Å Æå  çغ _ [].
_
b. ¸FØ ˜øæ Ł ı æåE  #ÆæE  |   Œº ı . . .
(Wilhelm 1904b = IG iii, App. (DTA) 103)98
a. To Hermes and Persephone I am sending this letter; that they may never bring these
up among men. O Justice, may I finally receive justice. Kallikrates son of Anaxikrates;
Eudidaktos; Olympiodoros [ . . . ] Theophil[os . . . ] -odoros, Zopyros, Pasion, Char-
inos, Kallinikos, Kineas, Apollodoros, . . . -theos, Philokles, Demophilos and their
advocates and if there is someone else of [their friends?], Demokrates who judges
on the trial, Mnesimachos, Antiphilos.
b. Lysis son of Dorotheos, Archinos, Charinos son of Menekles . . .
Interestingly, both texts lack the name of the sender, while the victims of the
curse are well-defined: in this respect, then, these two texts do not correspond to
Versnel’s definition of a typical prayer for justice. Moreover, the first document
(DTA 102) is not so much a letter as a narrative of sending a letter; the gods

97
This is the side A; side B concerns other people and is different in character (unlike in A, there is
no mention of an IØŒÆ, ‘injustice’, and the text begins directly with ŒÆå ı, ‘restrained’, followed by
the names of two boxers; forms of ŒÆ å Ø are used also for the names of the three women that follow,
to be restrained by Persephone, Hermes, and Hades). Short discussion in Gager 1992: 101–2; Eidinow
2007: 146, 229, and 378–9. See Wilhelm 1904b: 112–13 for the hypothesis that this tablet might afford
an opening into the life of courtesans.
98
Wilhelm’s text (1904b: 122–5), to which I refer for possible restorations and detailed commentary; his
interpretation is endorsed by Parker 2005: 128–30 (with discussion of the political background). Wilhelm’s
readings connect this tablet to judicial defixiones; the process was linked to naval affairs, and dated to the
years 325–322 bc. Gager 1992 maintains the text of Wünsch, as does Eidinow 2007: 230 and 380.
50 Greek Beginnings

themselves, the daimones and Persephone, are not directly addressed. To put it
differently: instead of a prescript in the third person, we have an epistolary
narrative in the first person present ( ø), which allows the author of the
curse not to name himself, and which gives the curse its performative value; this is
asyndetically coordinated to the statement, also in the first person but aorist
(Œ Æ), that the author has conveyed (presumably to these same underworld
gods) the victims of the curse.99 Thus, the epistolary narrative reflects two
moments: the actual utterance of the curse, and the continuity of the ritual action
implied in the sending of a letter. The second document (DTA 103) elaborates
along the same lines: the author opens by announcing that he is sending a letter to
Hermes and Persephone; this is followed by an expression of his wishes, both in
terms of the permanency of the curse (ÆFÆ must refer to the written tablet), and
of what he personally hopes to obtain; then the list of names is supplied.
There is thus an epistolary subtext to these curses, but their formulae do not
reflect those of any known ancient letter. In three further curses, two from Attica,
one from Euboea, of the same period as DTA 102 and 103, the names of the
Underworld gods are written on the external part of the lead sheet, like the address
in a letter; but the internal part, that is, the curse itself, does not present any
epistolary characteristic. The most detailed text, a ή   from Attica dated
to the fourth century bc, has on the outside the names of Hermes and Hekate in
the nominative ( ¯æ B åŁØ  ŒÆd  ¯ŒÅ åŁ Æ); inside, the curse:
[!] æ[ØŒ ] æe e  ¯æ B e åŁØ  ŒÆd [c  ¯]-
ŒÅ åŁ Æ ŒÆÆ Łø· ˆÆº Å, lØ ! æ []-
ŒøØ, ŒÆÆø æe  ¯æ B åŁ ØŒe ŒÆd  ¯ŒÅ åŁ Æ ŒÆÆ[]-
ø· ŒÆd ‰ y  › ºı  ¼Ø  ŒÆd łıåæ, oø KŒ <E>  ŒÆd a KŒ <>ø ¼Ø Æ
[Œ]-
5 Æd łıåæa ø ŒÆd  E ’ KŒ <> (ı) L  æd K (F) ºª Ø  ŒÆd  (ı)º ı Æ .
(...)
Let Pherenikos be bound before Hermes of the underworld and Hekate of the
underworld. I bind Pherenikos’s (girl) Galene to Hermes of the underworld and to
Hekate of the underworld I bind (her). And just as this lead is worthless and cold, so
let that man and his property be worthless and cold, and those who are with him who
have spoken and counseled concerning me. ( . . . )100
As for the other two texts, they are addressed to the Praxidikai and Hermes
(æe a | —æÆØŒÆ, | æe  ¯æ[ B), and to Hekate and Persephone ([ ¯]ŒÅ
ŒÆÆåŁ [Æ]) | ŒÆd — æØçÅ); their internal side is barely legible, but one seems
to consist simply of a list of names, while in the other the verbs ŒÆƪæçø and

99
Eidinow’s translation (2007: 378) seems to imply that the letter conveys Tibitis and her friends
(‘I am sending a letter to the gods below and Persephone, conveying Tibitis’); cf. instead Gager 1992:
201–2: ‘I am sending a letter, and deliver (to them) Tribitis.’ See Graf 1997: 130–1 for the relationship
between the oral recitation of the spell and its inscription (contemporaneous: but also in the letter
above not much time needed to have passed); and Carastro 2006: 164–6 for remarks on the role of
writing (seen as more and different than simply a transcription of the oral curse: ‘écrire n’est pas
transcrire’).
100
DTA 107 = Gager 1992, no. 40; I give here only the first half of the curse, in Gager’s translation;
the second half proceeds along very similar lines, but the indication that any other supporters of
Pherenikos must be bound too (ŒÆd Y [Ø] ¼ºº  ! æ ŒøØ ØŒ[  . . . ŒÆÆ Łø) permits one to
deduce a legal context. Interestingly, Galene appeared also in DTA 102 B: cf. Wilhelm 1904b: 112.
 Writing and Letter Writing—The Evidence 51

I Œ æø (I ŒÅæø, ‘I banish’) can be recognized.101 Clearly, these three

are not judicial prayers, but rather binding curses; they are also rather dissimilar
within themselves, with the first using a ‘wish-style’ formula and analogy to
expand on the binding formula, while the second goes for the simple list of
names, and the third for the bare ‘binding’ formula.102
Finally, three documents, two from Attica and one from Sicily, all probably later
than those we have seen (second–first century bc?), present the motif of ‘sending a
gift’ to the gods of the underworld, a gesture that is in one instance explicitly
associated with a letter. So begins one of the Attic documents:
˜Æ Ø åŁ øØ ŒÆd BØ åŁ - | ÆØ ŒÆd  E åŁ  Ø AØ |  ø Hæ  . . . |
ŒÆÆåŁØ[Æ] ÆPe `․․․ | 5 ․"$% IªÆŁ, L ÆP F . . . (DTA 99)
To the chthonian daimon and the chthonian goddess and all the chthonian divinities
I send a gift . . . 103
The text comprises approximately fifteen lines, but owing to its bad state, the
nature of the gift is unclear. Another curse from Attica, now lost, but probably
from the same grave as the previous one (and thus from the same milieu) also has
the formula  ø Hæ . A long and complex text from Lilybaeum in Sicily,
dated to the second century bc, and written from right to left, is more specific: it
opens with a straight address to Hermes ‘the Binder’ and the Telchines ( Æ
 ı, Œ- | ø  ¯æ B Œ å , 1–2); it then ‘sends’ the gift, the handsome girl Prima
(Hæ  e  ø | ÆØŒÅ, 5–6), who is described in detail; it then states that
‘Allia Prima, of her the letter and . . . I write’ (ººÆ —æE Æ Æ- | Å c
K Ø[ ºc] | ŒÆd e . . . ªæçø, 22–3).104 Again, there is no mention of injustice:
it seems clear that these are ‘binding’ ŒÆ  Ø; but it is interesting to find the
term ‘epistole’ in connection with the sending of a gift.105
That the ‘writtenness’ of the curses is important, and that specific attention is
dedicated to it, is clear from the care given to the way these documents are (mis)
inscribed (e.g. retrograde, with some names written backwards, from bottom to
top, with drawings, with nonsense letters . . . ); the fourth century ή  
from Athens DTA 67, for instance, which is not addressed to any specific

101
The texts are Ziebarth, Neue Verfluchungstafeln 20 (from Euboea, c. third century bc) and 26
(from Attica, undated). Cf. Faraone 1991: 4; Versnel 2002: 48–9 and 60–3.
102
On the three main ‘styles’ of curses and their formulae see Faraone 1991; Eidinow 2007: 144–52,
and 150 for DTA 107.
103
DTA 99; cf. also the very similar, and now lost, SGD I 54 (Austin, Proceedings of the British
School at Athens 27 (1925–6), 73–4) = López Jimeno 1999, no. 36, probably first century bc, and
probably from the same grave as DTA 99; on both documents, Versnel 2009: 345 n. 90.
104
SGD 109 = Brugnone 1984: 158–62 no. 184 (= SEG 34, 952) = Eidinow 2007: 430–2 (with full
text and translation; Eidinow, however, relies on the text of Gabrici, which has been superseded by that
of Brugnone). On the motif, see Versnel 2009: 345; López Jimeno 1990 (and for this text, 139–40).
105
As Versnel 2009: 345 says, ‘this tallies nicely with expressions containing forms of å ÆØ
(  ÆØ/ Ø Ł ), asking the gods of the Underword to receive some persons’, mentioned in
some lead tablets from Morgantina (SGD 118, 119, and 121, cf. Eidinow 2007: 432–3). A gift is mentioned
also in a tablet from Olbia Pontica, dated between the third and the first century bc, and published in
Bravo 1987 (SEG 37, 673; translation in Gager 1992, no. 48). The author of the curse addresses the dead
(the curse functions through analogy), and promises a gift if he will restrain his opponents. The situation
is thus different, and there are no epistolary markers; but the text present an interesting shift from ‘we’ to
‘I/me’, which may point to the ritual context of the casting of the curse (Bravo 1987; Gager 1992: 138).
52 Greek Beginnings

divinities but relies simply on analogy, calls attention to the fact that the text is
written backwards: the curse states that ‘Just as these (words) are cold and
backwards (u æ ÆFÆ łıåæa ŒÆd KÆæ æÆ), in the same way may the
words of Krates be cold and backwards ( oø a ˚æÅ  a Þ ÆÆ łıåæa
[ŒÆd] | [KÆæ] æÆ ª[ Ø] ), and those of the informers with them and the
judges . . . ’.106
Just as importantly, a few texts stress that they are (or might potentially be) ‘for
reading’. Thus, DT 52, a late-fourth-century curse from Athens, begins by listing
the names of the offenders, and then continues: ‘I bind Kerkis, both the words and
the deeds of Kerkis and the tongue, before those youth who died unmarried, and
whenever they read these words (ŒÆd ›Æ- |  y Ø ÆFÆ IƪHØ), then will
be the time for Kerkis to speak ( ˚æŒØØ ŒÆd e _ çŁ ÆŁÆØ). I bind Theon,
him and his girls and his craft and his resources and his _work and his words and
_ _ _
deeds. O Hermes who binds, keep these things and read (these words) for as long
as they are living ( ¯æ B åŁØ []ÆFÆ | 15 f Œ] å ŒÆd IƪHŁØ ÆFÆ ø
i s Ø ÇHØ)’.107 Fascinatingly, the reading of the _tablet by the dead, a reading
which will never happen, but which would release the victim from the curse, is
contrasted to the request to Hermes to ‘keep these’ and ‘read these’ for as long as
the intended victims of the curse are alive: Hermes’ repeated reading will maintain
the curse activated.108 Conversely, reading by a mortal, and more precisely by the
person who wrote the curse, is in a few instances specified as the condition for
deactivating the curse: thus, an early curse/prayer from Pella stipulates that ‘I
entrust (this curse) to Makron and the daimones. And were I ever to unfold and
read these words again after digging (the tablet) up, only then should Dionyso-
phon marry, but not before.’109
What can we conclude from this survey? First, references to the epistolary
format are found in both judicial prayers and ŒÆ  Ø; letter writing can thus
function as a paradigm for both. As Versnel puts it, ‘Continuous reading by the
divine addressee, then, is at least one of the objectives of writing a letter to the god
and depositing it at a place where the god is thought to be present, be it his
sanctuary [ . . . ] or the nether world via a grave, a pit or a well.’110 Second, the
reference to letters, that is to a type of document in which the sender gives at the
very opening, in the prescript, his own name and that of the addressee, is not
sufficient to overcome the curse writer’s reluctance to name herself or himself, not

106
Cf. Eidinow 2007: 167, with other instances of analogy, and 363–4 for full text and translation;
Curbera 1999: 162–3.
107
Jordan 1999: 119–20, with an excellent commentary; Eidinow 2007: 393–4 (but her translation is
unreliable).
108
Versnel 2002: 61–2. The possibility that the dead may read the message is evoked, only to
highlight its impossibility, in two very similar texts from the same grave, DT 43 and DT 44 (c.300 bc,
possibly from Megara, or from Arcadia): DT 43 states that ‘Whenever you, o Pasianax [probably the
deceased] read these letters—But neither will you, o Pasianax, ever read these letters, nor will ever
Neophanes son of Agasiboulos, direct a lawsuit. But just as you, o Pasianax, lie inert here, so may also
Neophanes be inert and become nothing.’ Cf. Bravo 1987 and Voutiras 1999; overall discussion in
Versnel 2002: 60–2; Nisoli 2007: 39–41.
109
Voutiras 1998 (SEG 43, 434), ll. 3–4; cf. Versnel 2009: 317–19. At l. 7, the request to ‘keep [this
piece of writing?] for me’, shows how important it is that a readable text continues to exist.
110
Versnel 2009: 62.
 Writing and Letter Writing—The Evidence 53

even in the two instances of ‘letter to the underworld’ which were clearly judicial
prayers, where the name of the author is, at least in Versnel’s definition, expected.
This implies that even if the distinction between ‘straight’ ŒÆ  Ø and judicial
prayers is worth maintaining (and not only because of its undeniable heuristic
value, but because it does correspond to differences within the corpus of ancient
curses), there is indeed a ‘borderland’, a fairly consistent grey area between the
two.111 Third, the curses that present explicit reference to the letter format are
comparatively rare: interestingly, they mostly date from the fourth century bc,
that is from a moment when both curses and letter writing were still looking for
their appropriate, written, format.112
Another group of documents found in graves presents texts that have been
compared with letters: the gold lamellae commonly labelled as ‘Orphic’ or ‘Bacchic-
Orphic’.113 Gold lamellae deposited in tombs, inscribed or uninscribed, have been
found all over the Greek world: in southern Italy, northern Greece, and Crete, but
also in the Peloponnese, in Sicily and in Rome, on Lesbos and possibly in Asia
Minor. Most are from the fourth century bc, although one from Hipponion
(southern Italy) may be as early as the fifth century bc, while the one lamella
from Rome may be as late as the second or third century ad.114 These lamellae
have been organized variously, on the basis of content, and/or of provenance.
Some of the long hexametrical texts consist of instructions and performative
injunctions addressed to the deceased, which frame a first-person declaration of
identity by the deceased himself, in the central part of the text; others are mainly
in the first person, consisting in a series of statements made by the soul to an
interlocutor, the underworld goddess, who answers back in a performative
speech-act to state that the defunct ‘shall be god’; other texts are much shorter,
and simply give the name of the deceased, or add the qualification Å.115 Five
among the shorter lamellae published to date appear to take up an epistolary
scheme, with the name of the addressee(s) in the dative followed in three instances
by åÆæ Ø; moreover, in three instances the text also gives the name of the sender.
Even so, it is unclear that they should be considered letters. Here are the texts:
1. Bernabé and Jiménez San Cristóbal 2008 L 16b = Graf and Johnston 2007
no. 31, gold lamella from a cist grave close to a temple from Pella/Dion, end

111
Emphasis on this borderland in Jordan 1999: 115. For this reason, I hesitate to follow the further
inference of Versnel 2009: 62 that ‘in the field of the conventional defixio the motive of fixing the
spoken curse formula is more dominant than in the prayer for justice, which naturally and demon-
strably is generally [?] phrased as a letter or at least as a direct form of communication with the god’.
112
The difference between the Greek curses and the Egyptian ‘letters to the dead’ is most clearly set
out in Johnston 1999: 90–5.
113
On the impropriety of such denominations, and on the kind of mystery cults (Dionysiac/
Eleusinian?) one might imagine behind them, see Calame 2008; Calame 2009 [2006]: 210–28; Ferrari
2007. See Curbera 1999: 161–2 for the possible connection between the lead sheets of the defixiones and
the golden lamellae (both addressed to underworld gods; both on thin sheets of metal, shaped as
leaves).
114
See the map and convenient presentation in Graf and Johnston 2007: 2–49. It is important to
realize that these are fluid texts, linked to specific ritual contexts: Ferrari 2010 speaks of a ‘palaeotype’.
115
Detailed analysis of the pragmatic and enunciative aspects in Calame 2009 [2006]: 177–210; see
also Ferrari 2010.
54 Greek Beginnings

of fourth century bc: ! æ çÅØ | —  Ø  Å | P  , ‘To

Persephone Poseidippus, pious initiate’.
2. Bernabé and Jiménez San Cristóbal 2008 L 16k = Graf and Johnston 2007
no. 37, gold lamella from a Hellenistic grave in Vergina: !غÅ ! æ çÅØ
åÆæ Ø, ‘Philiste to Persephone, rejoice’.
3. Bernabé and Jiménez San Cristóbal 2008 L15a, gold lamella from Heraclea
in Macedonia, date uncertain: !غø æÆ HØ ˜  Ø åæ (Ø), ‘Philotera to
the Lord, rejoice’.
4. Bernabé and Jiménez San Cristóbal 2008 L 15 = Graf and Johnston 2007 no.
15, gold lamella from Eleutherna in Crete, folded once, 2nd/1st cent. bc:
[—º ]øØ ŒÆd !- | [ æ]  Ø åÆæ Ø, ‘To Pluto and Persephone, rejoice’.
5. Bernabé and Jiménez San Cristóbal 2008 L 14 = Graf and Johnston 2007
no. 17, gold lamella from Rethymnon, dated to between 25 bc and 40 ad,
unfolded: —º øØ | ! æ çÅ[Ø], ‘To Pluto. To Persephone’.116
It has been argued that these five documents should be considered part of a
larger ensemble of sixteen ‘short’ lamellae, bearing in most cases only the name of
the departed.117 How to interpret the function of these ‘short’ lamellae is difficult.
As Johnston has suggested, they might be proxies, intended to speak on behalf of
the soul, presumably unable to speak—in which case there would be an ‘epistolary’
side, since the lamella would function as a message that the dead brings along to
hand over to the infernal gods.118 A detailed case for an epistolary (and non-
dedicatory) interpretation of the short lamellae has been advanced by Dickie, on
the basis of a comparison between the short lamella, no. 1 above, and the opening
of two quasi-identical lamellae, symmetrically disposed over the chest of a woman,
in a fourth-century tomb at Pelinna, in Thessaly.119 The Pelinna lamellae open
with two hexameters:
F ŁÆ  ŒÆd F Kª ı, æØºØ , ¼ ÆØ HØ
N E ! æ çÆØ  ‹Ø BŒåØ  ÆPe ºı .
Now you have died and now you have come into being,
O thrice happy one, on this same day.
Tell Persephone that the Bacchic One himself released you.120

Based on the presence in the second line of the Pelinna tablets of a verbum dicendi
(the injunction expressed with the infinitive N E, ‘tell’) followed by the dative ‘to

116
Unless otherwise indicated, texts from Graf, in Graf and Johnston 2007; translations Johnston,
with modifications.
117
Johnston, in Graf and Johnston 2007: 94–5 (cf. 134–6 and 183), distinguishes between longer
texts (primarily mnemonic devices, giving instructions to the deceased), and proxies; the latter ‘may
also be understood broadly as mnemonic devices’, since the reason why the soul itself is unable to speak
(and thus needs a proxy) may be postmortem confusion.
118
Johnston, in Graf and Johnston 2007: 134. Note however the doubts of Calame 2008: 303–4 on the
legitimacy of considering these short texts and the longer lamellae as expressions of the same religiosity.
119
Dickie 1995 (cf. Graf and Johnston 2007 no. 26 a and b; Bernabé and Jiménez San Cristóbal 2008
L 7a and b).
120
Text and translation in Graf and Johnston 2007: 36–7, with commentary at 131–3; detailed
analysis of the enunciative shifts in Calame 2009 [2006]: 208–10; correlation of these shifts to changes
in rhythm (hexameters for the first voice, the leader of the funeral rites; lyric verses for the second voice,
Persephone) in Ferrari 2010. See also Bernabé and Jiménez San Cristóbal 2008: 61–94.
 Writing and Letter Writing—The Evidence 55

Persephone’, Dickie argues that also in the short lamellae the datives are not
datives of dedication, but should be interpreted as ‘To Persephone, (tell . . . )’.
Thus, while in the longer texts which are similar to the Pelinna tablets the
deceased is told to tell (and he is also told what to say), in the shorter tablets the
deceased is presented as addressing the divinity—through the tablet, of course,
which results in something close to the prescript of a letter. However, there is no
‘body’ of the letter, no message, in the short lamellae. Dickie proposes a further
comparison with two ‘long’ poetic lamellae from Thurii that present the deceased
as speaking in the first person and stating the request he is going to make: ‘Now
I come, come as a suppliant to Persephone, so that she may kindly send me to the
seats of the pure’.121 A recently published lamella from Pherae offers a splendid
example of such a request, expressed in the first person and evidently intended for
an underworld god, probably Persephone: ‘Send me to the bands of initiates; I am
able to perform the [holy] rites of Demeter Chthonia and Mountain Mother’.122
In the case of the short lamellae, we would have to assume that the tablet just
provides the opening address (‘To Persephone’ or ‘To Persephone, greetings’),
while the request itself is left unformulated.
And yet, considering the importance of the declaration of identity in most of
the longer lamellae, it might be simpler to interpret the short texts that only
contain the name of the deceased along similar lines, as declarations of identity.
It is true that the identity asserted in the longer texts is the one shared by the
initiated, expressed in the formula ‘I am a son of Earth and starry Sky’; but in a
recently published lamella from Amphipolis, dated to the fourth/third century
bc, the deceased, while asserting her ritual purity, also states her (earthly)
identity: ‘Pure and sacred to Dionysos Bacchios I am; Archeboule daughter of
Antidoros’.123 The short texts with the name of the deceased and dative of an
underworld god could be declarations of identity, addressed (or dedicated) to
the god. Viewed in this light, the only problematic short texts are those
presenting an addressee in the dative, followed by åÆæ Ø. But one may wonder
whether åÆæ Ø is really to be understood here as a greeting; rather it may be
an invitation to the gods to rejoice over the gift of a golden lamella and of the
person now dead, according to a more fundamental meaning of åÆæ Ø (invi-
tations to rejoice, expressed through the repeated use of the verb åÆæ Ø, are
actually present in one of the gold lamellae from Thurii).124 In the pragmatic
situation in which these texts function, the latter interpretation may be more
appropriate; however, the fact remains that the enunciative stance of these texts

121
Graf and Johnston 2007, no. 6 and 7, ll. 6–7 = Bernabé and Jiménez San Cristóbal 2008 L 10 a
and b, 6–7; cf. Dickie 1995: 82.
122
I follow here the text and translation proposed by Ferrari and Prauscello 2007; cf. Ferrari 2010:
212–13. Graf and Johnston 2007 no. 28 and Bernabé and Jiménez Cristóbal 2008 L 13a restore a
reference to Dionysos at the end of l. 1 (cf. SEG 55, 612).
123
Graf and Johnston 2007 no. 30 = Bernabé and Jiménez San Cristóbal 2008 L 16n. Note also the
(late) lamella from Rome, Graf and Johnston 2007, no. 9 = Bernabé and Jiménez San Cristóbal 2008
L 11, where the dead woman is addressed by her name. On the collective identity of the mystai, Calame
2009 [2006]: 224–8.
124
Graf and Johnston 2007, no. 3 = Bernabé and Jiménez San Cristóbal 2008 L 8. For the exact
meaning and function of åÆEæ here (once, initially, ‘welcome’; and twice, later, ‘rejoice’) see Ferrari
2010: 213–14.
56 Greek Beginnings

is that of a third-person address, much closer to that of letters than to that of

prayers.125

2 . 3 . C ON C L U S I O N

Although the first written document mentioned in Greek literature is a letter—the

folded tablet given by Proetus to Bellerophontes with the injunction to put the
carrier to death—communication across a spatial distance (letter writing) was not
among the earliest uses of writing in Greece. Even bearing in mind that the loss of
certain types of evidence may skew our interpretation, we have to wait until the sixth
century for the first letters. One important point emerging from the survey offered
above of the uses of writing in archaic Greece is the existence of strong regional
differentiations; similar differences apply also to the use of letter writing (or more
precisely, for this period, to the use of writing for long-distance communication),
with some regions (and some milieus) being much more productive than others.
The survey of the earliest examples of written long-distance communication
has shown that the letter is initially an oral message couched in writing; certain
verbs, such as Kغºø, present themselves automatically to the writer, since
they would have been used also for the oral order; similarly, the names of sender
and addressee are felt to be necessary, and appear in most instances from early on.
Sometimes the name of the addressee only is made, or alternatively only that of
the sender: it may be questioned whether such documents should be considered
letters. The address on the outer part of the letter is present from the earliest
instances, but it is not a systematic characteristic. As for the letter itself, only in the
fourth century does a prescript emerge, with the transformation and standardiza-
tion of the ancient, oral form ›  EÆ fiH  EØ  ºª Ø (‘A to B thus says’, found,
as we shall see, in the letters embedded in Herodotus’ Histories) into the new form
›  EÆ fiH  EØ åÆæ Ø (‘A to B, greetings’). The transition is marked by letters
such as that of Mnesiergos (no. 39), where we still have a form, K غ (‘A
sends to B’) that may be put on a par with ºª Ø, as in the old address; instead of
the demonstrative  (‘this’), however, now the verb directs the salutation,
åÆæ Ø.126 Transitional forms are observable throughout the fifth century and
into the fourth. This fits with the Greek picture of the development of the
epistolary greeting that we shall find in Lucian: while it would be dangerous to
rely too much on Lucian’s statement about the relative lateness of the salutation
with chairein in official contexts, his claim seems to be borne out by the evi-
dence.127 In private contexts, the first time åÆæ Ø is attested is at the end of a lead

125
On the initial meaning of åÆæ Ø see Wachter 1998, Calame 2005, Day 2010: 232–80. Comparison
with the enunciative stance of epigrams addressing passers-by with åÆæ Ø, åÆæ  (see below, n. 128)
might also be rewarding: see meanwhile Porciani 1997: 160–2 (but I would disagree with his evaluation of
epistolary communications as functioning through ‘modalità ben collaudate’ by the mid-sixth century).
126
Koskenniemi 1956: 156–8 makes the sensible point that a messenger in an oral context will have
begun by transmitting a greeting. The letter from Hermonassa (no. 18), with its initial vocative and its
absolute use of Kغºø, is particularly interesting in this respect.
127
Discussed below, 90–8. The Greeks will be shocked by Alexander’s use of the early format
(›  EÆ fiH  EØ  ºª Ø instead of åÆæ Ø: Plut. Phoc. 17, Aelian. VH 1. 25, but already Duris and
 Writing and Letter Writing—The Evidence 57

letter from Emporion dated to c.500 (appendix 1, no. 23), and at the opening of
another letter from the same area, dated to the same period (no. 25); but in both
instances the verb is in the second person imperative, and in the second letter the
sender’s name is not given; to find the format expected, it is necessary to wait until
the fourth century, with the letters from Panticapaeum (no. 10), from Olbia
(no. 12), and from Nikonion (no. 20), all with formulations themselves not
altogether completely ‘standard’.
The substitution of the imperative (åÆEæ ) by the infinitive (åÆæ Ø, with an
intended ºª Ø) marks a significant change. The imperative is a direct address: the
text (and through it the person who wrote it) speaks out directly, without any
mediation, just as in the early funerary inscriptions, or in the ‘epistolary curses’ we
have seen.128 This will not be the case for letters: typically, they are introduced in
the third person, and then move to first person speech.129 This structure is usually
explained by the derivation of the letter from the ‘messenger’s report’; but on its
own this is probably not enough, since some letters maintain the first person
throughout, and since the final change to the standard third-person prescript
seems to have occurred only in the fourth century, when literacy must have been
relatively widespread. Thus, there may have been some other reason: possibly the
desire, necessity, or impulse of first marking (or at least acknowledging) a
distance, before moving to the first-person speech.
Here, the contrast with the curses appears clearly: even in the curses that
present themselves as letters, the speaker speaks always, from the very beginning,
in the first person; those curses in which the term KØ º appears are in fact
first-person narratives of the sending of a message ( ø), which is the text
itself. The same verb  ø, in the first person, is used in those curses where a gift
is being sent to the underworld gods. Curses are of course an extremely personal
mode of expression, even though the name of the sender usually does not—for
obvious reasons—appear;130 the first person has a performative function. From
this perspective, a fascinating connection appears between the epistolary curses
(all coming from Attica) and what has been interpreted as the ‘letter’ of Pasion
(appendix 1, no. 42, same region and same period), in which, in the commonly
accepted interpretation, the sender addresses in the first person a third party

Chares, see below, 165), because by then the prescript with åÆæ Ø had imposed itself, and the other
format was felt as oriental: Gerhard 1905: 53 and n. 156, 60–2 (with further references), and
Koskenniemi 1956: 156. On the various formats in use by the Macedonian chancery, and in particular
the distinction between letter and diagramma, see the illuminating study of Mari 2006.
128
The inscriptions may address the passers-by, or the dead person: compare CEG i, 127; 396; CEG
i, 522; 530. 1 and 4; 655; 719; 822 (åÆEæ ); CEG i, 4; 80; 108. 1; 162; CEG ii, 487; 492; 520. 5; 677; 861. 2;
868. 1 and 6; 901 (åÆæ  ), with Stehle 1997: 312–17 (on the communicative aspects of epigrams,
without focus on åÆæ Ø), and Day 2010: 237–8, 246–53 (focusing on the semantics of åÆæ Ø).
129
Besides the instances of åÆEæ , the presence of a nominative/vocative where later the prescript
will be found is also a form of direct address (so in appendix 1, no. 7, ‘Protagoras! . . . ’; in no. 18; in no.
31, possibly because this is but a short note; the same applies to the Attic graffiti nos. 35, 36, and 37).
130
Exceptions (most, it would seem, from erotic curses) where the name of the sender is given are
listed in Versnel 1991: 63 and n. 16: DTA 231, 260, 261, 270, 271; SGD 151–3, 155, 156, 158–61; see also
Jordan SGD, pp. 186 and 191; possibly also SGD 91, and an unpublished fifth-century defixio from
Apollonia, ‘written by Aristokrate daughter of Deidis’, mentioned in Slavova 2009: 207–8 (cf. BE 2010
no. 434).
58 Greek Beginnings

([—]Æø <˜>ØŒÆØæå<øØ> Kغºø ÆıæøÆ Ø øæ ÆŁÆØ Œº), asking

him to punish other persons.131 The context seems to be a legal one, just as was
the case for one category of curses, which may explain the use of this type of
language in a letter; alternatively, this may well be a curse, written as a letter
([—]Æø <˜>ØŒÆØæå Kغºø ÆıæøÆ Ø øæ ÆŁÆØ Œº), but present-
ing the remarkable oddity of a curser naming himself.
What emerges from this comparison is the original affinity, from the point
of view of the enunciation, of both curses (qua messages to the Underworld
gods) and letters; funerary and dedicatory epigrams, where forms of åÆæ Ø
are also used in direct addresses to the passers-by, also fall in this group. But
relatively quickly the streams diverged, and letter writing settled for an imper-
sonal, third-person opening, which maintained some distance and was clearly
distinguished from spoken language, while directness remained a hallmark of
curses and funerary inscriptions.

131
See above, n. 88, and the discussion in appendix 1, no. 42.
 3
Writing and Letter Writing
Representations

Any treatment of a society’s cultural life that limits itself to the socio-economic is
bound to give only a very partial picture. All the more so in a situation in which
there is no guarantee that the documentary evidence we possess corresponds,
proportionally, to the actual types of documents in use (this principle applies
especially to the archaic and classical period, but it is fundamentally valid for all of
ancient Greek history). Let us then enter the terrain of representations and expand
on the discussion of the early diffusion, social status, and uses of writing con-
ducted in the previous chapter by looking at a specific group of representations of
writing: those dealing with its introduction or invention.1 I propose first to plot
the changes in representations and aetiologies of writing tout court, and then to
look more specifically at the aetiologies of letter writing; a discussion of the way
the Greeks perceived and explained the introductory formula (the prescript) will
conclude this part. My contention is that this may reveal to us something of the
way letter writing was conceived.
Famously, the Homeric poems do not mention writing—with one notorious
exception.2 In what amounts to the first allusion to a form of writing in Greek
literature, the writing at issue is epistolary. In the sixth book of the Iliad, Glaucus
recounts how Proetus, king of Tiryns, sent his ancestor Bellerophontes to Lycia
with a letter, in response to the lying accusations of his wife Anteia:


  Ø ¸ıŒÅ,
æ  ‹ ª Æ Æ ºıªæ,
ªæłÆ K
ÆŒØ
ıŒ fiH ŁıçŁ æÆ
ºº
EÆØ  MªØ fiz
ŁæfiH, Zçæ I
ºØ .
(Il. 6. 168–70; cf. BÆ, Il. 6. 176 and 178)

1
See Svenbro 1993 for a wide-ranging panorama of Greek representations of writing; interesting
discussion of Greek perceptions of writing in Vasunia 2001: 136–55, and esp. 148–51 for its inventors;
Vasunia, however, puts together all forms of writing, and suggests, following Steiner 1994, that the craft
was viewed negatively.
2
The other Iliadic passage where a reference to writing has been suspected is Iliad 7. 175–89: the
heroes mark tokens (ŒºBæ K Å Æ  ŒÆ ) and put them in a helmet; a herald (ŒBæı) then goes
around showing the lot that came out, until Ajax recognizes it (ªH b Œº æı BÆ N) and rejoices.
Here, the marks on the tokens are simply signs, recognizable by the person who made them, but devoid
of any meaning for the others; they function as symbolic signs, as a language (Ajax rejoices), but cannot
be confused with writing. See Heubeck 1979: 127–8; and Steiner 1994: 10–16, who stresses the
importance of differentiating between messages based on linguistic communication and semata.
60 Greek Beginnings

He sent him to Lycia, and gave him baneful signs, which he inscribed in a folded tablet,
many and soul-destroying, and told him to show them to his father-in-law, so that he
might perish.
The strangeness of this unique reference to writing had already excited the
curiosity of ancient readers, who tried to solve the difficulty by suggesting that the
signs, semata, scratched on the tablets were eidola, drawings, or else pictographic
symbols, ƒæa ÇfiØÆ, such as those that Egyptians used to explain things (so schol.
Hom. Il. 6. 168–9a; one of the meanings of the Greek verb ªæçø is ‘to draw’).3
Even so, these signs would have been designed to transmit a specific message—
they would imply communicating through writing, even if not through alphabetic
writing. It is true that, as Steiner has pointed out, there is an interesting shift in
focus from the baneful signs ( Æ Æ ºıªæ), prominent at the moment of the
inscription on the tablet, to the tablet as a sign ( BÆ) in itself, received by the
Lycian king (Il. 6. 176 and esp. 178 BÆ ŒÆŒ ).4 But ultimately, the transmission
of the message works; it is difficult to avoid the conclusion that the passage refers
to some kind of writing, given the unequivocal reference to a tablet (a folded one
moreover!)5 How to explain this unique allusion to writing?
This story combines two well-known and fairly widespread folkloric motifs,
attested separately also in the Old Testament: the married woman spurned by a
potential lover who tries to exact revenge for the rejection through her husband;6
and the letter that enjoins the addressee to kill the bearer, as in the story of David
and Uriah (2 Sam. 11) and in the Sumerian Sargon legend.7 In the first, the letter
accomplishes the trick: Uriah gives it to Joab; he is assigned by the latter to a

3
Detailed discussion of the scholia—and of the entire episode—in Heubeck 1979: 128–43; see also
Graziosi and Haubold 2010: 124–5. On graphein in Homer, see Heubeck 1979: 141, who highlights the
difference between the use of grapho in the Proetus–Bellerophontes story (Proetus writes semata on a
tablet) and all the other uses of (epi-)graphein in the Homeric poems (where the object of graphein is
the support of the incision). For a wider contextualization see Saussy 1996.
4
Important discussion in Steiner 1994: 10–17; the main points already in Heubeck 1979: 139–40.
5
So Burkert 1983. Mylonas Shear 1998 makes a strong case for the existence of Mycenaean wooden
folding tablets; this need not imply that we should see in the Homeric folded tablet a reminiscence of
Mycenaean literacy—the tablet might also reflect contemporary Near Eastern practices. Wooden
writing boards are attested in the Greek area already around 800 bc: Payton 1991; Symington 1991.
6
Cf. the story of the wife of Potiphar, Gen. 39: 1–20 (the ‘Zuleika theme’, following the name given
to the woman by the rabbinic tradition, discussed in West 1997: 365); the Egyptian story of the ‘Two
brothers’, known by a papyrus dated to 1225 bc (Pritchard 1969: 23–5); and, in the Greek tradition, the
instances discussed in Longo 1981: 62–6, and below, ch. 5.
7
Cf. Burkert 1983: 52, who makes the point that a ‘femme fatale’ (Bathsheba) has a role in the story
of David and Uriah as well. For the Sargon story see Alster 1987: 169–73, West 1997: 365–7, and Black,
Cunningham, Robson, and Zolyomi 2004: 40–4 (whose translation I follow). The Motif-Index of Folk
Literature (Thompson 1955–8) lists under K 978 (a man bears a letter enjoining that he be put to death)
Irish, Icelandic, Spanish, Jewish, Indian, Buddhist, and Japanese traditions; the variant K 511 (letter
ordering the death of the bearer, modified during its transmission, so that the bearer is instead
honoured) comprises Il. 6. 168 ff. (the grammata lugra of Bellerophontes—although the letter is not
modified, the hero indeed survives), and the Danish Hamlet legend, told by Saxo Grammaticus, Hist.
Dan. 3–4, 86–106, and rendered famous by its Shakespearean reworking. Yet another variant is listed
under K 1612 (letter resulting in the death of the sender), and H 918 (letter imposing on its bearer
heroic deeds). Further Greek parallels in Longo 1981: 62–6, and below, 140, 170–1, 222–3.
 Writing and Letter Writing: Representations 61

dangerous task, as requested in the letter by David; and he dies. In the Sargon
legend, a narrative going back to the Old Babylonian period (or possibly even the
Ur III period) and recounting the rise to power of Sargon, the focus on the
deception allowed by sealed writing is much stronger, and the use of an envelope
is explicitly mentioned as part of the plot:
In those days, although writing words on tablets existed, putting tablets into envelopes
did not yet exist. King Ur-Zababa despatched Sargon, the creature of the gods, to
Lugal-zage-si in Unug with a message written on clay, which was about murdering
Sargon. (B 53–6)8
The connection with the story of Bellerophontes is fairly close: for if in the Iliad
no stress is put on either the use of writing or the existence of envelopes, the
message given to Bellerophontes by Proetus is written on a folded tablet (K
ÆŒØ

ıŒ fiH). Clearly, the folding here has the same function as an envelope: it is
needed to keep Bellerophontes in the dark as to the content of the message.
Similarly, it is clear that the success of Ur-Zababa’s plan depends entirely on the
fact that Sargon, the messenger, will not be aware of the content of the letter. So
although this is not spelt out (being probably lost in the missing part), either
envelopes were invented for the occasion, the message was enclosed in one, and
Sargon opened it; or there were yet no envelopes, the message was simply on a
tablet, and Sargon took advantage of this deficiency in communication.9 Interest-
ingly, a femme fatale may have been part of the Sargon story too: the text we have
is incomplete, but tablet C opens with a fragmentary reference to a woman: ‘With
the wife of Lugal-Zagesi [ . . . ] | She . . . [ –—– ] her femininity as a shelter [ – ]’.10
Is the connection between Sargon and letter writing purely accidental—i.e.,
linked to the specific needs of this story, as is evidently the case for the Beller-
ophontes story? Possibly; it is, however, suggestive that in detailing Sargon’s
initiatives, the Chronicle of Early Kings notes that ‘He stationed his court officials
at intervals of five double hours.’ This same detail appears also in the neo-Assyrian
omen collection: ‘[omen] of Sargon who put his palace in order, stationed (his
court officials) at intervals of five double hours, [ . . . ] the young men stood before
him and said, ‘Let it not be so, whither should we go?’’).11 Sargon apparently set up

8
Note also that the first attempt of Sargon at eliminating his cup-bearer relies on the help of his
chief smith, Beliš-tikal, characterized by him as ‘man of my choosing, who can write tablets’ (B 30);
writing is important throughout the story. The chief smith is asked to throw Sargon and a bronze hand-
mirror into a mould for statues: substitution is clearly at work here.
9
Invention of envelope: Alster 1987; differently Cooper 1993: 18 n. 36, who deems it impossible
that the invention of envelopes should be attributed to Ur-Zababa, and who proposes that because
envelopes were not in use, Sargon was able to read the letter and escape from the trap. See also
Komoróczy 1975 (text, translation, and discussion); Cooper and Heimpel 1983.
10
TRS 73 rev., see Cooper and Heimpel 1983: 77; Black, Cunningham, Robson, and Zolyomi
2004: 44.
11
Chronicle of Early Kings, Grayson 1975: 153 (Chron. 20, 7); neo-Assyrian Omen collection: King
1907: 32 (omen no. 3, viii, 28–9), with Grayson 1975: 236. On Sargon (king 2334–2279 bc following the
Sumerian king’s list) see Grayson 1975: 235–6 (collection of sources). On the relationship between the
omen collections (earlier) and the Chronicle of Early Kings, Grayson 1975: 45–6. The royal correspond-
ence of the Assyrian kings (published in the State Archives of Assyria series; see <http://www.ucl.ac.uk/
sargon/royalcorrespondence/>) shows the importance of communication for the eighth-century bc
Assyrian empire.
62 Greek Beginnings

a chain of provinces throughout his kingdom administered by his own officials

stationed at short intervals from one another—a measure that chimes in well with
the role of courier bearing letters attributed to him by the tradition (the officials
could have passed the letters from post to post).
The story of Bellerophontes may thus have been originally a Near Eastern story
that at some point entered the Greek world, possibly through a Hittite medi-
ation;12 it certainly sits uneasily in the Homeric poems as we have them, since
writing is, in the epos, downplayed as much as possible. But it was fully accepted
as Greek by the Greek audiences of the eighth century and later; and it proved
extremely productive. As Rosenmeyer points out, the story of the letter sent by
Proetus to destroy Bellerophontes encapsulates the three major themes that will
colour letter writing through most of Greek literature—and indeed Greek history:
the deceitful letter that causes the death of its carrier; the association between
letters and deceitful women; and the basic notion of a ‘letter of recommendation’,
something very present in all ancient manuals on letter writing.13
Because of the importance of this letter as a ‘blueprint’ for later perceptions of
epistolary communication, its position in the wider context of epic communi-
cation is worth further analysis. Communication in the epos happens through
face-to-face speech, or through mediators, sent by someone to transmit a message
to someone else. Mediators comprise figures such as the ¼ªªºØ (messengers) or
the Œ æıŒ (heralds), divine or human, but also other characters who accomplish
the same function; a letter is also an instrument of mediated communication.
Although mediated communication of the type just described has a fairly import-
ant place in the epos (and in the Greek world), one may, within limits, subscribe to
Bassi’s statement, that Greek epic poetry, and the Iliad in particular, ‘validates an
idealized mode of direct first-person speech between elite males, in opposition to
mediated speech (including oral and written messages) as suspect and feminine’.14
It would certainly be excessive to mark all epic instances of mediated communi-
cation as feminine and deceptive; but the unique epic instance of a communi-
cation mediated through writing is clearly associated with deceptiveness, and if
man-to-man combat and face-to-face speech constitute one end of a spectrum,
Proetus and his secret message are clearly at the opposite end.15
The fact that a letter is the first instance of writing in Greek literature might lead
one to suppose that letter writing played an important role in Greek aetiologies of
writing; yet this does not seem to have been the case. The first author to mention
the invention of writing seems to have been Stesichorus, who in the second book

12
Oriental story: Powell 1997: 28. Katz 1998 argues for an Indo-European derivation of Bellero-
phontes’ name, to be interpreted as ‘slayer of the eel-snake’ and connected to Hittite Illuyankaš.
13
Rosenmeyer 2001: 42–4. The tragic versions of the story of Bellerophontes are discussed below, 222–3.
14
Bassi 1998: 44, and 55–9, where Bassi argues that the decision to send a written message reflects
Proetus’ diminutive masculinity (already challenged by his wife’s love for another man). Graziosi and
Haubold 2010: 122–3 point out that ‘Anteia is the only character whose words Glaucus reports
verbatim’, and contrast this with the rebuke addressed by Telemachus to Penelope, Hom. Od. 1.
356–9, ‘speaking is a concern of men’: Anteia is in many ways exceptional.
15
More on mediated communication and its characteristics below, 101–3. Llewelyn 1994: 22 high-
lights a widespread anxiety connected to letter writing in the ancient world: the earliest references to
letters in the Septuagint involve arrangements for the death of third persons (of Uriah; of Naboth, in 1
Kings 21: 9–10, a letter written by a woman; and of the seventy sons of Ahab, in 2 Kings 10: 1–10).
 Writing and Letter Writing: Representations 63

of his Oresteia attributed this feat to Palamedes; we do not know in what context
writing was mentioned (we owe our information to a fleeting reference in the
scholia to Dionysius Thrax), but in a work focused on Orestes, an allusion rather
than an extended narrative is to be expected, and this implies a pre-existing
narrative, of which we have no trace.16 Otherwise, stories about the invention of
writing begin to proliferate at the end of the sixth century and at the beginning
of the fifth; in striking contrast to the Near Eastern accounts (e.g. the story of
Enmerkar) and to ‘conventional wisdom’, these accounts do not seem to have
much to do with letter–writing.17 It is a pity that for the early period we have only
fragments of what may have been fairly extensive discussions of the origin of the
alphabet, because this makes it very difficult to locate exactly the moment in
which, and the reasons why, specific developments took place. It is at any rate
clear that at some point, the traditions narrating of a foreign and very ancient
origin of the alphabet began to be combined with those highlighting the role
played by Greek culture heroes, through positing a number of intermediary steps;
similarly, from the fifth century onwards, communication appears as one of the
reasons for the invention of writing, beside the preservation of memories. Rela-
tively early, and at any rate already in the fourth century bc, these stories were
combined in what Jeffery has aptly named ‘the Theory of Division’—a parcelling
up of the invention among multiple inventors, who each added a few letters to an
initial group, in the typical way in which divergent genealogies and stories were
also reconciled.18 In what follows, Hellenistic accounts of the invention of writing
will not be discussed: by the end of the fifth century, stories that connect the
invention of writing to letter writing are already in existence, and our focus will be
on how these narratives situate themselves in respect to other accounts of the
invention of writing.

3 . 1 . W RI T I N G AS A C R A F T I M P O R T E D F R O M O U T S I D E
T H E G R E E K W OR L D

If looked at from the angle of the function attributed to the craft, the ancient
Greek accounts of the invention of writing can be roughly organized into three

16
Schol. Dion. Thr. 183. 14 = PMG F 213; see Jeffery 1967: 152. Note that the stoicheia of
Stesichorus may have been just numbers. Kleingünther 1933: 28 suggests that the attribution of
inventions to Palamedes (besides writing, also measures and weights) may have been an Argive
innovation of the seventh century, aimed at asserting Argive priority in respect to Aeginetan and
Lydian developments in the same area; he moreover links this with the reforms attributed to Pheidon of
Argos—there is indeed an interesting cluster around Argos.
17
Story of Enmerkar and ‘conventional wisdom’: above, 23–4 and nn. 1–3. Text and translation of
the main sources on the origins of the alphabet are reproduced in appendix 2. General account of the
theories of the Greeks on the origin of their alphabet: Jeffery 1967; Garcea 2002; Schneider 2004. One
cannot but agree with Heubeck (1979: 108): the Greeks already in the sixth and fifth century had no
clear notion of where their alphabet had come from.
18
Jeffery 1967: 155. Garcea 2002 organizes most ancient sources into a tradition that assumed an
initial Greek alphabet of sixteen letters, and a tradition that counted eighteen primitive characters.
Schneider 2004 offers an extremely useful review of the material.
64 Greek Beginnings

groups. A first group is constituted by those stories, first attested in the early
Milesian writers (Anaximander, Dionysius, Hecataeus), as well as, famously, in
Herodotus, that assert a foreign origin of writing, and that do not seem to attribute
any specific, well-defined function to the new craft.19 The moment of the trans-
mission of writing to the Greeks is located in the heroic age; the craft is seen as
having been invented in Egypt or Phoenicia, with Assyria being added to the
possibilities only later. In the role of mediators are Danaus or Cadmus, the
Phoenicians as a group, or also Phoenix son of Agenor and brother of Cadmus;
the transmission to the Greeks takes place in either Thebes or the Argolid. The
figures of the inventors are rather obscure, and at any rate do not present any
personal characteristics that might make them likely candidates for such an
invention. The use made by the Ionians of terms such as çØØŒ ØÆ and
çØØŒªæÆçø, attested in two early-fifth-century inscriptions from Teos in
Ionia and indicating, respectively, the letters of the alphabet and the act of writing,
will certainly have oriented interpretation and stimulated the aetiological imagin-
ation (note also the use of
ØØŒÇ and
ØØŒÆ  in two late-sixth-century
inscriptions from Crete).20 But the outcome of the quest to justify these designa-
tions could wildly diverge, leading to the Phoenicians, to the use of red paint, or to
the habit of writing on palm leaves;21 the existence of such terms thus is not in
itself sufficient to explain the circulation of narratives of an oriental, and especially
Egyptian, origin of writing. The reason why these foreign figures are linked to
writing must then be their provenance from civilizations perceived as extremely
ancient. At play here are the early and dense contacts between Ionia and Egypt,
from the settlement at Naukratis to the movements of Greek mercenaries to the
travels of Hecataeus, and the notion of the very ancient wisdom of Egypt: Danaus
provides the mythical link between Egypt and Greece.

19
Milesians: Anaximander FGrH 9 F 3, Dionysius FGrH 687 F 1, Hecataeus FGrH 1 F 20, with focus
on Danaus; Herodotus 5. 57–61, on Cadmus. Phoenicians: Soph. Poimenes F 514 Radt. Cadmus and/or
Phoenicians: schol. Dion. Thr. 183. 1 (Ephorus, FGrH 70 F 105a), 184. 20 (FGrH 70 F 105b), 190. 19–20;
Clem. Alex. Str. 1, 16, 75. 1 (FGrH 70 F 105c); Dion. Scytobr., FGrH 32 F 8 (D.S. 3. 67. 1); D.S. 5. 58. 3;
Nonnus, Dionysiaca 4. 259–64. The Egyptian Menes is also introduced among the inventors by the
Athenian Anticlides: FGrH 140 F 11a and b = Plin. NH 7. 193 and schol. Dion. Thr. 183. 11. The
context in which Anticlides mentioned the invention is lost; but the reference to Phoroneus makes it
likely that in what followed, Danaus was introduced as the mediator to Greece. In the case of Theuth
(Plato, Phaedr. 274c and Phileb. 18b), there is a purpose to the invention, although Plato disagrees; but
it is remembrance, not long-distance communication. See further appendix 2.
20
ØØŒ ØÆ: in the Teiorum dirae (480–450 bc), M.–L. 30, 35–41. ØØŒªæÆçø: at l. 20 of a
similar curse inscription from Teos, also dated to 480–450 bc, and published in Hermann 1981 (SEG
31, 985). Crete: cf. the so-called contract of the scribe Spensithios (SEG 27, 631, from Afrati/Lyttos), in
charge of
ØØŒÇ b [
] ºØ ŒÆd ÆF= a Æ ØÆ; and IC II xii 11, from Eleutherna. Other
inscriptions referring to phoinikeia / phoinikika include: IG ii² 1456.42–3, an inventory from the
Athenian acropolis dated post 341/340, mentioning an ivory gorgoneion ªæÆ]-[ ]Æ Ø،،a å[
(cf. IG ii² 1485, inventory of end of fourth century bc); a reference in the Lindian chronicle, FGrH/BNJ
532 F 2 (B) ll. 15–17, to a cauldron dedicated by Cadmus and inscribed in Phoenician letters; IG xii 2,
96 and IG xii 2, 97, both from Mytilene in Lesbos, both preserving references to a phoinikographos, in a
(fragmentary) context where a grammateus also appears, besides a tantalizing connection to Hermes,
marked especially in the second inscription: O ç]ØØŒ ªæÆç 0EæÆØ (see Labarre 1996: 168–9). Cf. also
Hesych.  1799 KŒçØÆØ. Discussion in Chantraine 1972; Corcella 1986.
21
Schol Dion. Thr. 184. 23. For the palm leaves see Ahl 1967, who suggests that they may refer to
Mycenaean tablets shaped as palm leaves (and not to Phoenician palms).
 Writing and Letter Writing: Representations 65

Importantly, these stories share the characteristic that no particular function is

ever given to writing: it is simply invented, or rather transmitted. Some attention,
in particular in the long Herodotean account concerning Cadmus, is given to the
relationship between writing and the language it notates (as one would expect, for
a craft imported from another civilization), and to the changes that writing
underwent in time, described here with the verbs  ƺºø and  ÆææıŁÇø.22
The main contribution of the Greeks is thus seen in the change of rhythmos
imposed on pre-existing material; once again, it is important to stress that
throughout the relatively long Herodotean excursus, not a word is spent on the
function of the new craft; the same applies to the other narratives—with one
exception, a fragment of Critias, whose interpretation is disputed.
In the context of an elegiac poem listing specialties of cities or peoples, Critias
mentions the Phoenicians as the inventors of the letters, qualifying the latter with
the adjective IººªÆ, a hapax: ‘And the Phoenicians invented the letters alex-
iloga’, ØŒ  sæ ªæÆ IººªÆ (Critias D.–K. 88 F 2. 10 = Athen. 1. 28
bc). Three main interpretations have been advanced, all hinging on the specific
meaning to be attributed to the hapax alexilogos. Some understand that the Phoen-
icians discovered the letters ‘that defend the logos’, by preserving it; others give to
logos a pregnant meaning, and understand that the letters defend the logos by
helping men to think and speak. In both cases, writing would be subservient, or at
any rate complementary, to spoken discourse. The third interpretation takes
IººªÆ to mean ‘that defend from speech’, thus pitting writing against speech
(permanency, clarity, possibility of checking a text and thinking consistently, as it
were, as opposed to verbiage).23 It is a pity that the interpretation of this single verse
is so disputed, because it could give us a definite indication on what a specific
individual, the Athenian Critias, active in the second half of the fifth century bc,
considered the main use of writing. While the overall evidence from the fifth
century speaks for an interpretation privileging the use of writing to preserve
memory, the doubt remains that the oligarchic Critias, from whom one might
expect the odd sly gibe against the ‘face to face’ culture of Athenian democracy, may
have been playing on the meanings of a term, alexiloga, probably coined by him. At
any rate, here Critias, unlike most of the narratives linking writing with the foreign
origins, qualifies the invention; but this is an exception, and possibly due to the fact
that the invention is framed within a list of civilizing discoveries. Later accounts
preserve the general scheme outlined above, but combine it with stories that give the
initial role to Greek heroes or gods: the Greeks were the first to use writing, having
received it from the gods; but after the flood, they forgot about it (all written
monuments would also have perished), so that writing—again a very ‘neutralized’
writing—had to be reintroduced from the outside.24

22
Hdt. 5. 57–61; see Jeffery 1967: 153; Corcella 1986: 54–5. For the meaning of ‘change of rhythm’
(not shape of individual letters, but graphic sequence close to the oral movement of the voice) see
Calame 1993: 788–95; Magini 1998: 23–4.
23
For the first interpretation, see e.g. Fantuzzi 1984; for the second, Pfeiffer 1968: 24; for the third
(a somewhat Derridean reading of Critias), Ferlauto 1990b. In the following hexameter Thebes is
mentioned, as the place that first produced a chariot seat: mental association over a figure such as Cadmus?
24
This is e.g. the version of D.S. 5. 37 and of the scholia Vaticana in Dion. Thr. 185. 24–186. 2 (see
appendix 2); it is very close to the late (from the first millennium onwards) Mesopotamian accounts, as
attested in Berossos (FGrH/BNJ 680 F 4).
66 Greek Beginnings

3. 2. W R I T ING A S T H E I NV ENT I ON O F A GR E E K
C U L T U R E HE R O

A second group is formed by the stories ascribing the discovery to a Greek culture
hero or god. The names of the inventors are up to a point interchangeable, and
even include some historical figures: Prometheus, Orpheus, Musaeus, the Muses,
Hermes, Sisyphus, and Palamedes, as well as Pythagoras and Simonides all appear,
in different periods and contexts.25 The links between these characters and writing
are various. Sometimes, the discovery is presented as part of a more general
civilizing mission (Prometheus, Palamedes, Hermes); alternatively, the peculiar
qualities of the inventor (e.g. cunning intelligence for Hermes or Sisyphus,
memory in the case of Simonides, song for Orpheus and Musaeus) explain his
connection to the craft of writing. In this group, a focus on function is much more
prominent, possibly because writing, when it appears in this context, is mostly
presented not per se, autonomously, but as part of a series of civilizing inventions;
the new instrument, the letters, primarily enable the preservation of memories,
and promote the arts. Thus for instance the Titan Prometheus claims, in a famous
speech of the homonymous drama:
q  Pb ÆP E h  åÆ  ŒÆæ
h IŁı qæ h  ŒÆæ
ı
Łæı ÆØ, Iºº ¼ æ ªÅ e
A

æÆ ,    çØ I ºa Kªg
¼ æø ØÆ   ı Œæ ı  Ø.
ŒÆd c IæØŁ , å çØ  ø,
KÅFæ ÆP E, ªæÆ ø  ıŁ Ø,
 Å ±
 ø, ı  æ KæªÅ. (Aesch. PV 454–61)26
They had no reliable sign either of winter or of flowery spring or of fruitful summer,
but managed everything without judgement, until I showed them the risings of the
stars and their settings, difficult to distinguish. Yes, and number, too, most eminent of
sciences, I discovered for them, and the assemblages of letters, memory of everything,
occupation that gives birth to the Muses.
Prometheus gives men tools to orient themselves, in space and time, by
teaching them to discern the rising and the setting of the stars, the science of
numbers, and the assemblage of letters, all grouped under the heading of
ç Æ Æ (v. 459). Writing, intended as the ability to assemble letters, is defined
by its potential for memory and as something that helps produce poetry: it

25
On first inventors see Kleingünther 1933; Kurke 1999, esp. 248–54; and Gera 2003: 112–81,
esp. 113–25. Interchangeability: in Aesch. PV 460–1 Prometheus is the inventor of writing; but a
scholion to Aesch. PV 458 (= TrGF 3 F 182a) states that elsewhere Aeschylus had named Palamedes as
the inventor of these same arts. The scholion may, however, refer here only to the teaching of the
movements of the stars.
26
The manuscript tradition of v. 461 is divided between  Å ±
 ø and  Å Ł ±
 ø.
The second reading severs the connection of 461 with 460: Prometheus would have taught men the
assemblage of letters, as well as memory, mother of the Muses’ art. Detailed discussion in Ferlauto
1990a. On the whole, the arguments for the first reading (correspondence of ήd c at 459 and  at
460, balancing of the sentences with each time the invention and an apposition) appear convincing.
 Writing and Letter Writing: Representations 67

effectively replaces Mnemosyne as the mother of the Muses.27 The use of KæªÅ
to qualify the memory resulting from writing is particularly interesting, because it
can be related to the emergence of a language of ‘making’ song that, as Ford has
shown, places the emphasis on the construction of poetry, rather than on its
performance.28 It is worth noting here, however, that paradoxically, although the
story attributing the discovery of the alphabet to Prometheus has come to
dominate Western tradition, it is itself hardly traditional: the very detailed list of
his gifts to mankind has been often considered an Aeschylean innovation, mod-
elled on (some of) the inventions more commonly attributed to Palamedes.29
From the fifth century onwards, the invention of writing is also attributed to
Orpheus, Linus, and Musaeus.30 The reason for the invention is, in the case of
Orpheus, probably poetry, because of his connection with the Muses (the same
applies to Musaeus), as well as wisdom. This is explicitly stated in an epigram,
supposedly inscribed on the tomb of Orpheus and quoted in a speech attributed to
Alcidamas:
ªæÆ Æ b c
æH  Oæçf K ªŒ,
Ææa ı H ÆŁ, ‰ ŒÆd K
d fiH
 Æ Ø ÆP F źE a K
تæÆ Æ·
ı ø
æ
º fi B’ OæçÆ ¨æB fi Œ ŁÅŒÆ,
n Œ  łØø Zf łº  Ø ºØ
ˇNªæı çº ıƒ , n  HæÆŒºB’ KÆ,
æg IŁæ
Ø ªæÆ Æ ŒÆd çÅ. ([Alcid.] Od. 24)
Orpheus first introduced the letters, having learnt them from the Muses, as is shown
also by the epigram on his tomb:

27
See Nieddu 2004: 47. Ferlauto 1990a suggests interpreting ªæÆ Æ as ØåEÆ (meaningful
elements rather than purely graphic ones), and understands this to mean ‘combinations of letters
articulated orally’: the Titan would have given mankind an articulated language, able to remember
everything and to produce poetry. While such an interpretation is unlikely (if only because slightly later
in the play there is a pointed allusion to the metaphorical writing of the Titan’s instructions by Io, in the
tablets of her mind), the phonic quality of the Greek alphabet renders it close to a transcription of
spoken language, and many early Greek descriptions of writing characterize it as a double of spoken
language: besides Ferlauto 1990a, see Calame 1993 (on the rhythmos of the letters in Hdt. 5. 61); Magini
1998.
28
Ford 2002: 131–9; for tragedy, see esp. 137–8. As Piccaluga 1988: 41 notes, Prometheus does not
‘create’ the letters: he shows how to put together (or how to understand) elements pre-existing in
nature.
29
Kleingünther 1933: 83 and n. 38; Sommerstein 2000: 122; see also below. Prometheus’ connection
with writing depends solely on the homonymous drama. On the evolution of the figure of Prometheus
see Saïd 1985: 115–30 and 131–54; Gantz 1993: 153–66; West 1997: 581; Gera 2003: 120–2. The
question of the drama’s authenticity is not important here; West 1990: 51–72 denies Aeschylean
authorship, but arguments based on inconsistencies or structural incoherence are hardly convincing
(see Saïd 1985 for a very different approach). The date to c.430 bc endorsed by Sommerstein 2000: 121
would align the play with the interest in writing that emerges during this period. Jeffery 1967: 155 n. 9
stresses the uniqueness of the mise en scène of Prometheus in the Prometheus vinctus: ‘clearly the poet
did not intend that the great Titan, transformed by him from a minor cult-hero of Attic potters into the
first civilizing force of mankind, should be seriously ranked by later scholars as a claimant with the
mortals Palamedes and the rest’.
30
Linus: Boardman 1992, Bremmer 1999; Musaeus: Kauffmann-Samaras 1992, Heinze 1999;
Orpheus: Linforth 1931 and 1941; Calame 2000; Steiner 1994: 194-201; Graf and Johnston 2007. For
the iconography, see Garezou 1994 and 2009. Orpheus teacher of Linus: schol. Eur. Rhes. 895 =
Asclepiades of Tragilus FGrH 12 F 6b, D.S. 3. 67. 2; teacher of Musaeus: D.S. 4. 25. 1.
68 Greek Beginnings
The Thracians buried Orpheus here, the servant of the Muses | whom lofty-ruling
Zeus slew with the smoking thunderbolt | the dear son of Oiagros, who taught
Herakles, | having discovered writing and wisdom for mankind.
While the date of the speech is uncertain, the—clearly fictional—epigram could
have been composed in the Athenian milieu of the last quarter of the fifth
century;31 we have from that period numerous testimonies of an association
between Orpheus and books or writing, both literary and iconographic.32
Versions that portray writing as a craft imported from the outside had localized
the place of the transmission mainly in Argos or Thebes. The ‘Orphic’ version of
the invention of writing is not really localized: the epigram adduced in its support
had been inscribed on Orpheus’ tomb, in Thrace, or so Alcidamas states, but it
circulated (and was possibly created) in Athens; it does not say anything about the
place of the invention. The same lack of localization is apparent in the version in
which writing is given to mankind by Prometheus, or in those that set the
invention in the Greek camp at Troy; it is also reflected in the fact that in these
stories the letters are simply called ªæÆ Æ.
For these traditions had to contend with the theory of the Milesians, that
writing had originated in older civilizations (such as Egypt, or later, Assyria);
similarly, Herodotus had claimed that the name phoinikeia reflected an ancient
reality and was to be explained by the Phoenician origin of writing. When this
observation imposed itself, the problem became to avoid that name, or to explain
it away in a Greek way. As a result, a number of strategies were put in place that
managed to retain the invention on Greek ground. Thus, in discussing the entry
phoinikeia grammata (‘Phoenician letters’), the Suda (ç 787) and Photius (ç 652
Porson) preserve the following information:
¸ıd ŒÆd ø a ªæÆ Æ I
e ØŒ F ª æ F æ  ·  Ø b
I غªı Ø ˚æB , ‰ æŁÅ I
e F ªæçØ K çØŒø
 ºØ. ( . . . )
Lydians and Ionians say that the letters received their name from Phoenix son of
Agenor, who discovered them. But the Cretans disagree with them, saying that the
name derives from writing on the leaves of palm-trees. ( . . . )
The Ionian and Lydian theories mentioned at the opening of the passage are
simply variants of those traditions that linked writing to the Phoenicians. Doing
away with Cadmus and concentrating on Phoenix offered the advantage of
explaining the word phoinikeia as applied to letters through the (Greek) name

31
The question of the authorship and date of the Odysseus, or, Against Palamedes (see O’Sullivan
2008 for a date after the first century ad) has to be kept separate from that of the date of the epigram, on
which see Linforth 1931 and Linforth 1941: 15.
32
e.g. Eur. Hipp. 954; Eur. Alc. 966–9; Plato Resp. 2. 364e3–365a3; and four Attic vases, all dated to
the second half of the fifth century: hydria Basle, Antikenmuseum BS 481, LIMC, s.v. ‘Orpheus’, no. 68;
hydria from Attica, now Dunedin (N.Z.), Otago Museum E 48. 266, ARV2 1174. 1, LIMC, s.v.
‘Orpheus’, no. 69; red figure oinochoe from Basle, BS 1416, CVA Switzerland 8 (Basel iv), p. 39; cup
from Naples, now Cambridge, Fitzwilliam Museum, ARV2 1401. 1, LIMC, s.v. ‘Orpheus’, no. 70.
Discussion in Schmidt 1972; Steiner 1994: 195–7. Fascinatingly, the connection between the head of
Orpheus and writing is taken up in Etruscan iconography, and extended to include Palamedes: see De
Puma 2001, and below, n. 63.
 Writing and Letter Writing: Representations 69

of their alleged discoverer rather than from that of a foreign people.33 But Phoenix
son of Agenor, brother of Cadmus and of Europa, may have been felt to be still too
close to the Phoenicians: for the discovery of the letters is attributed to quite a few
other Greek ‘Phoenixes’, besides the son of Agenor. According to Duris, the letters
owed their name to Phoenix the teacher of Achilles;34 why Duris should have
mentioned this in the eighth book of his Macedonica is unclear (possibly as an
annotation to Macedonian education?) At any rate, for Duris the invention had
taken place at the time of the Trojan war. Alexandros of Rhodes, for his part,
attributed the invention to another Phoenix, the son of Europa and Pronapes (or
Pronapos), who discovered the letters in Crete, as a result of which he was killed
out of jealousy by Rhadamanthys.35 The remark about the death of Phoenix at the
hands of Rhadamanthys because of his discovery of the letters is interesting: as
Schneider remarks, the stories about Palamedes, Prometheus, and Linus show that
inventing letters does not always produce positive results for the discoverer.36 Just
as interesting is to find that the comic poet Anaxandrides attributed to Rhada-
manthys and Palamedes the invention of geloia, jokes brought as a contribution to
the banquet by those who did not bring any other contribution (the asymboloi):37
these accounts are all related to each other. The chronological level is again that of
the Trojan war: Pronapes/Pronapos (better known as Pronapides, and usually an
Athenian, not a Cretan) was, according to Dionysius Scytobrachion (FGrH/BNJ 32
F 8), a contemporary of Orpheus and teacher of Homer, who had used ‘Pelasgian’
letters; his son, who invented ‘Phoenician’ letters, was a contemporary of Homer.
But local versions circulated as well. These firmly situated the invention of
writing on Greek soil—indeed, on a specific part of it. One in particular is worth
discussing, because it shows how important a topic writing and its uses were in
fifth-century Athens.
In the same entry ‘phoinikeia grammata’, the Suda and Photius preserve the
following information going back to Skamon, a fourth-century historian from
Mytilene on the island of Lesbos and author of a work On Inventions:
Œø  K fi B ı æfi Æ H ¯æÅ ø I
e ،ŠB `Œ Æø OÆ ŁBÆØ.
ıŁ ÆØ b y  Iæ ø b
Æø ¼
ÆØ, ª ŁÆØ b ÆP fiH ŁıªÆ æÆ `ªºÆıæ,
¯æ Å, —æ · c b ،Š Ø
ÆæŁ s Æ ºı B ÆØ· Øe ŒÆd ØØŒ ØÆ
a ªæÆ Æ e `Œ ÆøÆ, ıº   Ø ØB I
EÆØ fi B ŁıªÆ æ. (FGrH
476 F 3 = BNJ 476 F 3)

33
For Phoenix son of Agenor and Cadmus see Pherecydes, FGrH 3 F 21; M. L. West 1985: 82–3. For
the notion (problematic in my opinion) that in a lost Euripidean play (Phrixos B, fr. 819 Kannicht)
Cadmus might have taken this name after migrating to Greece, having originally been named Phoenix,
see Gantz 1993: 209–10.
34
Duris FGrH/BNJ 76 F 6 = schol. Dion. Thr. 184. 27–9 (and again anonymously at schol. Dion.
Thr. 32. 12–13, 182. 16–17 and 185. 9; see appendix 2.)
35
Schol. Dion. Thr. 184. 29–185. 2 (and again anonymously at schol. Dion. Thr. 192. 7). Jacoby, ad
FGrH 32 F 8, c. 67, qualifies this as ‘Schwindelnotiz’ (with a question mark), built on the tradition
concerning an Athenian Pronapides (also mentioned in schol. Dion. Thr. 183. 20–1), who would have
‘arranged that the letters be written as we write now’.
36
Schneider 2006: 135–6, with further references (but the case of Prometheus is not really similar to
the others). See also Piccaluga 1988 for general considerations in this direction.
37
Athen. 14. 614 c (fr. 10 K.–A.): ÆÆæÅ ’ K ˆæ ÆÆ fi ŒÆd æ a H ªºø çÅ d
ª ŁÆØ  !ÆÆŁı ŒÆd —ƺÆ Å, ºªø o ø· ŒÆ Ø
ºº ª
F. | e I º yæ
ªºEÆ ºªØ  !ÆÆŁı ŒÆd —ƺÆ Å.
70 Greek Beginnings
Skamon, in the second book of his On Inventions, says that they were named after
Phoenice, the daughter of Actaeon. Legend has it that he was childless when it came to
male children, but that to him daughters were born, Aglaurus, Erse, and Pandrosus.
But Phoenice died while still a parthenos; for this reason Actaeon named the letters
‘Phoenician’, desiring to bestow some honour upon his daughter.
The scholia to Dionysius Thrax show that the same story was also narrated by
two other fourth-century historians, Andron (of Halicarnassus or Ephesus) and
Menecrates of Olynthus: it must have enjoyed some diffusion.38
None of the three historians who discussed this story is Athenian; yet the name
of the king, Actaeon, and even more so the names of the three other daughters,
Aglaurus, Erse, and Pandrosus, point to Athens. Actaeon, the king of Acte (the
original name of Attica), is a double of Cecrops, the autochthonous first king of
Athens: the two share the same three daughters; moreover, Cecrops, to whom a
number of civilizing inventions were attributed, is said by Tacitus to have invented
writing.39 But a Phoenice is otherwise unknown: her name was clearly invented to
account for the by then common denomination of phoinikeia for the early
characters of the alphabet. Here, the discoverer, Actaeon, is at the same time a
name-giver: he assigned a name to his daughter, and at her death, he assigned her
name to the letters (also his daughters, if only metaphorically), so that she might
be remembered. Svenbro has brilliantly elucidated the way in which this story was
created: it is closely modelled on that of Cadmus, first king of Thebes, and also
father of four daughters, Agave, Autonoe, Ino, and Semele. Agave, Ino, and
Semele all end, for various individual reasons, in a disastrous way, exactly as the
daughters of Cecrops did, when they opened the chest in which Athena had put
Erichthonius. As for Autonoe—‘she who thinks for herself ’, the fourth daughter
of Cadmus—she had a son, named Actaeon: as Svenbro says, these accounts seem
to ‘inter-think each other’.40 Building a purely Attic story on the Theban model
allowed the Athenians to claim priority for the invention of writing by linking it to
the very origins of their city. It is impossible to ascertain whether Skamon’s
account goes back to Hellanicus (according to tradition, his father), who had
also written a book On Discoveries; but clearly, if the three versions are found in
authors of the fourth century bc coming from three fairly different areas of the
Greek world, the story must go back to the fifth century, and one would expect it
to have originated in an Athenian milieu.
More interesting, but just as difficult to answer, is the question of the meaning
of this story. Already Jacoby had seen that this is an Attic invention; should we put
it on the same level as those concerning Prometheus and other Greek inventors, or
should we assign a special significance to the Athenian origin of the discoverer?
And if the latter, as seems likely in view of the elaborate construction of the story,

38
On Skamon see Jacoby, FGrH 476 F 3; Kaldellis, BNJ 476 F 3. For Andron (of Halicarnassus), see
FGrH 10 F 9 and Toye, BNJ 10 F 9. Bollansée, FGrHCont IV, 1005 F 5/6, thinks, however, that schol. Dion.
Thr. 184. 20 refers to Andron of Ephesus. Menecrates is mentioned in the same sentence of the scholia to
Dionysius Thrax; nothing else is known of him, and the fourth-century date is only an inference deriving
from his origin (Olynthus was entirely destroyed in 348 bc) and his association with Skamon and Andron.
39
Svenbro 1993: 81–2; Tacitus, Ann. 11. 14. Note that Cicero attributed to Cecrops the invention of
funeral rites (leg. 2. 63); other civilizing inventions are listed in Scherf 1999.
40
Svenbro 1993: 82. On Cadmus’ daughters, see Gantz 1993: 472–81.
 Writing and Letter Writing: Representations 71

should this be interpreted as a sign of Skamon’s (or rather his source’s) goodwill
towards Athens? As a rejection of the Ionian or, more specifically, Milesian
tradition focused on Danaus? As a rejection of the Theban tradition focused on
Cadmus? Would the Athenians, in hearing this story, have been struck by the
‘deadly’ character of writing, its immobility, or would they rather have perceived it
as the living counterpart of the dead girl, as able to produce children for her?
It is impossible to give a definite answer to such questions. Jacoby saw this story
as a specific instance of a larger pattern of ad hoc traditions, a pattern according to
which already in the fifth century Greek gods, heroes, or humans are occasionally
mentioned as inventors of writing (so Prometheus, Palamedes, and possibly even
Musaeus) and linked to a definite place, without, however, ever being able to
match the Milesian theory of an Egyptian or Phoenician origin of writing.41 Three
more points may be made.
First, the complex manoeuvre of reinventing under an Athenian cover the
tradition about the arrival of the letters with Cadmus testifies to a relatively
important Athenian involvement with the question of the origins and purposes
of writing. Such an involvement is attested both at an ideological level and at a
practical one. To start with the latter, Athens underwent a reform of the alphabet
in 403 bc that led to the adoption of the Ionian alphabet, a reform that must have
occasioned considerable debate.42 As for the ideological involvement, Greek
drama—and in the fifth century this means mainly Athenian drama—brings to
the fore in an unprecedented way writing and its implications, attesting the lively
interest of both playwrights and their public in the issue; attention to writing and
its uses will remain a constant of drama.43
Second, aetiological tales that identify a Greek origin for writing, be they local or
Panhellenic, present a striking difference from the Greek stories that portray writing
as a craft imported from the Near East or Egypt, inasmuch as they foreground the
uses of writing. Thus, in this specific story, writing is connected with memory after
death, with kleos; a similar emphasis on writing’s ability to preserve memories
through time, as well as, less frequently, on writing’s creative potential, is typical of
the majority of the stories that see in writing the invention of a Greek culture hero.
Third, we may want to read the contrast between the two types of stories as a
contrast between a vision of writing as something that is deeply inscribed in the
nature of the world, but which ultimately—for this very reason—is beyond human
control (and thus often linked to death), and a second, more optimistic vision of
writing as a skill acquired from another civilization, imported, and thus ‘negoti-
able’, with all the implications this has in terms of alphabetization.44

41
Jacoby, FGrH 476 F 3, Kommentar. I would dissent as to the gods: they do not seem to play an
important role in the fifth-century traditions on writing. See e.g. Henrichs 2003a and b.
42
D’Angour 1993; Grassl 1972 suggests that this is why Herodotus dedicates a long (and polemical)
excursus on the arrival of writing in Greece with Cadmus.
43
Detailed discussion of writing in tragedy and comedy below, ch. 5. On the invention of alphabetic
letters by an unknown character, in Philemon fr. 10 K.–A., see below, 257–8.
44
So Piccaluga 1988: 42–3, who further sees in the second, ‘optimistic’ approach a ‘pathetic attempt
to redeem their [the letters’] disturbing alterity’. Piccaluga’s is a fascinating structuralist reading of the
evidence—some of her associations, however, are forced. For the at times negative connotations of
writing (and for the deep and disturbing presence of writing in the world and even within the human
body), see also Steiner 1994.
72 Greek Beginnings

3 . 3 . T H E IN V E N T I O N O F W R I T I N G AS L E T T E R W R I TI NG

There are, however, a few stories, all going back to the fifth century bc, in which
the main use of writing appears to be that of allowing—or enforcing, as the case
may be—communication at a distance: the Euripidean version of the story of
Palamedes; the story of the invention of writing by the Oriental queen Atossa, as
recounted by Hellanicus; and the narrative of the introduction of writing among
the Medes by Deioces, related in the first book of Herodotus’ Histories.45

3.3.1. An Ill-fated Greek Hero: Palamedes

Two factors combine in making Palamedes exceptionally interesting: in some

versions of his story, epistolary communication is presented as one of the original
purposes of writing (to the point that Detienne has characterized him, strikingly,
as ‘un héros épistolier’);46 and long-distance communication has an important
role in the plot itself, so that it is possible to see how the invention is considered in
this particular light. The story clearly means that epistolary communication is
perceived as an important aspect of writing; but what is the sense, what are the
values, of the connection between Palamedes and letter writing?
Palamedes is not mentioned in the Homeric poems, a fact much discussed by
the ancients—and the moderns. Powell makes much of a Euboean connection,
whose existence rests on the fact that Mount Caphareus in Euboea is the place
where Nauplius lights the devious fires that will cause the death of the Achaeans
by drowning. But the scholia to Euripides’ Orestes make clear that Euboea is not
the home of Nauplius: after having tried to obtain justice at Troy, Nauplius went
back home; then, having learnt that the Achaeans were leaving Troy, he went to
Euboea, waited for a storm, and lit fires.47 The only other sources for a Euboean
origin of Palamedes are a passage in an oration of Gregory of Nazianzus, in which
a Euboean origin is mentioned, with qualifications; and the remark made by a
later commentator on this very passage.48 This is rather thin evidence; the
Euboean connection is stressed mainly by those scholars who are interested in
finding in ancient traditions a reflection of modern theories on the spread of
literacy.

45
Rosenmeyer’s (2001: 26) remark on the ‘supposed invention of the imaginary letter in Greek
prose by Lesbonax, a rhetorician named by Lucian’ (without references) is not correct: Luc. De Salt. 69
mentions a Lesbonax interested in the movements of the arms of the dancers; the scholia add that this
Lesbonax, a rhetorician, was famous for various rhetorical works, and in particular for his Kæø ØŒÆd
K
Ø ºÆ (‘erotic letters’); but there is no question of invention.
46
Detienne 1989: 109. Detailed discussion of the variants in the story of Palamedes in Falcetto 2002;
see also Powell (1991: 233–6 and 1996: 25–6), who imagines a ‘real’, Euboean Palamedes, grosso modo
contemporary of Homer. See now the excellent discussion of Jenkins 2006.
47
Powell 1991: 234; contrast schol. Eur. Or. 432.
48
Gregory of Nazianzus, or. 4, 107 (Against Julian 1, 644): Y
æ ¯Pf › —ƺÆ Å (ironically,
in view of the use that some try to make of the Euboean origin, not mentioning the invention of
writing); and Pseudo–Nonnus, Scholia mythologica or. 4 hist. 62. Powell’s reference to Eudox. 321 Blass
(1997: 234 n. 34) derives I suppose from a misunderstanding of the abbreviation ‘Eudok’ in Lewy
(1897–1902: 1264): this is the Violarium of (Pseudo) Eudocia, a sixteenth-century forgery.
 Writing and Letter Writing: Representations 73

An Argive origin of the hero is much better attested: apart from the explicit,
although late, statements of Tacitus (Ann. 11. 14. 9–11) and of the Suda (
44), a
number of other sources point to it, if in a rather indirect way, either through
stressing the Argive connections of his father, Nauplius, and of his brother, Oeax,
or, as in Pausanias, through reference to local monuments.49 The assumption of
an Argive origin of the hero makes it possible to connect the emergence of the
figure of Palamedes in this area with the stories concerning the invention of
coinage, weights, and measures by Pheidon of Argos in the seventh century bc.
After all, the inventions that are most stably linked to Palamedes are letters,
weights and measures, and taxeis (ordering by ranks); and in a context in which
various other areas (Lydia and Aegina in particular) contended for the paternity of
civilizing inventions, going back to Palamedes may have appeared a valid way of
bolstering one’s claim to priority.50 In this context, it is worth noting that
Palamedes, son of Nauplius, is linked to Danaus, one of the heroes to whom
Greek tradition attributed the invention of writing. In his Argonautica (1. 133–8),
Apollonius Rhodius offers a genealogy in which Nauplius (Palamedes’ father;
Palamedes himself could not of course be mentioned in the Argonautica, since the
story precedes the Trojan war) descends from Danaus through his father Clyto-
naeus, son of Naubolus, son of Lernus, son of Proetus, son of Nauplius, son of
Poseidon and Amymone, herself daughter of Danaus. As for Danaus, by the fifth
century at the latest he appears in the Greek genealogies as son of Belos.51
None of the early versions of the story of Palamedes we possess mentions his
lineage, except for references to his father Nauplius and his connection to
Poseidon; but by the time of Virgil, Palamedes is directly and explicitly connected
genealogically to the family of Belos (and thus Danaus): Sinon refers to Palamedes
as Belides, a descendant of Belos (Virgil, Aeneid 2. 82).52 While such an explicit
connection is late, one can hardly avoid noticing that the only definite mention of
‘written signs’ in the Iliad is the folded tablet on which King Proetus, also a descendant
of Danaus and thus part of Palamedes’ family (although there are a number of
variants in the genealogical details), scratches semata lugra, a writing ‘avant la
lettre’—or more precisely, a writing produced before a younger member of the
same family could (re-?)invent the use of letters. A thread seems to have been running
through the family, from Danaus to Proetus to Palamedes—attracting along the way

49
Oeax, brother of Palamedes, is mentioned as one of the Argive citizens particularly hostile to
Orestes in Eur. Orestes 432. If the plot of the Euripidean Palamedes (produced before the Orestes: the
latter’ production is dated to 408 bc) contained an attempt at murdering Oeax, as suggested by Luppe
2011 on the basis of P.Mich. inv. 3020(A), Oeax’s hostility is even more understandable. For Pausanias
2. 20. 3 the remote antiquity of the temple of Tyche in Argos is proved by the fact that in it Palamedes
dedicated the dice (Œı) that he had just invented. Eustathius (in Il. 2. 308, 228.8) says that
according to some, Palamedes’ own draught (
 ) was to be seen in Argos. Kleingünther 1933:
27 rightly takes the Argive origin as self-evident.
50
Kleingünther 1933: 28; and 82 for the basic connection between Palamedes and the inventions of
numbers, weights and measures, letters, and ranks. Of course, these are contested traditions, and there
may have been a Euboean Palamedes, just as there may have been a Euboean Io (see Mitchell 2001); but
by the moment when writing (and epistolary writing) enter the story, the Argive tradition has imposed
itself, and it constitutes the background for the elaborations on the story of Palamedes.
51
See Gantz 1993: 203 for the upper part of the line (Io, Belos, Danaus, and Aegyptus); West 1997
442–55.
52
On Palamedes in Virgil see Horsfall 2008: 110–18.
74 Greek Beginnings

also Cadmus.53 Argos thus plays a pivotal role in many of the traditions on the
invention or transmission of writing;54 the craft figures quite importantly among the
prerogatives of the descendants of Inachos. It is probably not by accident that writing
is present specifically in the heroic stemma that more than any other was extended to
include the Eastern Mediterranean; but within this stemma, Palamedes appears as
thoroughly ‘Greek’.
The legend of Palamedes is certainly ancient. The hero played a relatively
important role in the Cypria: he uncovered the feigned madness of Odysseus,
and thus forced him to join the expedition against Troy; Odysseus avenged
himself, with the help of Diomedes, by causing the hero’s death through drowning
during a fishing expedition.55 A short sentence in Ps-Apollodorus’ Bibliotheke
(2. 1. 5) states that according to the tragic poets Palamedes was the son of
Nauplius and Klymene, but that in the Nostoi his mother was Philyra, while
Kerkops (reputedly author of an Aegimius) mentioned as his mother Hesione.
This shows that Palamedes, or at any rate his father Nauplius, played a role in the
Nostoi (fr. 1 Bernabé) and in the Hesiodic Aegimius (fr. 297 M.–W.), but also that
the traditions concerning the hero were still relatively unstable.56
It is unclear how far back in time the tradition reaches that assigns to him the
invention of letters: a fleeting reference in the scholia to Dionysius Thrax informs
us that Stesichorus in his Oresteia (fr. 213 PMG) attributed to Palamedes the
invention of ØåEÆ (letters, or possibly numbers); unfortunately, the context in
which the discovery was made is not given, nor its purpose. Palamedes can only
have been incidentally relevant to the Oresteia: he might have been mentioned as
the reason of the hatred of Oeax, Palamedes’ brother, for Orestes, attested in
Euripides’ Orestes (Eur. Or. 432). At any rate, from this the inference has been
made that Stesichorus was alluding to a story already known.57 Such an inference
is most likely correct, but it need not mean much in chronological terms:

53
Although mainly known for his Phoenician origin and his link with the city of Thebes, Cadmus
is, in the fifth century, part of the same family as Danaus: the extended narrative of Io’s story in
Bacchylides’ Dithyramb 19 leads to Cadmus son of Agenor, and from there to Dionysos. Discussion of
Cadmus’ genealogy in Edwards 1979: 23–9, 47–8, 165–8; see also Jacoby, ad FGrH 1 F 20 and FGrH 70
F 105; Hecataeus of Abdera, FGrH 264 F 6; Kleingünther 1933: 131. Fascinating discussion of the story
of Cadmus, its antiquity, and its meaning in Gruen 2010: 233–9.
54
The connection with Euboea dear to Powell should be considered, keeping in mind the possibility
of an Euboean Io, argued for by Mitchell 2001. At a late-sixth- to early-fifth-century level, at any rate,
our sources are clear on the Argive connection. An Argive presence can be felt even in Herodotus’
account: the expulsion of the Cadmeans from Boeotia by the Argives is mentioned at the beginning of
the excursus on writing (5. 57. 2) and at its end (61. 2): the Argives somehow cause writing to become
‘phoinikeia’ (before, the letters are still Cadmean, as on the tripods).
55
Resp. Procl. Chrest. Ep. 119–21 Severyns = Cypria arg. 30 ff. Bernabé; and Procl. Chrest. Ep. 166
Severyns = Cypria arg. 66 Bernabé, with also Paus. 10. 31, 2 = Cypr. fr. 30 Bernabé. For the debate on
the antiquity of the story of Palamedes, and in particular on whether Homer intentionally erased
Palamedes from his epics, see Usener 1994: 50–6 (who rightly stresses that it is uncertain whether
Palamedes’ sophia was already highlighted in the Cypria); Falcetto 2002: 11–12. According to the
antiquarian Polemon (F 32 Müller = Eust. in Il. 2. 308, 228. 2–4), a stone marked, on the plain of Troy,
the place where Palamedes used to play dice and draughts.
56
The fifth century marks a turning point. Note that Ps-Apollodorus does not explicitly say that in
the Nostoi Nauplius was Palamedes’ father; the earliest clear statement is in Eur. IA 198 (Palamedes son
of the son of Poseidon); cf. Jacoby ad FGrH 3 F 4.
57
Jeffery 1967: 152.
 Writing and Letter Writing: Representations 75

Stesichorus might have recounted the story of Palamedes at greater length in his
Nostoi, where the vengeance of Nauplius would have been pertinent. In this case,
the passage in the Oresteia might be interpreted as an allusion to a theme already
broached elsewhere by Stesichorus himself, and there would be no need to go
much further back in time for the story. The further inference that Hecataeus
would have rejected Palamedes’ claims to the invention in favour of Danaus is
unsupported by any evidence; while Palamedes might, it is true, have been
mentioned in an aside, there are no compelling reasons for Hecataeus to have
mentioned him.58
The next author to mention Palamedes after Stesichorus is Pindar, who in a
very fragmentary poem seems to state that Palamedes surpassed Odysseus in
sophia (fr. 260 S.–M.). The very fragmentary lines may be reconstructed so as to
refer to a dispute, possibly even to a hidden logos and a fishing net, but it is
impossible to know exactly what was going on. The (probable) fishing net might
show that we are still here in the tradition of the Cypria.
We have thus reached the fifth century. Importantly, it is only from this
moment onwards that the tradition is attested according to which the death of
the hero was caused by a false letter accusing him of betrayal and planted by
Odysseus. This seems to have been a new development, specific to drama. At any
rate, a passage of Polyaenus shows conclusively that the tragedians abandoned the
older version of the Cypria, implying a violent death by drowning, to focus instead
on a version implying a false accusation by Odysseus and the ensuing condemna-
tion of Palamedes after a debate in front of the army.59 A forged letter need not
have been always part of the dramatic plot, but the false accusation and the public
debate were: the new version offered to the playwrights the possibility of staging a
public debate, and of exploring issues linked to decision-making, trust, and
truth.60 Thus, in the course of the fifth and the fourth centuries bc a debate
concerning the functions and uses of writing developed around the story of
Palamedes. This debate started in drama but became central in the sophistic
and philosophical milieu, as shown by the Apology of Palamedes of Gorgias,
possibly by the Odysseus attributed to Alcidamas, and by a few other scattered
references.61 The relative scarcity of depictions of Palamedes may also be taken to

58
Jeffery 1967: 154, Heubeck 1979: 106, and Powell 1991: 5–6 all propose, against Jacoby (FGrH
1. 323–4), that Hecataeus in suggesting Danaus as the inventor of writing was implicitly polemicizing
against Stesichorus and the poetic tradition, who supported Palamedes. This, if it were true, would of
course imply a greater diffusion of the story of Palamedes than the fragments we have would lead us to
believe, and also a strong connection, at an early time, of the hero with writing.
59
Polyaen. 1 prooem. 12: x b ŒIŒE æÆ ªÅÆ Oı ø ƒ æƪø fi dfi ¼ı Ø. —ƺÆ Å
KŒÅ  Oı f K ØŒÆ Åæø fi H åÆØH 
ƺg ÆP F fi B ŒÅB fi ÆæÆæØŒe åæı , ŒÆd ›
ç Æ  H  Eºº ø KŒE lºø
æ Æ  ºø fi ŒÆd æÆ Åª Æ Ø. See Scodel 1980: 43–4 for the
difference between the epic and the tragic/rhetorical Palamedes.
60
Polyaenus’ summary is compressed, but no letter is mentioned, although the gold is. Cf. Engel
2000: 319: ‘Crucially, it is not clear whether Palamedes’ invention of writing and his being undone by a
forged letter were both regular details in the story, or even whether they were features of Sophocles’ and
Euripides’ [or rather Aeschylus’?] plays. The forged note was not always part of the story. Gorgias’
Palamedes points out that no such evidence exists’. Contra, e.g. Usener 1994: 56–64, for whom a letter
(or even two) was part of all the tragic plots.
61
O’Sullivan 2008 proposes a much later date for the Odysseus attributed to Alcidamas (after the
first century ad). Independently of the question of date and authenticity, I accept Scodel’s (1980: 46–7)
76 Greek Beginnings

confirm that outside tragedy and oratory the character of Palamedes did not
attract much interest: before the fifth century, the hero appears only once, on a
Corinthian pyxis of about 570 bc as part of a named group comprising Protesi-
laus, Nestor, Patroclus, and Achilles riding to the right, while Hector and
Memnon ride to the left.62 In fifth-century iconography Palamedes appears only
twice. An Attic red figure calyx crater dated to c.440 bc depicting a Nekyia has
Palamedes (inscribed "`¸`¯˜¯) leaning on a column before Persephone
and holding an oar (an allusion to the means by which Oeax informed his father
Nauplius of the events? or to the older story of his death by drowning, as is
commonly assumed?). And in Pausanias’ description of Polygnotus’ Nekyia in the
Lesche of the Cnidians at Delphi, a gathering of Odysseus’ victims has Palamedes
playing dice with Thersites, with Ajax close by.63 In its epistolary version, then, the
story of Palamedes begins as a fifth-century Athenian story, closely linked to
drama and oratory.64
Four Palamedes plays are known to have existed, by Aeschylus, Sophocles,
Euripides, and Astydamas the younger; they seem to have differed in quite a few
aspects. Unluckily, only a few fragments are preserved: two and four, respectively,
from Aeschylus’ and Sophocles’ plays; thirteen from the Euripidean Palamedes;
none from that of Astydamas. Any reconstruction, therefore, must perforce base
itself on the use of divergent accounts in later sources, with all the methodological
problems this entails.65 Not even the order in which the various Palamedes plays
were staged is certain: the first one was Aeschylus’; the Palamedes of Euripides was

argument that the Odysseus does not directly reflect any specific play, and that ‘all in his work that is
not free invention is no doubt a hodge-podge of bits adapted from various sources which could easily
be unconnected with Palamedes’. The arrow bearing a message, for instance, is first found in Herodotus
8. 128, and will have a long life as a stratagem for secret communication across enemy lines.
62
Pyxis Louvre E 609, Cf. Woodford 1994a: 147, no. 9, who notes that ‘it is unlikely that any specific
story is referred to’.
63
Crater New York MMA 08. 258. 21; and Paus. 10. 31. 1–3 (Woodford 1994a: 147, nos. 7 and 8).
For the later period, Woodford 1994a lists only four representations, and five more uncertain ones; see
also Woodford 1994b for a fuller discussion. Palamedes seems to have enjoyed a much greater success
among the Etruscans, with nine named depictions: particularly interesting is an Etruscan bronze
mirror of the third century bc, on which the hero (Palmithe) is depicted seated with a diptychon in
his lap; besides him (E)linai (Helen), Ziumithe (Diomedes), Euturpe, Umaele, and another masculine
figure (Krauskopf 1994, 148, no. 16, and 149; Da Puma 2001). See also above, n. 32.
64
For the phases in the development of the story see Usener 1994; Gantz 1993: 603–5; Jouan 1966:
339–63. List of sources in Lewy 1897–1902: 1264-73 (to which add Ion of Chios FGrH 392 F 2 = Athen.
10. 426 e, in which Palamedes, rather uniquely, figures as a mantis). Jouan and van Looy 2000: 503 note
that the false accusation is a fairly widespread folkloric motif (they compare the story of Joseph and
Benjamin in Genesis 44, and the trick by which the Delphians free themselves from Aesop, schol. Ar.
Vesp. 1446, Plut. de sera num. 12, 556 f, P.Oxy 1800, fr. 2), and feel that it is impossible to date the
moment it entered the story of Palamedes; Aeschylus is at any rate the first to have brought it on the
scene.
65
See the Notice in Jouan–Van Looy 2000: 487–513. Proposals of reconstruction of the plots of
these dramas: Scodel 1980: 43–63; Müller 1990; Usener 1994: 56–64; Falcetto 2002: 17–21 (for the plays
of Aeschylus and Sophocles) and 171–95 (for the Euripidean Palamedes). None of these authors could
use the fragment from the end of the hypothesis of the Palamedes published by Luppe (2011). To the
versions of the story I shall discuss should be added [Apollod.] Epit. 3. 8 (a fairly simple version); Serv.
in Aen. 2. 81 (close to Hyginus); and those details that can be gleaned from Gorgias’ Defence of
Palamedes, from the Odysseus attributed to Alcidamas, from Philostratus’ Heroicus, and from the prose
paraphrase of Euripides’ Philoctetes contained in Dion of Prusa’s speech 59. 1–11.
 Writing and Letter Writing: Representations 77

part of the trilogy comprising also Alexander and Troades, the latter usually dated
to 415 bc; as for the Sophoclean plays linked to the story (the Palamedes, but also
the Nauplius navigans), they might have preceded, but also followed, the Eur-
ipidean one. Thus, what follows next is best taken as an overall picture of the
main, common characteristics of the ‘tragic Palamedes’, although some differ-
ences in the plots will be highlighted. We shall then look more specifically at the
role of letter writing in Euripides’ play.
Our main source for the ‘tragic’ Palamedes is a scholion to Euripides’ Orestes
(432). According to the scholiast, Palamedes’ first intervention in favour of the
Achaeans took place during their involuntary stay at Aulis: he solved the difficul-
ties caused by the rationing of food by showing them the use of Phoenician
characters, presumably for numbering the rations (
æH  b a ØŒØÆ
ØÆ ªæÆ Æ ÆP f Y Å  ŒÆd I
ºÅ
 c ØÆc K  Ø
K
æƪÆ  Æ ). Here the idea of a Phoenician origin of writing is combined
with a Greek hero, Palamedes, and with distributive numbering. Rather than
seeing in this a memory of the ‘real’ origin of writing and of its initial purpose,
I would argue that this is a result of fifth-century and later amalgamation of the
various traditions concerning the invention of writing. On the same occasion
Palamedes was also said to have invented measures, as well as games of dice and
draughts, thereby offering the troops distraction from hunger and inaction
(Œı . . . ŒÆd  æÆ KFæ ŒÆd łBç). This provoked the jealousy of Agamem-
non, Odysseus, and Diomedes, however, who decided to destroy him. They
forced a Trojan prisoner to write a letter in Phrygian characters as if sent by
Priam to Palamedes, revealing the latter as a traitor (MªŒÆ Æ ªæłÆØ æıªØ
ªæÆ Ø
æd
æ Æ ‰
Ææa —æØı
æe —ƺÆ Å). Next, having
persuaded a servant of Palamedes to hide the letter and the ‘Trojan’ gold
under Palamedes’ bed, they accused the hero of betrayal and ordered a search
of his tent. The letter and the gold were found, and the Achaeans stoned
Palamedes to death.66 The scholion continues with the arrival at Troy of Nau-
plius, who has heard of the events, and with his request for justice. The army,
however, disregards his request, preferring to please their leaders. Nauplius thus
leaves for home, but prepares his revenge.67
Two small points are worth singling out. If we can trust the wording of the
scholion, in this version of the story Palamedes does not really ‘invent’ writing, but
simply transposes Phoenician characters (are these only the words of the

66
According to a number of scholars (notably Carl Robert; more recently Jouan 1966: 345, Müller
1990, Usener 1994: 57), the scholion preserves the intrigue of the Euripidean Palamedes, because a
scholiast to Euripides would most likely have given the Euripidean version, and because of the
accumulation of bad characters—Agamemnon, Diomedes, Odysseus—all united to destroy the hero,
something that is typical of the late Euripides (and more specifically of Troades, which was part of this
trilogy). Scodel 1980: 50–2 objects that this intrigue allows only one agon, since the letter and the gold
are discovered at the same time; she thus prefers to see here the intrigue of the Aeschylean homonym-
ous tragedy.
67
This last part—in particular the arrival of Nauplius at Troy—coincides with what we learn from
the fragmentary final part of the hypothesis of the Euripidean Palamedes recently published by Luppe
(2011). However, if Luppe’s reconstruction is correct, in the Euripidean play there was also an attempt
at killing Oeax by drowning: of this no mention is made in the scholion, nor of the message sent by
Oeax (Nauplius arrives ‘having heard’, ˝Æ
ºØ b IŒ Æ wŒ). See below.
78 Greek Beginnings

scholiast? or, if the term phoinikia to describe the letters was present in the source,
why were they called so—was there a link to his Phoenician lineage in the source
that the scholiast is following?). Moreover, the purpose of his invention is clearly
stated, namely to help with the rationing and distribution of food. However,
Palamedes’ ruin is caused by a letter, one written in Phrygian characters: writing
seems in the plot described to be conceived as already having a wide—and fairly
differentiated—currency. Not only that. While there seems to be no sense of a
difference in language (the Greeks do not need, at least in the compressed version
of the scholion, an interpreter to convey their message to the Trojan prisoner),
writing is coded in a variety of ways: pre-existing Phoenician characters are
adapted by Palamedes to the necessities of the Greeks and Phrygian characters
are used by the Trojans. This has the further implication that Palamedes is
convicted by a piece of evidence that the Greeks cannot decipher, for otherwise
they would not have needed a Trojan prisoner to write it. We possibly see here the
trace of a reflection on the ability of signs to work on their own, beyond their
meaning. This is a prominent feature of a number of ‘tragic’ letters, Euripidean in
particular.68
Hyginus’ fabula 105 preserves a slightly different version. Odysseus, deter-
mined to ruin Palamedes, convinces Agamemnon of the necessity to move the
camp for one day, and then during the night hides gold where the tent of
Palamedes had been. He then gives a letter for Priam (presumably written by
himself: the foreign aspect of the writing is not stressed in this account) to a
Phrygian slave, whom he despatches to Priam, having previously arranged for him
to be killed while on his journey. On the following day, when the army comes back
to the camp, the body is found, and with it the letter. It turns out that it contains a
message sent by Priam to Palamedes, promising him as much gold as Odysseus
has hidden in the tent if he will betray the Greeks according to the proposed
agreement. An agon followed in which Palamedes defended himself; to prove his
innocence, he asked for his tent to be inspected, but at this point the gold was
found and Palamedes put to death.69
Servius offers in his commentary to the Aeneid (2. 81) a fairly similar version,
but adds some novel elements, one of which is a genealogy: Palamedes descended
at a remove of seven generations from Belus (Servius is here discussing the
expression used by Virgil in Aen. 2. 82, Belidae Palamedis). It is impossible to
know when exactly the genealogical connection between Belus and Palamedes was
established (it is explicitly stated only in Virgil), but it might go back all the way to
the fifth century bc—after all, as we saw, Apollonius Rhodius provides a genealogy
connecting Danaus and Palamedes, and the link between Danaus and Belus is a
relatively early one. Another interesting aspect of Servius’ version, one that may
with some likelihood go back to the Sophoclean play, is that Odysseus is portrayed
as particularly devious: he is the one who, when presenting himself as a supporter

68
See below, 224–30, for a discussion of the function (and functioning) of Iphigenia’s letter in
Iphigenia among the Taurians. This Euripidean element should be taken into account in the discussion
of whether the scholion reflects the Euripidean or the Aeschylean Palamedes.
69
According to Scodel 1980: 47–54, this fabula represents the main lines of the Euripidean version.
Usener 1994: 62 imagines a ‘Verdoppelung’, a duplication of the letter motif, in what he identifies as the
Sophoclean play, which is not borne out by the texts.
 Writing and Letter Writing: Representations 79

of Palamedes after the letter has been found, suggests that in order to clear the
hero of any remaining suspicion his tent should be searched.70
While it is possible to notice differences in emphasis and while the range and
scope of Palamedes’ inventions show some variations, the plots make clear both
Palamedes’ standing as an inventor and the role played in his demise by a false
accusation, in most cases supported by a letter. Although already in the Cypria
Palamedes had shown his sophia by unmasking Odysseus’ stratagem, the first
explicit mention of inventions in connection with him had been in Stesichorus’
Oresteia. But as an inventor, Palamedes ‘takes off ’, together with a few other
figures, in the fifth century. Each of the Palamedes plays by Aeschylus, Sophocles,
and Euripides included a list of his inventions, pronounced by the hero or by
someone else in his defence; a list of discoveries was also in the Sophoclean
Nauplius katapleon (or pyrkaieus), relating the story of the vengeance of Pala-
medes’ father. That all these lists have been preserved, even if only in short quotes,
shows the interest of later authors for this aspect of the story. It is worth looking at
them in more detail, both for the indications they give of the different aspects of
the hero’s character stressed by the three tragedians, and because they offer the
possibility of contextualizing Palamedes’ invention of writing among his other
discoveries.71
Writing is not mentioned as one of Palamedes’ discoveries in the preserved
fragments of Aeschylus’ homonymous play. In fr. 181a Radt, probably from the
defence speech of Palamedes, the hero asserted, in terms very close to those used
by Prometheus in the Prometheus vinctus, that ‘I organized the life of the Greeks
and their allies (
 Å  Eºº ŒÆd ıåø), which previously had been as
chaotic as that of beasts. To begin with, I invented the ingenious art of number,
supreme among all techniques’.72 The invention of numbers implies that of the
letters, since, with the partial exception of Attica, where the acrophonic system
was used, numbers were expressed through the ordered sequel of the alphabetic
letters. It is thus possible that writing might have been mentioned later in the
speech, as is the case with the Prometheus; but it is impossible to know what
aspects of the craft would have been stressed. In fr. 182 Radt, from the same
speech, further specific inventions are recounted: ‘and I appointed brigade
( ÆØæåı) and company commanders (%ŒÆ  æåı) for the army, and
I taught them to distinguish their meals, to take breakfast, dinner, and thirdly
supper’. The speech was certainly much longer, and there is no guessing what else
could have been mentioned. The focus on military aspects and on distributive

70
See e.g. Müller 1990: 204–6 and Falcetto 2002: 20, who base this interpretation on Sophocles
fr. 479 Radt, a praise of Palamedes pronounced by someone else: a lying Odysseus, rather than
Palamedes’ father Nauplius (the other possibility).
71
A point stressed by Detienne 1989: 105–8. Kurke (1999: 248–53) understands Palamedes’ inven-
tions as symbolic systems intimately connected with the civic, democratic order. Yet in the context of
the story of Palamedes, I would sympathize with Steiner’s (1994) approach: writing seems to be
perceived as a craft potentially dangerous for the democratic city. See also Vasunia 2001: 149 and n. 29.
72
The fragment is quoted by Stobaeus directly after the similar tirade of the Prometheus vinctus (see
discussion above). Here and elsewhere for Aeschylus’ Palamedes I follow Sommerstein’s (2008) Loeb
translation. The numbering of the fragments is the same in both editions with the exception of fr. 451k
incertae sedis Radt, considered by Sommerstein (2000 and 2009) to be from the prologue of the
Palamedes and thus assigned by him the number 180a.
80 Greek Beginnings

organization is at any rate evident: these two fragments are very close to the
summary offered by the scholion to Euripides’ Orestes.
In Sophocles’ Palamedes the hero’s discoveries were recounted by someone else,
either Nauplius, or, as seems more likely, Odysseus being duplicitous. The main
theme of the single extant fragment is the resolution of the famine, probably in
conjunction with food rationing and with the invention of various pastimes
(inventing pastimes is often connected with surviving famines):73
‘Was it not he who drove famine away from them, be it said with reverence towards
the god, and he who discovered the cleverest ways of passing time for them when they
were resting after their struggle with the waves, draughts and dice, a pleasant remedy
against idleness?’ (fr. 479 Radt, trans. Lloyd-Jones)
In the same way Sophocles in the Nauplius navigans had another person, this
time Palamedes’ father, enumerate the benefits brought to the Greeks by his son:
‘And it was he who devised the wall for the army of the Argives; his was the invention
of weights, numbers and measures; he taught them to marshal armies thus and how to
know the heavenly signs (PæØ  Æ Æ). He was the first, too, who showed how
to count from one to ten and so to fifty and to a thousand; he showed the army how to
use beacons ( æÆ F çæıŒ øæÆ), and revealed things that earlier were hidden. He
discovered how to measure terms and periods of the stars, trustworthy signs (
Ø a
ÅÆ æØÆ) for those who watched while others slept, and for the shepherds of ships
at sea he found out the turnings of the Bear and the chilly setting of the Dogstar.’
(Soph. fr. 432 Radt, trans. Lloyd-Jones)
In this long series of eleven iambic lines, the army and practical wisdom are
very much on the forefront, possibly simply because of the location of the action
(the Achaean encampment at Troy); yet, what is stressed throughout is Pala-
medes’ ability at interpreting signs, and at rearranging them in a meaningful
way.74 Another fragment, probably from this same part of the Nauplius navigans,
picks up on aspects that had been highlighted in the fragment from the Sopho-
clean Palamedes: the hero invented ‘the draughts-boards with five lines (
 a

 ªæÆÆ) and the throwing of the dice’ (fr. 429 Radt). The discovery of
writing is potentially implied in this last fragment; however, nothing of what
remains of Aeschylus or Sophocles allows us to know how the invention was
presented or what aspects of it were stressed.75

73
Herodotus 1. 94, the story of how the Lydians who then colonized the Tyrrhenian coast survived
a famine lasting eighteen years by playing every other day, offers a good parallel—interestingly
Herodotus, while not mentioning alternative versions, stresses that it is the Lydians who pretend to
have invented the games of dice and knuckle-bones and ball, and all other games but draughts, Hdt. 1.
94. 3. Cf. the disapproval in Athen. 1. 19 A; and Kurke 1999: 248–54.
74
An aspect highlighted by Jenkins 2006: 19–20; see already Purcell 1995. One may compare with
Herodotus’ insistence on the necessity of rearranging the sequences of letters (their rhythmos) in order
for the Phoenician signs to function in Greece: cf. Calame 1993. The grouping of weights, numbers, and
measures acquires importance only in the second half of the fifth century, reflecting the striving
towards IŒæØÆ, precision, typical of the moment: Heinimann 1975.
75
In fact, only the discovery of numbers was certainly present in all tragedies: cf. Plato’s amused
comment (Resp. 522d) that in all tragedies Palamedes makes Agamemnon appear ridiculous, a leader
unable to count his soldiers—indeed, unable to count his feet. Bertrand 1999: 64–6 emphasizes the
connection between writing and war (for which see also below, on Gorgias’ Palamedes), and hence the
political meaning of writing; but he does not mention Palamedes.
 Writing and Letter Writing: Representations 81

It is only with a fragment from Euripides’ Palamedes that we gain an insight

into the value and the function attributed by Palamedes/(Euripides) to his inven-
tion. In this passage, Palamedes is speaking, almost certainly in the context of the
agon in which he defended himself against Odysseus:
a B ª º ŁÅ çæÆŒ OæŁ Æ  ,
¼çøÆ ŒÆd çøF Æ, ıººÆa ØŁd,
KÅFæ IŁæ
Ø Ø ªæÆ NÆØ,
u P
Ææ  Æ
 Æ 
bæ
ºÆŒe
IŒE ŒÆ YŒı
 K
 Æ ŁÆØ ŒÆºH,

ÆØ  I
Łfi Œ Æ åæÅ ø  æ
ªæłÆ Æ º
Ø, e ºÆ  Æ  NÆØ.
L  N æØ

ı Ø IŁæ
Ø ŒÆŒa
º  ØÆØæE, ŒPŒ Kfi A łıB ºªØ. (Eur. fr. 578 Kannicht)
‘Alone, I set up remedies against forgetfulness, without speech and yet speaking, by
creating syllables;76 I invented for mankind knowledge of the letters. Thus, one who is
away on the other side of the marine expanse knows perfectly well what is going on
there, at home, and when he dies he leaves for his children the measure of his fortune,
having recorded it,77 and the recipient knows it. The evils that afflict men when they
enter into quarrels a writing-tablet solves, and it doesn’t allow the telling of lies.’
Here, writing has lost its connection with rationing and organization; the
primary functions of the discovery now are the preservation of information
(memory) and the enabling of long-distance communication.78 The hero appears
fully convinced of the merits of his discovery; the examples of possible uses of
writing that follow illustrate the potentialities of the new medium. However, in the

76
The language is possibly technical, at any rate difficult. Various translations of this line have been
proposed: I have kept close to that suggested by Collard and Cropp 2008, who, however, also entertain
in a footnote the alternative possibilities of ‘consonants and vowels, by creating syllables’ and ‘by
creating consonants and vowels and syllables’. Scodel 1980: 91 accepts the correction çø  Æ and
takes ¼çøÆ çø  Æ to have here their later well-attested meaning of ‘surds and sonants’; Jouan and
van Looy 2000 also accept çø  Æ and translate as an asyndeton, ‘by creating consonants, vowels and
syllables’ (a variant is ‘by organizing consonants and vowels into syllables’, as Jenkins 2006: 20).
77
There are here two textual issues that impact importantly on the meaning. The first one concerns
l. 6: I
Łfi Œ Æ is the transmitted text, accepted by Kannicht. It can be understood as referring to
the person who is away, and then dies; but, as Kannicht says, this form is unparalleled in tragedy. Thus,
Neri 1997: 168 prints the words between cruces; Jouan and van Looy 2000 and Collard and Cropp 2008
accept Wecklein’s  e Łfi Œ Æ, ‘and if someone dies’ (so also Jenkins 2006: 20), which is also
problematic; see Neri 2007: 169 for full discussion and list of other proposals. With the transmitted text,
distance is the reason for writing the will; with Wecklein’s emendation, the reason becomes the
possibility of checking it. The second issue concerns l. 7, where both Kannicht and Collard and
Cropp 2008 print Scaliger’s emendation ªæłÆ Æ º
Ø. Stobaeus, who preserves the passage, has
here ªæłÆ Æ N
E; while the accusative plural in the transmitted text is problematic, the connection
between writing and speaking is tempting. Neri 2007: 168 thus accepts Boissonade’s ªæłÆ Æ ª’ N
E,
while Jouan and Van Looy print Gomperz’s ªæłÆ KØ
E, ‘tells, having recorded it’. See Neri 2007
for a full apparatus and list of loci similes.
78
For Turner 1975: 17 the purpose is memory only, as in Aeschylus and possibly Stesichorus.
Turner accepts, as everyone, that the second example of writing refers to the possibility of leaving a
testament; but in the first he sees an allusion to the Ionian historia, to documents such as the map of
Anaximander, the works of Hecataeus, and those of Herodotus later. I find this difficult to accept in
light of the ŒÆ YŒı, ‘at home’; moreover, it is hard to see what sense a reference to Ionian historia
would make in the context of the play.
82 Greek Beginnings

context of the play, these words of Palamedes are charged with heavy irony.79
A letter—a forged letter, feigning private communication between a Trojan and a
Greek, but in reality publicizing to the army the false accusation of a Greek against
another Greek—will cause the death of the hero. As for the advantages of writing
to dispose of one’s fortune, it is true that occasionally sources mention the
usefulness of writing in legal contexts, and especially to write down laws or
wills.80 However, a written testament, that is, a securely guarded text, was looked
at with mistrust in Athens even in the fourth century bc: thus goes yet another of
the advantages mentioned by Palamedes.81 Finally, as the protagonist realizes in
the course of the action, no tablet in itself can pretend to the truth.82 In fact, the
very last sentence of the fragment, on the possibility of solving disputes through
the use of tablets that do not allow falsehoods, announces the next moment in the
action: the discovery of the tablet accusing Palamedes of betrayal.
But if this final sentence on the truthfulness of tablets is so evidently related to
what follows, one may justifiably wonder whether the allusion to the usefulness of
writing for long-distance communication, far from being a traditional motif, does
not also owe its presence here to dramatic strategy: it anticipates an action that
takes place later in the drama, when Oeax, Palamedes’ brother, tries to send a
message to their father Nauplius.83 For Euripides’ play had a further epistolary
twist, which must have been important enough to be parodied in Aristophanes’
Thesmophoriazusae (vv. 768–84 = Eur. fr. 588a Kannicht). At this point in the
play, Euripides’ kinsman, having been unmasked as a woman, has already tried to
help himself out of his plight by using stratagems, scenes, and props taken from a
famous Euripidean play, the Telephus, but has failed in his endeavours. He now
wants to inform Euripides of the situation, and thus looks for a messenger, but
cannot find one; remembering the method used in Euripides’ own Palamedes, he
decides to inscribe oar blades and throw them in the sea. As he cannot find any
oars, the kinsman settles for votive tablets, which he then proceeds to throw in all
directions, happily noting that ‘they’re wooden too, just like oar blades’ (775). The

79
Irony: Scodel 1980: 116, who demonstrates that this ironic reversal permeated the entire trilogy—
and possibly the satyr drama as well (Scodel 1980: 122–37).
80
See e.g. Aesch. Eum. 273, on the written accounts of Hades. If laws and wills need to be written, it
is so that they may be remembered with precision and checked if necessary; this notion can thus be
subsumed under the more general idea of writing as a means of, and prompt for, memory (cf. Hedrick
1994).
81
Maffi 1988: 188–210, and esp. 203. See also below, 269–72, 292; and compare Ar. Vesp. 583–7:
Philocleon stresses, among the advantages that sitting in the juries gives him, the power deriving from
the fact that ‘if a father dying gives his heiress daughter to someone, we tell his will, and the shell so
solemnly placed over the seal, to go hang its head, and we give that girl to whoever talks us into it. And
for doing this we can’t be called to account’ (Œi I
Łfi Œø ›
Æ æ fi ø fiH ŒÆ ƺ
ø
ÆE
K
ŒºÅæ, | ŒºØ &E ÆŒæa c Œçƺc N
  fi B ØÆŁ ŒfiÅ | ŒÆd fi B Œ ªåfiÅ fi B
ı H E
ÅØ Ø K
 fi Å, |  Æ Å ‹ Ø i &A I غ Æ IÆ
 fi Å. | ŒÆd ÆF Iı
ŁıØ
æH). Philocleon here clearly thinks of a written will—a powerless will, however.
82
Jenkins’ (2006: 21) stress on the innovation implied in Odysseus’ invention of forgery is excessive:
the process of communication may be manipulated on the oral level as well, as the stratagems of
Themistocles and Leotychidas (Hdt. 8. 221. 2 and 9. 98. 3–4, discussed below, 116–18) show. Yet there
is in the play something that reminds one of the emphasis put by Thucydides on Nicias’ decision of
sending a letter (Thuc. 7. 8. 2, discussed below, 142–6).
83
Already noted by Jouan 1966: 351 n. 1. On this episode’s organization see Scodel 1980: 58–9.
 Writing and Letter Writing: Representations 83

scholia to these verses (schol. Ar. Thesm 770a and 770b) explain that Euripides in
his Palamedes had Oeax inscribe the story on oar blades that he then abandoned
to the sea so that at least one might reach Nauplius; the oars, floating over the
Aegean, were meant to reach Greece and to transmit the information.
Interestingly, the votive tablets of the kinsman do not produce an immediate
effect; it is only when the kinsman turns to yet another Euripidean tragedy and
resorts to playing Helen that Euripides/Menelaus arrives. A recently published
papyrus fragment preserving the end of the hypothesis of the Euripidean Palamedes
seems to imply that in the Palamedes too, Nauplius may have arrived in Troy
independently of the message—or at any rate that he received it when already under
way, so that the message sent by Oeax would have been of limited usefulness. More
precisely, the fragment of the hypothesis begins with the recovery by Nauplius of
one son, thrown into the sea and saved from drowning by the Nereids: possibly it
was Oeax himself that informed his father, and not the tablets at all.84
A connection between the stress on writing as a means of communication at a
distance, in the words pronounced by Palamedes in the agon, and Oeax’s strata-
gem is rendered likely by the fact that none of the mythographical sources
mentions either Oeax, who is a rather obscure character anyway, or the ‘bottle-
in-the-sea’ stratagem.85 Surely if some such story had been present also in
Aeschylus’ and Sophocles’ versions the scholia would have mentioned it. Interest-
ingly, Libanius (De Socrati silentio, 28) eliminates Oeax and presents a Palamedes
himself writing the narrative of the events on the very ships of the Achaeans
(presumably just before being stoned).86 The passage of the Thesmophoriazusae
too may be taken to imply the relative obscurity of the character Oeax, for there
the kinsman says that he will imitate the Palamedes, and ‘like that fellow’ (‰
KŒE, 775) inscribe oars: Oeax is not explicitly mentioned. Yet the scholia to the
passage make it abundantly clear that in Euripides’ Palamedes it was Oeax who
inscribed the story:
› ªaæ ¯PæØ
Å K fiH —ƺÆ Ø K
Å  e ˇYÆŒÆ e Iºçe —ƺÆ ı
K
تæłÆØ N a
º Æ87 e ŁÆ  ÆP F, ¥ Æ çæ ÆØ %Æı ÆE ºŁø Ø N

84
P.Mich. inv. 3020(a) ll. 1–5, with Luppe’s (2011) proposed restorations, runs as follow: I
/ or
ή 
-] | ŁÅ F  []b IÅ ŁÅ- |  
e g ˝æø ›
Æ- | [ ]cæ IÆŒØ  ˝Æ
ºØ-
| 4 [] I
غ Æ ªÆ[Ø] _ _ _
-
_ _ _ | [ºØ] K
ºı ; see the comments of Luppe 2011: 54–5.
85
See Müller 1992. Oeax is mentioned, with Nausimedon and Palamedes, as son of Nauplius in
[Apollod.] Bibl. 2 (23) 1. 5. 14; as the only brother of Palamedes in [Apollod.] Bibl. 3 (15) 2. 2. 1 and in
schol. Eur. Or. 432; Pherecydes (FGrH 3 F 4) mentioned a son Damastor. Out of desire for revenge,
Oeax came to help Aegisthus and Clytaemestra (according to Hyg. Fab. 117, he excited Clytaemestra’s
anger by reporting that Agamemnon had taken Cassandra as his wife); on a painting, possibly by
Polygnotus (in which case dated to the second quarter of the fifth century bc), seen by Pausanias (1. 22.
6) near the Propylaea in Athens, Orestes was depicted killing Aegisthus, while Pylades killed the sons of
Nauplius. In Pacuvius’ Doulorestes (vv. 113–15, 136, 137–8 Ribbeck3 = frr. 88, 89, 90, and 105 Schierl
2006) Oeax was promised to a daughter of Clytaemestra: probably Erigone, daughter of Aegisthus,
rather than Electra. Detailed discussion in Schierl 2006: 240–79, and esp. 243–4 for Oeax.
86
Liban. Decl. 2. 28: 
  ŒÆd —ƺÆ Å e ç Æ  H   Eºº ø
Ææa ŒÅ
I
ŁÆE, q Æ ªaæ ŒI ºø fi Øb @ı Ø ŒÆd ºÅ Ø, Iºº’ Pb KŒE
æe F ŁÆ ı Øø
A
KŒº , Iºº’ KB ÆP fiH ŒÆd ºªØ ŒÆd ªæçØ, ŒÆd ªæłÆ K
d F ºı F Æı ØŒF c %Æı F
åÅ 
ł fiH %Æı F
Æ æd ˝Æı
ºø fi c K
ØŁÆ Ø K
Ø º .
87
Here the Ravennas 429 has ÆF;
º Æ is a correction of Enger accepted by Holwerda, probably
because of schol. Ar. Thesm. 770bÆ (quoted in the next note). In view of Libanius’ text, maybe the
transmitted ‘ships’ should be maintained.
84 Greek Beginnings
e ˝Æ
ºØ e
Æ æÆ ÆP F ŒÆd I
ƪªºø Ø e ŁÆ  ÆP F.
(Schol. Ar. Thesm. 770a Holwerda)
for Euripides in the Palamedes had Oeax the brother of Palamedes inscribe on the oar
blades his death, so that being carried on their own they would arrive to Nauplius his
father and announce his death.88
In this case the scholia should be trusted: because Aristophanes does not
mention the name of Oeax anywhere (of course his point is a comparison between
the kinsman and the wise Palamedes), the commentators must have known the
story from somewhere else—from Euripides’ play or a summary of it.89 Libanius’
attribution to Palamedes of the writing on the oars may be explained on the
assumption that the hero instructed his brother, but even better with the obscurity
of the figure of Oeax, and with Libanius’ specific interest in Palamedes.90 At any
rate, the pointed references to Euripides in the Aristophanic parody, better
understandable if the stratagem is considered untraditional, an invention of the
‘frigid’ Euripides;91 the obscurity of Oeax (whose name, ‘Helm’, besides being
appropriate to the family’s connections with Poseidon, may have also been a
function of the story, as he steers information, through oars, and through his
recovery); and the absence of any references to the staging of a similar stratagem
by the earlier tragedians all speak for the non-traditional character of the story.92

88
Cf. the fairly similar schol. Ar. Thesm. 770bÆ Holwerda: u
æ ˇYÆ fiH ˝Æı
ºø fi ªæçØ K fiH
—ƺÆ Ø ¯PæØ
ı. › ªaæ ˇYÆ KªåÆæ Ø
ººÆE
º ÆØ a
æd e —ƺÆ Å ŒÆd IçÅ Ø N
ŁºÆ Æ, u  Øfi A ª ØØ e ˝Æ
ºØ
æ
 E.
89
See on this issue Scodel 1980: 58–9.
90
Libanius’ passage shows why the character of Palamedes acquired so much importance: a rich
tradition, starting early in the fourth century bc, puts together the unjust trial of Palamedes and that of
Socrates, with Palamedes becoming the archetypical figure of the just who is unjustly accused. See
Kannicht ad fr. 588 for references; Usener 1994: 65–72, who stresses the ‘politicization’ of Palamedes.
The earliest connection between Socrates and Palamedes is put in the mouth of Socrates himself, in
Xen. Apol. Socr. 26. Diog. Laert. 2. 44, who quotes the fragment ‘You have killed, have killed, the all-
wise innocent nightingale of the Muses’, states that Euripides had Socrates in mind (but then adds a
quote from Philochorus affirming that Euripides died before Socrates, as indeed is the case). In this
sense, the rapprochement between Palamedes and Prodicus (rather than Socrates) proposed by Wor-
man 2002: 183–4, in her illuminating study of the uses of the characters of Odysseus and Palamedes in
the last third of the fifth century and first quarter of the fourth, cannot be taken as central. Of course the
story of Palamedes could be used to think about any situation in which an ‘intellectual’ was unjustly
accused: thus also the name of Protagoras, who may have been tried for impiety in Athens in the 420s
(Diog. Laert. 9. 54), has been suggested. The paradox of the inventor who suffers from the beneficiaries
of his invention is fully brought out e.g. by Dio Chrys. Or. 13. 21.
91
The Palamedes as ‘frigid’ (łıåæ ): Ar. Thesm. 848. The song in Ar. Thesm. 768–84 (‘Hands of
mine, you must put your hand to an effective job. Tablets of planed board (
ØŒø  H º Ø),
accept the knife’s scratching, harbingers of my troubles (Œ æıŒÆ KH  åŁø)! Damn, this R is
troublesome. There we go, there we go! What a scratch! Be off then, travel every road, this way, that
way, and better hurry!’, in Henderson’s Loeb translation) probably reflects very closely Euripides’
highly emotional anapaests: see Kannicht ad fr. 588a, Jouan and van Looy (2000) fr. 9, Collard and
Cropp 2008 ad fr. 588a. and below, 261–2. Incidentally, the absence of Oeax from the scholion to the
Orestes and from Hyginus fabula 105 shows how perilous it is to make either of those summaries the
basis of a reconstruction of this play.
92
The new fragment of hypothesis P.Mich. inv. 3020(a), showing that—if one follows Luppe’s 2011
restorations and interpretation—the Achaeans attempted to kill Oeax by drowning might have been a
Euripidean stratagem to combine the new death through stoning, which resulted from the ‘epistolary
twist’ of the plot, with the traditional notion of a death by drowning, as in the fragment of the Kypria.
 Writing and Letter Writing: Representations 85

Moreover, it is worth noting that at the level of the plot, Oeax’s use of
communication in Euripides’ Palamedes neatly dovetails with what had happened
to Palamedes earlier in the same play. As pointed out by Jenkins, the latter is
undone by a letter having the wrong sender and addressee (purportedly Priam and
Palamedes; actually, Odysseus and the Greek army), thus a letter that does not
have a straight path. The former is unable to send his message directly to the
intended recipient, but somehow arranges things so that at least one of the many
inscribed oars is steered towards the addressee, Nauplius.93 This statement might,
however, have to be revised in light of the fragmentary hypothesis preserved in
P.Mich. inv. 3020(a): it is possible that Oeax’s messages too went to the wrong
addressees, since at some point the Achaeans attempted to drown Oeax, and they
must have had some reason for this; moreover, if the latter was saved by the
Nereids and recovered by the father, it is possible that the father eventually got his
information through face-to-face conversation. At any rate, if all this is correct,
the connection between the invention of writing by Palamedes (a fairly traditional
element, but mainly connected with the rationing of food) and long-distance
epistolary communication, whether successful or not, would have been a feature
specific to Euripides’ play.94
From this moment onwards, communication at a distance, whether involving
writing or not, will loom large in the stories about Palamedes. An instance very
close in time to Euripides’ play is the Palamedes of Gorgias.95 In this speech, the
hero argued his defence point by point, discussing opportunity, associates and
motive, and character. Palamedes opened his discussion of opportunity by stress-
ing that for him to have agreed with Priam to betray the Greeks, there should have
been a logos, a speech, and a meeting ( ıı Æ)—but how could this have
happened if there were no messengers nor tablets (Pb ªaæ IªªºÆ Øa
ªæÆÆ ø IçEŒ ÆØ ¼ı F çæ , 6)? Even admitting this, how could the
exchange have worked when they were speaking different languages (7)?96 The
explicit denial of the existence of a written message can be interpreted in two ways:
in a straightforward way, as implying that the story of the letter did not yet exist (it
is then possible to assume that this gave Euripides the idea of the forged letter); or
also, as Gorgias’ playing with a (recent) version of the story, whose knowledge he
assumes in the public, in which case Euripides’ play would have preceded Gorgias’
Palamedes. A choice between these possibilities, on the basis of the evidence we
have, is not possible.

93 94
Jenkins 2006: 23–4. Pace Detienne 1989.
95
The exact relationship between the play and the speech is uncertain. Scodel 1980: 90 is ‘convinced
that Euripides used the Palamedes of Gorgias as a source’ for the agon; but she also stresses (91) that
there was no letter in Gorgias, for if there had been a letter, the arguments brought forward at Gorg.
Pal. 6–8 and 22–4, would not make sense. Jenkins 2006: 28 n. 25 is less sure. Usener 1994: 65–9 stresses
the difference of Gorgias’ project. An excellent overall reading of the speech, in terms of the possibility
of the logos to communicate, and in particular, to communicate something which is not, in Morgan
2000: 119–22.
96
Good discussion in Jenkins 2006: 26–9 (although I disagree with the translation he proposes of
the list of inventions; see rather the excellent ones by Untersteiner 1967 and Buchheim 1989, who seem
to me to have caught the ambiguity of the words used here). Tablets and languages: see above 77–8.
86 Greek Beginnings

After the problem of how communication might have been set up, the state-
ment ‘Still, assume that what did not happen has happened’ (11) led to the
discussion of further points (associates and motive), and then of character (of
the accuser and of Palamedes himself). In this context, a list of inventions was
given: the hero substantiated his claim concerning character by stating that he had
been the greatest benefactor (ªÆ Pæª Å) of his audience, of all Greeks, and
of all mankind, those living and those to come:
 ªaæ i K
Å  e IŁæ
Ø 
æØ K I
æı ŒÆd ŒŒ Å K
IŒ ı, Ø 
ºØŒa æg ªØ  N
ºŒ Æ Æ,  ı  ªæÆ
f
çºÆŒÆ [ ] F ØŒÆı, ªæÆ    Å ZæªÆ,  æÆ  ŒÆd ÆŁa
ıƺºÆªH P
æı ØƺºÆª, IæØŁ   åæÅ ø çºÆŒÆ,
ıæ   ŒæÆ  ı
ŒÆd Æå ı Iªªºı,
   åºB ¼ºı
 ØÆ æØ ; (Gorg. Pal. 30)
‘For who else has made human life practicable from impracticable97 and organized
from disorganized, inventing military structures of the greatest advantage for aggres-
sive expeditions, and written laws, guardians of justice, and letters, a tool of memory,
and measures and weights, easy standards of commercial exchanges, and number as
guardian of riches, and beacons as very powerful and swift messengers, and draughts,
the harmless game of leisure?’
This passage concentrates in one long list most of the inventions that are
attributed to Palamedes in fifth-century tragedy; here too, while Palamedes,
both in introducing the list and in then capitalizing on it, saw the inventions as
indisputable progress, the words with which the inventions themselves are de-
scribed leave room for anxiety. Palamedes claims to have invented two instru-
ments central to the politike techne, divisions of the army (war) and written laws
(peace); but his military reforms are meant to allow pleonektemata, a term whose
meaning ranges from ‘gain, acquisition’ (in the best of cases) to ‘undue gain, act of
overreaching, arrogance, greed’; as for the writing of the law, it allows for
consistency, but not necessarily for justice;98 the use of the term phylax, ‘guardian’,
for both the activity of the laws (to guard justice) and the activity of the number
(to guard riches) points to an inherent imbalance (why guard riches, in an equal
system—Palamedes could have chosen among quite a few other possibilities to
illustrate the advantages of numbers). These are extremely slight hints, probably
meant to show that even when trying to point out the most positive aspects of
inventions, it is impossible to avoid that the language one is using may convey
some blame. Interestingly, however, in a speech centred on the issue of the ability
of the logos to communicate, there is no mention of communication through
writing among the inventions. Writing appears, but as an instrument having as its
main function the preservation of memories, while the conveyance of long-
distance information is limited to the invention of beacons, a traditional element
of the story. As for the ability of the logos to communicate, the speech itself and
the outcome of the story show that Gorgias took a very dim view of it.99

97

æØ expresses at the same time the notion of a path, and that of finding one’s way through it;
with the help of resources (hence the meaning ‘resourceful’ for a person who is porimos) one can pass.
98
The Euripidean Hecuba could have explained this to him (Eur. Hec. 864–7); see discussion below,
210–11.
99
Morgan 2000: 121. The final appeal of Palamedes (35) reminds one closely of the wish uttered by
all of the main characters of Euripides’ Hippolytus, that the truth of deeds could be made clear through
 Writing and Letter Writing: Representations 87

Let us now consider the inventions of Palamedes as a whole. Remarkably, those

that systematically turn bad in his story by showing an extraordinary liability to
distortions are linked to communication (in the case of Gorgias’ Palamedes, the
failure is radically extended to all communication).100 Such is the case with letter
writing; such is also the case with the use of luminous signals as a code to transmit
messages through the night, whose invention is also in some versions attributed to
Palamedes. This second invention, meant to benefit humanity, was used by the
Greeks to destroy Troy: a fire lit by Sinon gave to the Greeks hiding in Tenedus the
signal that the Trojan horse was inside the citadel ([Apollod.] Epit. 5. 19). The
invention will be even more forcefully distorted by Palamedes’ father, Nauplius:
already the Nostoi recounted how, to avenge the death of his son, Nauplius had
corrupted the wives of the heroes during their absence at Troy, and then, at the
end of the war, had lit fires that signified death for the Greeks coming back from
Troy, by bringing them to shipwreck on the rocks of Euboea.
These same luminous beacons had already been productively used in Attic
drama to explore the ambiguities of meaning through the fire signals at the
opening of Aeschylus’ Agamemnon, set up by a woman (Aesch. Ag. 312–13),
but a woman defined as Iæ ıº, ‘of male counsel’ (Ag. 11). These fire signals
represent an Aeschylean innovation: there is no trace of a beacon chain in earlier
versions of the story. Clytaemestra’s two speeches (281–316 and 320–50) make it
clear how the transmission of the message on the one hand, and its meaning (i.e.
its interpretation) on the other, are two different, separate things.101 Signs and
meanings do not necessarily hang together. But the beacon scene, in this context
an innovation, might have triggered in the spectators of the play a recollection of
the story according to which Nauplius had, through beacons, provoked the
destruction of the Greek fleet. Slightly later in the Agamemnon, a messenger will
describe the destruction of the Greek fleet during a hurricane, stating that ‘fire and
water, two elements resolutely opposed before, conjured together’ for the destruc-
tion of the Greeks (Ag. 650–1: ı Æ ªæ, Z  åŁØ Ø e
æ,
Fæ ŒÆd
ŁºÆ Æ). This is a statement difficult to understand, as no lightning or fire is
explicitly otherwise mentioned as participating in the destruction of the fleet. But
the presence of fire here might allude to Nauplius’ own fires on Mount Caphareus
in Euboea: Clytaemestra’s beacons would thus retrospectively confuse themselves
with those of Nauplius.102

words. On the ‘triad’ weights, measures, and number see Soph. fr. 432 Radt, with the important
contribution of Heinimann 1975, who, however, seems excessively optimistic about the connotations of
this triad. In this passage, I would not be so certain.
100
Of course other inventions too can turn sour: Scodel 1980: 129 points out that notwithstanding
the hero’s discovery of law, he ends being worsted in a debate. The Euripidean questioning must have
been very deep indeed in the Trojan plays of 415 (see Scodel 1980, passim, but esp. 105–21).
101
Brilliant discussion in Goldhill 1984: 49–51 (much of what follows in Goldhill is relevant to the
broader issue of the perception of oral speech and debate versus written communication).
102
Nauplius’ children did intervene in some versions of the story of the Atridae: in Eur. Or. 432
Orestes in talking to Menelaus acknowledges that Oeax hates him, because of the role of Agamemnon
in the death of his father; Hygin. Fab. 117, 1 and Dictys 6. 2 relate that Oeax informed Clytaemestra of
Agamemnon’s return with Cassandra; Paus. 1. 22. 6 describes a painting on which the children of
Nauplius came to the help of Aegisthus.
88 Greek Beginnings

Thus a discourse on signals and the meaning of signs is certainly an early

component of the myth of Palamedes, going back to the Nostoi and to Nauplius’
revenge; this discourse was pursued in drama and elsewhere, also independently of
the figure of Palamedes. The connection between (alphabetic) writing and long-
distance communication is, however, secondary; the Greek (or Athenian?) defiance
towards letter writing and its uses finds a reflection in the fifth- and fourth-
century treatments of the myth of Palamedes.103

3.3.2. The Queen and the Tyrant

Two texts looking at the invention of writing specifically as a means for long-
distance communication help to further define the connotations that the Greeks
associated with letter writing. According to a fragment of the historian Hellanicus
preserved in two different, late sources, the oriental queen Atossa received an
education similar to that of a man; when she succeeded to her father’s throne, she
introduced some reforms, among which were the custom of wearing a tiara and
the ample oriental trousers called anaxyrides, the custom of employing in one’s
service eunuchs, and that of administering justice by way of letters (ŒÆd Øa
غø a I
Œæ Ø
ØE ŁÆØ, FGrH 4 F 178a = Anon. De Mulier. 7).104 The
introduction of letters to administer justice is mentioned in the context of a series
of innovations that, from the Greek point of view, are all marked negatively, as
inversions of the normal order: a woman should not be brought up as a man, and
should not take up kingship: giving power to a woman is dangerous, and queens
are often rather irresponsible characters. The wearing of a tiara and of oriental
trousers (the latter also worn by the Amazons, according to Greek mythology), as
well as the decision to use eunuchs are also usually viewed as negative things. It
thus seems reasonable to attribute a markedly oriental—and negative—character
also to the last of Atossa’s innovations, the institution of judgements through
letters instead of the oral, ‘democratic’ process (known for instance from Athenian
practice). Women in Greek tradition are linked to letter writing in a number of
intriguing ways (Gorgo at Sparta, for instance, or Sappho, as portrayed in the

103
Also in later tradition letters will remain linked to Palamedes: in the speech Odysseus, or Against
Palamedes for Treason attributed to Alcidamas (probably a product of the second sophistic: O’Sullivan
2008; but see Jenkins 2006: 30–6, who considers it an answer to Gorgias’ Palamedes), a message
engraved on an arrow that misses its aim is the main (but paradoxically absent) item of accusation;
according to Ampelius, a letter of Palamedes was preserved, together with some other rather extraor-
dinary documents, in a temple in Sikyon (Amp. Lib. Mem. 8. 5); and Tzetzes (Alleg. Il., Prol. 402–5) had
Palamedes carry to Troy, in the context of an embassy aiming at persuading the Trojans to return
Helen peacefully, a letter by Clytaemestra exhorting her sister to come back.
104
Cf. Hellanicus FGrH 4 F 178b = Tatian. —æe  ‚ººÅÆ, 1: ŒÆd K
Ø ºa ı  Ø [yæ] &
—æ H
  &ªÅ ÆÅ ªı , ŒÆŁ çÅ Ø  EººØŒ· @  Æ b ZÆ ÆP fi B [q]). This Atossa is not
the wife of Darius I and mother of Xerxes, but an Assyrian queen, assimilated at times (possibly first by
Ctesias?) to Semiramis. The Greek chronographical tradition locates her reign in the eighteenth place
in the list of Assyrian kings compiled by Castor (FGrH 250 F 1). This would correspond to the years
1497–1476; cf. Jacoby’s commentary ad FGrH 4 F 178. Justin 1. 2. 1 ff. preserves a related account: his
Semiramis has many traits in common with Atossa (trousers, tiara), but the innovation of giving royal
answers through intermediaries is attributed to her son Ninyas (1. 2. 12). On these traditions see
Porciani 1997: 157–60; Gera 1997: 141–50.
 Writing and Letter Writing: Representations 89

homonymous play by the comic poet Antiphanes); but women are never listed
among the inventors that benefited humanity. This is a unique case: only an
oriental woman could have been an inventor, and from such an odd inventor one
should not expect anything straightforward. Where Hellanicus came across this
information is not known; a comparison with the more general tendencies of his
work has led Kleingünther to suspect that this information originates with him, or
at any rate that it does not depend on an earlier historiographical tradition.105
This same invention—the introduction of letters to administer justice, and
more generally for administrative purposes—is attributed by Herodotus (1. 99–
100) to Deioces the Mede, in a context that accompanies the institution of a
centralized and absolutist power. Deioces, after first having gained a reputation for
justice in publicly solving disputes, refuses to keep putting his skills to the service
of the community unless he is given sovereign power; he then fortifies his
residence, and orders that nobody may have access to the king, and that all
communications must be through intermediaries, that is, messengers (Ø
Iªªºø); similarly, concerning the administration of justice, Deioces orders
that all suits be written down and transmitted to him for consideration; his
decisions are similarly sent out in writing. Herodotus explains this strategy in
terms of the need for Deioces to create a distance between the ruler and the other
Medes of his generation (  æçØ): the new ruler needs this distance, otherwise
his contemporaries, knowing that they come from no lesser families than Deioces
and that they are not his inferiors, might resent him and possibly even attempt to
overthrow him. By hiding from their view, Deioces can present himself as
different (% æE  çØ ŒØ r ÆØ c ›æH Ø). Paradoxically the letter, an instru-
ment meant to allow and facilitate communication notwithstanding distance, is
also the instrument that allows communication while imposing distance—this is
the message of these stories, not accidentally (dis)located in the world of the Near
East.106 Deioces’ request that communication between himself and his subjects
take place only through intermediaries and writing has also another important
consequence: because each person writes individually to the king, and receives an
individual answer, the ‘horizontal’ communication between the members of the
community, the collective dialogue on judicial matters, is severed.107

3 . 4 . A G R EE K N A RR A T I V E O F T H E B EG I N NI NG S O F A G E N R E

Let us now focus on how the Greeks represented to themselves the origin of the
formulae they used in their letters. Ancient perceptions and representations of the
development of letter writing are a crucial part of the analysis, since it is through

105
Women and letters: see below, ch. 5. Hellanicus’ sources: Kleingünther 1933: 125–31.
106
Porciani 1997: 159–60. Steiner 1994: 8 (and 130–2 specifically on Deioces) considers this to be
true of all forms of writing. Bertrand 1999: 86–8 compares with Plato, Statesman 295a–e and Laws 859a
(writing as the sign of the absence of the legislator/king). See also below, 126 n. 76, on the role played by
direct (but individual, not public) contact in the story of Maeandrius of Samos.
107
Bertrand 1999: 87–8.
90 Greek Beginnings

them that we are able to perceive the ancient discourse, as well as the ideological
take, on letter writing.
The diversity in the form of the earliest letters, when compared to the stand-
ardized form later taken by epistolary communication, had not eluded the an-
cients, and, as one would expect, there were questions on how this form had
originated and on its relationship with oral messages. An ancient commentator on
the Plutus of Aristophanes mentions a monograph on the greeting åÆæØ by a
certain Dionysius; however, the treatise is lost.108 The best testimony of the
ancient discussions on the form and the evolution of Greek epistolary writing is
a small treatise by Lucian, A Slip of the Tongue in Greeting. Its value as ‘hard
evidence’ is limited, because Lucian could not make use of archaic private or
public letters; he was writing in a period (the second century ad) when the
evolution we are interested in had long taken place; he had to rely on letters
and forms of greetings transmitted through literary tradition, and shaped by it.
More importantly, Lucian himself is not a neutral observer interested in the how
and when and why. Nonetheless, his discussion opens a fascinating vista on what a
cultivated Greek writer of the second sophistic thought of a genre, letter writing in
its various forms, that by then had imposed itself as central (collections of letters,
authentic or forged, will know an extraordinary growth in importance, from the
Hellenistic period onwards).109
Lucian’s A Slip of the Tongue in Greeting opens with the narrator recounting an
error he has just committed in a greeting: he has used for a morning visit a type of
greeting, ªØÆØ (literally: ‘be in good health’), that is appropriate for the
evening. In order to console himself for this error, he decides to write a small
treatise on the subject, to prove that his slip is not inexcusable. The narrator begins
by sketching the three main types of greeting: åÆæØ, s
æ Ø, ªØÆØ. Of
these, åÆæØ (‘Rejoice’, or ‘Joy’) appears to be the most ancient: and the narrator
proceeds to give instances of the use of åÆæØ in both oral and written greetings.
Examples taken from oral greetings in tragedy and the epos show that the term
could be used at any moment of the day, whenever two persons met, and even at
the moment of parting company (Luc. Laps. 2).110 After this statement about the
temporal indeterminacy of the greeting with åÆæØ, the narrator goes on to stress
the semantic indeterminacy of the term: ‘Nor was it [the address with åÆæØ]
necessarily significative of friendliness; it could express hatred and the determin-
ation to see no more of another. To wish much joy to someone indicates that one

108
Schol. Ar. Plut. 322 c:
æd F K fi B ıÅŁÆfi åÆæØ F  K ÆE K
Ø ºÆE, ªªæÆ
ÆØ
˜Øı ø fi  غ
æd ÆP F, ‘Concerning the greeting used in common intercourse and in the
letters, Dionysius has written a monograph on it’. Who this Dionysius may have been is unknown:
Chantry 1994 ad loc. suggests Dionysius of Halicarnassus or Dionysius Thrax. See further Gerhard
1904: 40–8; Sykutris 1931: 189–92.
109
For the fortune of pseudepigraphical epistolary collections see Rosenmeyer 2001: 193–223; for
the manipulations of the letter form in the period of the second sophistic, Koenig 2007; Hodkinson
2007a and 2007b.
110
The double use, as a greeting at the moment of meeting, but also at the parting, explains (and
gives more pregnancy to) the use of åÆæØ in funerary epigrams; see also the åÆEæ/åÆæ  at the end
of most Homeric hymns, with Calame 2005. Wachter 1998 sees the origin of the greeting with åÆæØ
in the request to someone else (god or man) to accept a gift; see Calame 2005: 26 and n. 22 for the
notion of reciprocal pleasure entailed by the use of åÆæØ.
 Writing and Letter Writing: Representations 91

ceases to care about them.’111 Of course this serves the cause of the narrator, and
one might be tempted to take this as an over-interpretation on the part of Lucian.
But when construed with KA and an accusative (of a person or a thing), åÆæØ
does indeed mean ‘to dismiss from one’s mind’, ‘to renounce’; this same sense is
found also in constructions with verba dicendi (the overall effect is ironical, as an
instance from Euripides’ Hippolytus shows, c ˚
æØ
ºº Kªg åÆæØ ºªø,
‘As for Cypris, I send her many greetings’, 113). Moreover, the composed forms
K
ØåÆæø and ŒÆ ÆåÆæø are attested with the meaning of ‘to rejoice over’ (usually
a disaster that has befallen an enemy), and ‘to rejoice at the expense of someone
else’.112
After thus establishing that the commonest greeting, åÆæØ, was indeterminate
in terms of when it was used, and could have very negative connotations, the
narrator proceeds to examine the first official uses of the term. The first person to
have made use of åÆæØ as a greeting is found to have been Philippides, the
runner who, according to Lucian, announced to the Athenian archons the news of
the victory at Marathon:
—æH  ’ ÆP e غØ

Å › &ææ Æ ºª ÆØ I
e ÆæÆŁH Iªªººø c
΁ N
E
æe f ¼æå Æ ŒÆŁÅı ŒÆd
çæ ØŒ Æ 
bæ F ºı B
åÅ, 'Ææ , ØŒH, ŒÆd F  N
g ıÆ
ŁÆE fi B IªªºÆ
fi ŒÆd fiH åÆæØ
ıŒ
F ÆØ. (Lucian, Laps. 3)
Philippides the dispatch-runner is said to have been the first, when bringing the news
of the victory of Marathon, to have said to the archons seated and tense about the issue
of the battle, ‘Rejoice, we win’. And having said this he died with his message,
breathing his last in the word ‘joy’.
The other instance of a greeting with åÆæØ offered by the narrator is again an
official one, but this time it is in a written, epistolary form: Cleon was allegedly the
first to compose a letter beginning with åÆæØ, when he announced to the
Athenians the good news of the victory at Sphacteria; after him, the narrator
hastens to add, Nicias writing from Sicily went back to the old habit of beginning
in medias res.113
The rest of the small treatise (nineteen paragraphs) does not mention the
greeting åÆæØ any more: it is exclusively dedicated to s
æ Ø and ªØÆØ.
The proportion shows unequivocally how biased the discussion is: because he had
erroneously greeted his patron with ªØÆØ instead of åÆæØ, the narrator, after
having established in what conditions åÆæØ, the most common epistolary
greeting, was used, and after having made clear both explicitly and implicitly,
through his choice of examples, that the salutation åÆæØ is not always really

111
Lucian, Laps. 2: ŒÆd P   çغçæ Å ÆP E q F  º, Iººa ŒÆd I
åŁÆ ŒÆd F
ÅŒ Ø åæ  ŁÆØ Iºº ºØ. e ªF ÆŒæa åÆæØ çæ ÆØ e ÅŒ Ø çæ ØE źE.
112
For the meanings of åÆæø, K
ØåÆæø and ŒÆ ÆåÆæø see LSJ, s.v., and Chantraine 1999: 1240–1.
Two Herodotean instances of situations in which ŒÆ ÆåÆæø is used in a very ambiguous, or
downright negative way (in the sense of rejoicing over the other’s misfortunes) are discussed below,
119–21.
113
Luc. Laps. 3: K K
Ø ºB b IæåB fi ˚ºø › ŁÅÆE Åƪøªe I
e çÆŒ ÅæÆ
æH 
åÆæØ
æhŁÅŒ PƪªºØÇ  c ŒÅ c KŒEŁ ŒÆd c H 
Ææ ØÆ H –ºø Ø. ŒÆd ‹ø ª
 ’ KŒE › ˝ØŒÆ I
e ØŒºÆ K
Ø ººø K fiH IæåÆø fi H K
Ø ºH ØØ I
’ ÆP H
Iæ H
æƪ ø.
92 Greek Beginnings

auspicious (note the extraordinary insistence on the coincidence between saluta-

tion and death in the story of Philippides), concentrates on positive instances of
the use of the other greetings.114
Keeping in mind this evident bias, we may try to evaluate the information
Lucian gives about the use of åÆæØ. The first important element is that Lucian
puts together oral and written forms of greeting. Of course he is relying for all of
his examples on the literary (written) tradition, but he begins with instances of
oral greeting and communication, and then moves towards written communi-
cation. He thus sees oral and written communication as a continuum. The one
formal difference is due to the different enunciative situation: oral greetings (as
presented in literary texts) are always delivered as imperatives (åÆEæ, åÆæ ),
because the person is present and facing the addressee; written greetings are
always in the infinitive (åÆæØ, with an implied ºªØ), because the greeting is
mediated.115
Second, it is mildly surprising, to say the least, to find that the first instance of
the use of this type of greeting, even with the restriction to an official context,
should be dated to 490 bc: because the previous references to the epos and to
tragedy had shown, if there had been any need of it, that the term was, at least in
the literary tradition, well attested as a normal form of greeting. 'ÆæØ is also
well attested as a form of greeting in sepulchral inscriptions, starting at least from
the mid-sixth century bc, in the imperative, as an apostrophe either to the passers-
by, with åÆæ , or to the deceased, with åÆEæ.116 Interestingly, however, the term
is not found in the earliest documentary letters, nor in the earliest literary letters,
those preserved in Herodotus and Thucydides. So, after all, there may be some-
thing in Lucian’s relatively late dating for this type of greeting.
Third, the story of the hemerodromos who announces 'Ææ , ØŒH, and
having said as much, expires, is a brilliant opportunity for a narrator whose
narrative takes as its point of departure the fact that he has erroneously said

114
Possibly taking some liberties along the way. Plato uses s
æ Ø, as Lucian says (Laps. 4); if
Plato Letter 3. 315 a 6 opens with åÆæØ, this is followed by a discussion in which preference is given to
s
æ Ø (as in Plat. Charm. 164 d–e). The preserved letters of Epicurus, however, have åÆæØ (see
To Herodotus, 35; To Pythocles, 83; To Menoeceus, 121), and not ªØÆØ, as Luc. Laps. 6 claims. Diog.
Laert. 3. 61 concurs with Lucian; but his reference to Epicurus and Cleon in the catalogue of Plato’s
works, ¯
Ø ºÆd æØ ŒÆŒÆ, MŁØŒÆ—K Æx  ªæÆç s
æ Ø, ¯
Œıæ b s تØ, ˚ºø
åÆæØ, is a clear give-away of the fact that this comes from the same source. The same applies to the
Suda  3664, ¯s
æ Ø: o ø & K
تæÆçc H K
Ø ºH —º ø. ªæÆł b K
Ø ºa ت´. N d b
F MŁØŒF Yı. s تØ: o ø ¯
Œıæ K
ªæÆç: åÆæØ b K
ªæÆç ˚ºø. Suda å 156
('ÆØæ ø ºÅæH: KæıŁæØ ø çºıÆæH. ŒÆd åÆæØ, NÇØ, ÇØ, also in Et. Gen., Et. M. 808.2, 3),
putting together blushing at talking nonsense, and åÆæØ with the meaning of ‘moaning’, ‘groaning’,
also corresponds closely to the opening of Lucian’s treatise, in which the narrator begins by blushing at
his blunder. Koskenniemi 1956: 163 believes, however, in the correctness of Lucian and Diogenes’
information; see Inwood 2007: 143 for the view that the preserved letters of Epicurus have been
‘normalized’ through the substitution of åÆæØ for the original ªØÆØ.
115
Only with the first century ad do apostrophic epistolary opening formulae with the optative
åÆæØ appear; the imperative apostrophic epistolary greeting åÆEæ is attested from the second century
ad. see Gerhard 1905: 38–52 (and Gerhard 1905: 27–38 with Lallot 1997: i, 228–32 for a fascinating
close discussion of Apollonius Dyscolus’ take on the letter prescript, in On Syntax 3. 63–77, 329–41
Uhlig); Exler 1923: 53–4 and 73; Klauck 2006: 18–19; and compare with the early letters discussed
above, 35–47.
116
See above, 57 n. 128.
 Writing and Letter Writing: Representations 93

ªØÆØ (‘be in good health’), instead of pronouncing the usual morning greeting
åÆæØ. In opening his survey, the narrator had indeed stated—tongue-in-cheek,
of course—that he had initially thought his task would be a difficult one, but that
to his surprise he had found a lot of material.117 But exactly how good an
opportunity is such a story—how ancient is it? Can we trust it?
The hemerodromos Pheidippides (Philippides?) makes his first appearance in
Herodotus, in connection with the battle of Marathon.118 In 6. 105–6, Herodotus
narrates that when the Athenian strategoi realized that the Persians had landed in
Marathon, they sent Pheidippides to Sparta to ask for help. He covered the
distance (c.240 km) on foot in two days, meeting Pan along the way; he arrived
in Sparta on the day following his departure from Athens, and delivered to the
Spartan magistrates a message comprising an address in the vocative, followed
directly by the body of the message—but no greetings.119 He then did not die, but
came back to Athens, to report to the Athenians his encounter with Pan, and the
Spartans’ refusal to move before the full moon. Neither Herodotus nor any other
ancient source ever mentions his death; Lucian’s version is unique. And it refers to
a very different ‘run’: in Lucian, Philippides covers the distance between Marathon
and Athens—this is the origin of the length of our modern marathon-run.
There were, however, at least two other versions concerning the run from
Marathon to Athens; both are preserved by Plutarch. In his De glor. Ath. 347c
(= Heracleides fr. 156 Wehrli), Plutarch states that according to Heracleides of
Pontus, a near contemporary of Plato, the news of Marathon was brought to
Athens by Thersippus of Eroiades; ‘but most writers say that it was Eucles who ran
in his armour, hot from the battle, and bursting in through the doors of the most
important magistrates, said only “rejoice” and “we are happy”, and then immedi-
ately expired.’120 Lucian’s version appears to conflate the Herodotean story and
that attributed by Plutarch to ‘most writers’, possibly in order to make fun of the
communis opinio, as Frost has suggested. Unluckily Plutarch does not give any
details on what Heracleides said happened to Thersippus (another horsey name,
interestingly) after bringing the news, nor is it self-evident that we should attribute

117
The tongue-in-cheek approach is betrayed even at the level of terminology: Lucian says,
æå  b s B ªæÆçB
ı I
æø fi K  ŁÆØ fiþÅ fiH
æº Æ Ø (Laps. 2), ‘I began to
write expecting to find the problem impossible to solve’— K ıªåø means ‘to meet’ (here: with a
problem), but an  ıØ is one of the commonest types of letter, the petition (described, after
Bikerman 1930, in Koskenniemi 1956: 159–60).
118
Pheidippides is the reading of the better manuscript family; all other Herodotean manuscripts,
and all the later tradition, have Philippides. Badian 1979 defends (convincingly, to my mind) Pheidip-
pides; Frost 1979 takes the other view.
119
Hdt. 6. 106. 1–2: I
،  b K
d f ¼æå Æ ºª ‘t ¸ÆŒÆØ ØØ, ŁÅÆEØ ø
 ÆØ ç Ø ÅŁB ÆØ [ . . . ]’. The story is recounted also by Paus. 1. 28. 4 and 8. 54. 6; Poll. 3. 148. 16;
schol. Ael. Arist. p. 51 and 251 Frommel; Nep. Milt. 4.3; Plin. NH 7. 20. 84 (whence Solin. Coll. 1. 98
p. 25. 10 Mommsen). Cf. Frost 1979; Badian 1979.
120
Frost 1979 discusses the relationship between these and other stories of extraordinary runs with
death at their end. The text of Plutarch has been subjected to many corrections, one of which concerns
the announcement: where all codices have  F    N
E ‘åÆæ ’ ŒÆd ‘åÆæ’ (which is also
the text of Nachstädt’s Teubner edition), the Budé editors Thiolier and Frazier, following Cobet and
because of Lucian Laps. 3, write ‘åÆæ ’ ŒÆd ‘ØŒH’. But, as Gallo and Mocci point out (Plutarco, La
gloria di Atene (Naples, 1992) ), the text is meaningful as it is; the notion of reciprocity, which is
structurally part of the greeting åÆEæ, also speaks for the transmitted text.
94 Greek Beginnings

to Thersippus the same message as that of Eucles. This means that we cannot
know whether the greeting with åÆæ  had entered the tradition already by the
time of Heracleides; at any rate, it is unlikely to have been much earlier.121
What of the other instance offered by the narrator of a greeting with åÆæØ, the
one concerning Cleon? Can we use it to pinpoint a transformation, in the
Athenian context, of the epistolary opening form of address? This is again a
‘first’, but this time, an epistolary first use of åÆæØ, supposedly taking place in
424 bc, after the Athenian capture of the Spartans blockaded on the island of
Sphacteria. Two passages from comic plays of this period mention a special
greeting; their late commentators interpret these greetings in connection with
the story also told by Lucian.
In the antepirrhema that closes the parabasis of Aristophanes’ Clouds, the
coryphaeus mentions an encounter between the Clouds and the Moon:
&å’ &E Fæ’ IçæA ŁÆØ
Ææ Œı ŁÆ,
& º Å ı ıåF ’ &E K
 غ çæ ÆØ

æH Æ b åÆæØ ŁÅÆØ Ø ŒÆd E ıåØ·
r Æ ŁıÆØ çÆ Œ· Øa ªaæ

ŁÆØ,
TçºF A –
Æ Æ P º ªØ Iºº KçÆH. (Ar. Nub. 607–11)
When we were ready to set out on our journey here, the Moon happened to run into us
and instructed us first of all to convey her greetings to the Athenians and their allies;
and then she said she was annoyed, because she had been shamefully treated despite
having helped you all not with mere words but with plain action.
(Trans. Henderson modified)
As pointed out by all commentators, the inclusion in the greeting of the allies
points to official language.122 But it is very difficult to know whether Aristophanes
is here simply alluding to a message sent through a messenger, and transmitted
orally (as for instance in Ar. Lys. 1074–5, where the coryphaeus addresses the
Spartan representatives with ¼æ ¸Œø,
æH Æ  Ø åÆæ , | r Y
ÆŁ
&E,
H å  lŒ , ‘Men of Lacedaemon, first greetings to you, then tell us,
for what reason you have come’), or whether the public could hear in the greeting
with åÆæØ, preceded by K
 غ, an allusion to the formula which will become
standard in epistolary writing. As we have seen, K
Ø ººø and K
Ø º are not
yet specific enough.123 Even more difficult is the question of what exactly is the
punchline of this passage: is Aristophanes having fun out of using substandard

121
The anecdote may come from Heracleides’ Dialogues, a work in which Heracleides used
historical figures to make philosophical points; Wehrli 1967 classifies the fragment among the
‘Historisches unbestimmter Herkunft’, adding—his only comment—that ‘it must be an accident’ (‘es
muß Zufall sein’) if in fr. 149 the first beneficiary of Solon’s law that the state should take care of the
wounded in war is called Thersippos; fr. 149 is also transmitted in Plutarch, Solon 31.2.
122
Van Leeuwen 1898 ad v. 609 refers to Ar. Av. 879 and to the parodic Pherecr. fr. 39 K.–A., from
the Graes (Old Women): ŁÅÆÆØ ÆP ÆE  ŒÆd ÆE ıåØ (‘To the Athenian women and their
allies’). Kassel and Austin add a number of further parallels, both literary and epigraphical, among
which Thuc. 4. 117. 3. Interestingly, in fr. 38 K.–A. of Pherecrates’ Graes it is question of sending a dove
as a messenger (I

ł Iªªºº Æ e
æØ æ ): frr. 38 and 39 K.–A. might have been
connected; with a dove as a messenger, the message would have had to consist of a sign or signs,
although not necessarily written ones.
123
Discussion above, 17–18, 38, 44–6, and 56; below, 118, 199–200, 241.
 Writing and Letter Writing: Representations 95

official language? Is he punning on the meanings of åÆæø? As we have seen, the

term covers quite a range of meanings, and in this specific passage it does not
necessarily imply an innocent greeting.124 At any rate, verbs of ‘saying’ appear
twice (çæ ÆØ and çÆ Œ) in the transmission of the Moon’s message; this means
that even though it is possible to see an allusion to official epistolary communi-
cation in this passage, such communication is still being moulded along the
parameters of oral messages.
The scholia to this passage offer further information: they report a story
according to which the greeting alludes to a letter sent by Cleon to the Athenians
after the victory at Sphacteria (the first Clouds was produced in 423, just after the
victory).125 Although Cleon is not mentioned anywhere in the antepirrhema of
Clouds, he is the central character of the epirrhema; and there is scholarly
consensus on the fact that the epirrhema and antepirrhema of Clouds are closely
related. For our purposes, particularly interesting is the reference in the epirrhema
to the Moon going out of her path at the moment of Cleon’s election to the
strategia (Nub. 584), for it forms a pendant to the Moon’s message in the
antepirrhema; the audience might have perceived the message being now sent as
related to the Moon’s former action and consequently to Cleon.126 A real letter by
Cleon might thus—but just ‘might’—be behind the message of the Aristophanic
Moon.
Eupolis fr. 331 K.–A. may also allude to this same letter by Cleon:

æH  ªaæ &A, t ˚ºø, | åÆæØ
æ E
Æ
ººa ºı
H c
ºØ127
For you were the first, Cleon, to address us with ‘Rejoice’ (or ‘Forget it!’), you who
caused so much grief to the city.
On this fragment the ancient commentators provide information which, while
potentially extremely interesting, is not univocal, besides being very difficult to
evaluate. We have no ancient indication of the play from which the fragment
came. Hypotheses have of course been advanced. Bergk proposed the Chrysoun
genos (Golden Race), a play that could have been brought on stage in 424 bc, that
is, at the right moment for an allusion to Sphacteria (but a date of 426 bc has been
recently suggested for the play, which would lead to the exclusion of any links with

124
Most translations of Aristophanic plays tend to translate åÆæØ, here and elsewhere, simply
with ‘greetings’; yet LSJ lists most comic instances of åÆæØ (this one too) under III. 2c, ‘dismiss,
renounce’. It may have worked as a kind of standing joke, to be made explicit by the gestures and
intonation of the actor.
125
Schol. Ar. Nub. 609a:
æH Æ b åÆæØ· e ˚º çÆ Ø I
e B —ºı ŒÆd çÆŒ ÅæÆ E
ŁÅÆØ K
Ø ºº Æ ‘åÆæØ’
æŁEÆØ· ‹
æ ªª K åæ Ø.
126
Parts of the first Clouds were integrated in the reworking of the play. The epirrhema certainly
comes from the first version of Clouds (Cf. Dover 1968: lxxx–xcviii, and ad 584); the closeness of the
antepirrhema to the epirrhema in meter, number of verses, and character (Dover 1968, ad 607–26) is
such as to make it very likely that both belonged to the first version of Clouds. Note in particular the
nice contrast between the support accorded visibly (KçÆH) by the Moon (v. 611) and her eclipse at
the election of Cleon (v. 584).
127
Kaibel in 1895 suggested reading
æH  (adverbial) rather than
æH  in Eupolis fr. 331
K.–A. (‘for first you said to us åÆæØ . . . )’, comparing Ar. Nub. 608–9, where there is also an adverbial

æH Æ ; this would have solved the issue of Cleon’s being the ‘first’ greeting. He was followed by
Gerhard 1905: 48–9; Kassel–Austin maintain the transmitted
æH .
96 Greek Beginnings

the Pylos/Sphacteria affair).128 Storey suggests that the fragment might rather be
attributed to the Demoi.129 This play featured the return to Athens of four great
leaders, Solon, Aristides, Miltiades, and Pericles; Cleon might have been put
forward as one of the candidates to return to Athens in the course of an initial
scene of necromancy, or of raising of the dead.130 The difficulty with this is that
Demoi is usually thought to have been performed in 412 bc, although 417 bc and
410 bc have also been proposed, and while an allusion to Cleon in a play of this
period is possible, it is less likely than in a play performed in the period of Cleon’s
leadership. Cleon is explicitly made fun of only once more in all of Eupolis’
fragments, in fr. 316, from the Chrysoun genos; yet this may be an accident of
transmission, as jokes concerning him will have figured in other plays.131 Kassel
and Austin have left the fragment among those incertae fabulae; similarly, Olson
is non-committal as to which play the fragment came from, but suggests for it a
date to 424–422 bc, which would suit the context I have been sketching.132
The lexicographer Moeris, who transmits the fragment, describes Cleon’s
åÆæØ as the first one ever used in a letter:
åÆæØ K K
Ø ºB
fi
æH  ºª ÆØ ªæłÆØ ˚ºø ŁÅÆØ  a < e> ºÆE c
—º. Ł e ŒøØŒe K
Ø Œ
 Æ N
E· (the text of the fragment follows)
Cleon is said to have been the first to write ‘greetings’ in a letter to the Athenians after
taking Pylos. For this reason the comedian, making fun of him, said: . . . 133
A series of authors concur with this statement: besides Lucian (the oldest
datable source for this story), other late sources affirm that Cleon was the first
to send an official letter to the Athenians, greeting them with åÆæØ. In particular,
a scholiast to the Plutus of Aristophanes preserves (or at least, believes he
preserves) the beginning of the letter in question.134 In it, Cleon addressed the
Athenians directly:

128 129
Storey 2003: 266–7. Storey 1995–6: 142–3.
130
Storey 2003: 112–14 dates the Demoi to 417 or 416 bc; Telò and Porciani 2002 to 410 bc. The
idea that Cleon might have been one of the characters from the Underworld is suggested in Storey
1995–6: 143; important general discussion of the play in Storey 2003: 111–74. On the issue of Cleon’s
letter, Gerhard 1905: 38–41 is worth consulting.
131
For references to Cleon in Eupolis see Storey 2003: 343–4 and 384–6.
132
Olson 2007: 191.
133
Moeris å 37 (213, 31) speaks simply of ‘a’ comic poet; the attribution to Eupolis results from the
emendation of a passage of the Suda, attributing the story to the fourth-century comic poet Eubulus
(Suda, ed. Adler, åÆæØ 162: F  K d
æ ƪ æı Ø I
ƺºÆ ø m ŒÆd
æ æåø· K Ł 
ŒÆ Iæåa H K
Ø ºH. . . . Ołb ÆE K
Ø ºÆE
æ ŁBÆ Ø Çı Ø· ±
ºH  o ø
Iºº ºØ K
Ø ººØ, x  @Æ Ø —ºıŒæ Ø  ºªØ.
æH  b ˚ºøÆ çÅ d ¯hıº ›
ŒøØŒe o ø K
Ø EºÆØ E ŁÅÆØ I
e çÆŒ ÅæÆ, Kç’ fiz ŒÆd 
æÅ ŁBÆØ· IªH, ‹ Ø ŒÆd
ƒ
ƺÆØd KåæH  ŒÆd
æ Ū æı o ø Iºº ºı, P   e
æH  K ıªå , ‰ &E,
Iººa ŒÆd Øƺı Ø I
Iºº ºø, I d F ªØÆØ ŒÆd KææH ŁÆØ, åÆæØ ØŒº  Iºº ºØ).
Fritzsche’s emendation has been accepted by Bergk, Commentationum de reliquiis comoediae Atticae
antiquae libri duo (Leipzig, 1838), 362, by Meineke, and by Kassel–Austin (see their apparatus ad loc.),
but not by Adler.
134
The scholiast is discussing a passage of the Plutus in which Chremylus addresses the chorus, vv.
322–3: åÆæØ b A K Ø tæ Å ÆØ | IæåÆE XÅ
æ ƪæØ ŒÆd Æ
æ , ‘To say,
“Greetings, my dear neighbours!” is old-fashioned and musty’); Chremylus then uses the more formal
(? cf. Plutus 1042) I
ÇÆØ.
 Writing and Letter Writing: Representations 97
˚ºø ŁÅÆø fi B ıºB
fi ŒÆd fiH  ø
fi åÆæØ. (Schol. Ar. Plut. 322c Chantry)
Cleon to the boule and the demos of the Athenians, greetings.
There are however ancient opponents of this story, even though what is
contested is never that Cleon sent a written message, but only its exceptional
status. Thus, a scholion to the Clouds states: ‘Notwithstanding what some people
say, Cleon was not the first one to write letters in this way.’135
Did Cleon indeed send a letter greeting the Athenians with åÆæØ? Was this
letter really the first one addressing the Athenian assembly with a åÆæØ, or does
the ‘first’ refer rather to the fact that Cleon would have been the first to write to the
Athenians with news of the victory? Should we consider this entire story a later
construct, based on a text, that of Eupolis, whose specific context was lost, and on
its connection—not necessarily correct—with the already discussed passage of the
Clouds? The latter is Storey’s position: for him, the greeting with åÆæØ was
already common at the time of Cleon, and the main point of the Eupolis fragment
is to be found not in the use of åÆæØ in an official, public context, but in the
contrast, already present in tragedy, between joy and suffering, åÆæØ and
ºı
E.136 This may indeed have been the principal effect aimed at by the poet,
who may, moreover, have played on the other meaning of åÆæØ. Thus, even
though here it is difficult to reach certainty, the story is likely to have been made
up; at any rate, Lucian’s account is very much open to suspicion.137
However, the greeting with åÆæØ cannot have been frequent in official
communications. The numerous passages cited by Storey in support of his
claim that the official greeting with åÆæØ was common by the second half of
the fifth century do not offer good parallels.138 These passages may be divided into
two groups: those where the greeting is directed to the native land (so e.g. Eupolis
fr. 99. 35: t ªB
Æ æØÆ åÆEæ· b ªaæ . . . ); and those where the greeting is

135
Schol. Ar. Nub. 609bÆ: IæåÆE q Ł
æ  Ø K ÆE K
Ø ºÆE e åÆæØ. P ªæ, u Ø,
˚ºø
æH  o ø K
 غ ŁÅÆØ KŒ çÆŒ ÅæÆ. (Same in schol. Ar. Nub. 609b.) Similar
opposition is registered in the Suda å 164: 'ÆæØ: IæåÆE Ł e K
Ø ºÆE
æ ØŁÆØ e åÆæØ.
ƒ b ºª  ˚ºøÆ
æH  åæ Æ ŁÆØ  fi ø, ł ÆØ, and in Suda å 162.
136
Storey 1995–6: 142, comparing Aesch. fr. 266 Radt: ŒÆd   åÆæØ   ºı
E ŁÆØ æ ,
and Soph. Aj. 555.
137
There might have been at some point in the transmission of these stories some intertextual
‘horse-play’ around the name Pheidippides (an amplification of Pheidippos, ‘he who uses sparingly of,
and so who cares for, his horse’). Lucian tells a unique story of a greeting pronounced with the runner’s
last breath, a variant on a well known Herodotean story, of which another variant is known, whose hero
is a certain Thersippos (‘Confident in his horse’?). In Lucian’s account, this is followed by Cleon’s
greeting—alluded to possibly in Clouds, a piece presenting among its characters a Pheidippides. The
story of how Pheidippides was given his name is the object of a long monologue on the part of
Strepsiades; and the moon, which played a role in the Herodotean story, plays a role too in the
epirrhema and antepirrhema of Clouds. The name itself is rare (the two only instances in LGPN ii are
the hemerodromos of Herodotus and Lucian, and the Aristophanic character.) Note, however, a
Pheidippidas from Thera, and a Pheidippides from Eretria in Euboea: LGPN i s.v. There are also, of
course, in Attica and elsewhere a number of people called Philippides or Philippidas (‘Lover of horses’),
a much more common name, as well as forty-seven instances of people called Pheidippos.
138
Storey mentions Eupolis fr. 99. 35; Aeschylus, Agamemnon 538; Aristophanes, Pax 582. To these
one could add Aeschylus fr. 144 Radt, from the Mysoi; Aeschylus, Agamemnon 22 and 508; Euripides
fr. 558 and 696 Kannicht, from the Oineus and the Telephus resp.; Euripides, Orestes 356 ff.; and
Aristophanes, Acharnians 729 ff., to remain within the fifth century.
98 Greek Beginnings

addressed to a god or to someone who has just arrived: that is to say, a situation
completely different from that of a messenger who on arrival pronounces a
greeting. Thus, in Aristophanes, Pax 582, the chorus greets Peace, who has just
arrived (see also Pax 523); thus also in Aeschylus, Agamemnon 538, the greeting
(ŒBæı åÆØH åÆEæ H I
e æÆ F) is pronounced by the chorus (and not by
the messenger).139 None of them is even remotely close to the type of message that
Cleon sent, according to the scholiast on the Plutus. Neither the fifth-century
documentary letters, nor those cited by the ancient historians (including Nicias’
letter, sent in 414 bc, which, as Lucian says, ‘went back to the old habit’ of
beginning without an introduction—and in this context it is unimportant whether
it is a Thucydidean rewriting) present a standardized epistolary prescript. Cleon’s
greeting finds a comparison only in oral and informal apostrophes. Thus, part of
the joke, both in Eupolis and in the Clouds, may have had to do with the use of
åÆæØ in an official context (not necessarily in a letter).140 At the same time, it is
clear that there must have been some kind of debate over the right form of (public)
address: in the early fourth century Aristophanes could have Chremylus in the
Plutus dismiss the greeting with åÆæØ as ‘old-fashioned and musty’.141

3 . 5 . C ON C L U S I O N

The analysis of the traditions on the invention of writing has uncovered the
existence of a profound division between everyday written prose and letter
writing. The early narratives on the invention of writing had not focused on the
function of the new craft, which was presented as something imported from the
outside, and adopted more or less without difficulty. Slightly later traditions had
attributed the invention of writing to a Greek culture-hero of Panhellenic or local
status. These traditions had highlighted the ability of the written word to preserve
information and had emphasized its potential for creativity, two aspects that have
little to do with letter writing. Letters come to prominence and are problematized
in a specifically Athenian context during the second part of the fifth century,
through the development of the story of Palamedes first in drama and then among
the sophists. In these accounts, letter writing is viewed negatively, as something
that it is difficult to control. The accent is here on the difficulty of communication
in general, and then specifically on the difficulties of epistolary communication.
How far this perception of letter writing was shared outside Attica and outside the
questioning of the tragedians and the discussions of the sophists is difficult to tell.
And yet the two other accounts that include writing—epistolary writing, that is—
among a list of innovations, preserved in Herodotus and Hellanicus, also present a
strongly negative picture. According to these narratives, the purpose of letter

139
Cf. Fraenkel 1950, ad 251 ff., ad 538 and 539, and ad 22, with numerous further examples.
140
Note also Olson 2007: 211: ‘Cleon’s innovation may actually have been to include a common
colloquial expression in a formal state communication’, referring to Fritzsche.
141
Of course this is comic too; Aristophanes may be playing on tragedy’s elevated style (e.g. the
address of Aegeus in Eur. Med. 663–4:  ØÆ, åÆEæ: F ªaæ
æØ | ŒººØ Pd r 

æ çøE çºı.)
 Writing and Letter Writing: Representations 99

writing is to enforce a distance, to render face-to-face communication impossible;

not by accident is it invented by an Oriental queen or a Median tyrant.142 I would
thus suggest that in Greece letter writing remained a marginal activity for a
relatively long time, from its beginnings until the late fifth century. Not surpris-
ingly, communication at a distance would have been viewed negatively in a face-
to-face society; only in the fourth century did the habit of writing letters spread
throughout society, and only in that same century did letters acquire a literary
status.
Centuries later Lucian still preserves something of this uneasiness in his
narrative of the evolution of the greeting formulae, when he stresses that Cleon
was the first to use åÆæØ in an official document and notes that Nicias went back
to the ‘old habit’ of beginning dispatches in medias res—that is, writing messages
not couched in letter format. Lucian’s sources are lost to us; but a few fragments of
Old Comedy which imitate official communication would seem to confirm that, at
least in an Athenian milieu in the fifth century bc, certain types of greetings were
considered inappropriate for official communication. A detailed analysis of
writing and letter–writing in tragedy and comedy (below, chapter 5) will permit
further refinement of the analysis proposed here; meanwhile, in the next chapter,
we shall look further into the connection between oral and written messages, and
between letter–writing and tyrannical control.

142
Actually, in the story of Atossa as in that of Deioces, writing is probably imagined as pre-
existing; but it comes into prominence only once turned into writing-for-communication, i.e. into
letter writing. It is for this reason that I have discussed these narratives with those on the invention of
writing.
 4
When a Letter and Why?
Narrative Strategies in the Ancient Historians

Embedded letters are found in all sort of literary works: in tragedies and comedies,
in novels, and in historical works. Their function is sometimes to advance the
intrigue, sometimes to provide an insight into the thoughts of the characters,
sometimes to provide the internal and external reader (or spectator) with infor-
mation shared in a complicit way while leaving other characters in the dark.
Recent contributions have helped illuminate the role and uses of letters in drama
and in the novel.1 Some attention has also been paid to their functioning in
historiography: Rosenmeyer dedicates a chapter to a discussion of letters in
Herodotus and Thucydides; these same authors had already been central, from
different perspectives, to the studies of van den Hout, Longo, and Porciani.2 But if
one is interested in the historical development of the letter form and in its
changing ideological connotations, it becomes necessary to pursue the analysis
into the fourth century and beyond, for this is when letter writing becomes one of
the most important means of communication. Focusing on the use of letters in
historical narratives from Herodotus to Polybius offers the opportunity to survey
developments and changes—or, conversely, the constants—in the use of writing
for long-distance communication within one specific genre (historiography), and
to obtain a more rounded picture of the modalities according to which Greek
poleis chose to interact with other poleis, leagues, and kings, both in the fifth
century and in the changed circumstances of the early Hellenistic period.
Besides the decision to chart the use of letters in the works of the ancient
historians over a relatively long time-span (fifth to second century bc), I have
made two other choices that should allow a more nuanced understanding of the
use of written communication. First, I shall evaluate the use of letters for commu-
nicating against the use of oral messages for the same purposes; for messages
conveyed orally, by means of the voice of a messenger, and information conveyed
in writing, by means of a letter, are both instances of mediated, ‘second-hand’
communication.3 Bassi has pointed out that this mediated communication

1
Among the most important recent studies are the papers in Holzberg 1994, and Rosenmeyer 2001;
for epistolary novels, Hodkinson 2007a. For letters in tragedy, see below, ch. 5.
2
Rosenmeyer 2001: 45–60; van den Hout 1949; Longo 1978 and 1981; Porciani 1997.
3
Letters (or inscriptions, oracles, etc.) inserted in a narrative may from a formal point of view be
considered as ‘speeches’ (even non-vocalized elements such as thoughts or inscriptions, in themselves
102 Greek Beginnings

(or ‘scripted communication’, as she suggestively chooses to define it) occupies a

middle position in the dichotomy between orality and literacy: it is relayed
communication, and not face-to-face speech, but as such can take oral or written
form. It thus seems worth investigating whether differences linked to the medium
are noticeable.4 A written message is by definition inscribed on something mater-
ial (parchment, lead, even someone’s skin) and sent by a sender, A, to a distant
addressee, B, through a third person, the messenger; among the speeches
appearing in the sources, I shall, therefore, consider only those concerning a
situation in which a messenger transmits a message, oral or written, from a
character A to a distant addressee B—in short, only what can be strictly defined
as ‘mediated communication’.5
Second, I shall evaluate how ancient historians employed ‘mediated speech’
(whether oral or written, but with a special focus on letters) against the backdrop
of their use of other types of ‘written reported speech’, such as inscriptions or
documents: this approach makes it possible to gauge the respective importance of
their choices.
Because of the very nature of historical narrative, it is impossible to assume that
the decision to present the transmission of information in the oral code
(a messenger relating someone else’s words) or in the written code (a letter)
rests exclusively with the author, as would be the case with a novelist: such a
choice could be due to external, ‘objective’ factors (the historian, in his enquiries,
might actually have been told that a letter had been sent, or that a messenger had
been entrusted with the delivery of an oral message).
The following enquiry thus unfolds within a terrain marked by the complex
interface of historiographical choices and cultural practice. Putting the findings
into the larger context of the historian’s work (tracing internal connections in the
narrative, for instance) as well as contextualizing the findings within the larger
Greek contemporary cultural context may, however, help to assess the balance
between the two.

deprived of sound, would have acquired a vocalization in the performance: Bers 1996: 9). Lang 1984:
80–131 includes in her list of Herodotean speeches all voices other than that of the narrator, hence
letters too—the difficulty is in distinguishing between ‘speeches’ in oratio obliqua and ‘speeches’ related
by the narrator (Lang 1984: 90, 155 n. 6 and 131; de Bakker 2007: 6–7). The same difficulty in defining
the notion of ‘speech’ explains the discrepancy between the lists of speeches in Thucydides compiled by
Jebb 1880 and Blass 1887: 231 (41), Luschnat 1970: 1159–63 (more than 60), and West 1973: 3–15 (141
‘speeches’, divided into fifty-two speeches in oratio recta, eighty-six in oratio obliqua, and three
combining direct and indirect speech; he also includes the letters).
4
Bassi 1998: 9. The transmission of information through spontaneous conveyors, such as travellers,
who may accidentally learn of events that they then report to others, is not ‘mediated communication’,
since the initiative of the individual, who is responsible for what he chooses to report, is paramount. Cf.
Longo 1981: 13–30; Longo 1997: 657–60. On news, see Lewis 1996; Russell 1999. The spontaneous
diffusion of news in Herodotus takes varied forms: the ºª ºº concerning Solon, his wisdom, and
his voyages, which reaches Croesus without any specific intermediaries (1. 30. 2); the çØ concerning
the return of Peisistratus (1. 60. 5) or the birth of Cyrus (1. 122. 3); the Iªªº Å concerning the victory
of Marathon (7. 1. 1; 8. 99. 1); the ç Å of the victory of Plataea, spreading itself miraculously at the
moment of Mycale (9. 100). The words of the Pythia ‘arrive’ ( Ø Ø KºŁF Ø E Ø  Ø, 1. 56. 1),
exactly as the chresmos (1. 75. 2) and the manteion (2. 111. 3) do; letters too ‘arrive’ (3. 43). Finally,
news can be diffused spontaneously by ‘someone’, as in Hdt. 1. 43. 3 or 5. 33. 3.
5
Longo 1981: 27. The modalities of mediated communication, oral and written, are discussed in
Longo 1981: 43–58 and 59–63. See also Bassi 1998: 43 for a definition of ‘scripted communication’; this
corresponds to what de Bakker 2007: 49 terms ‘transported speech’.
 When a Letter and Why? 103

4. 1. M E D I A TE D L O N G - D I S T A NC E C OM M U N I C A T I O N
I N HE RO D O T U S

Looking at the interplay between oral and written messages in Herodotus has
implications for the understanding of both the Histories as a whole, and the
cultural world of fifth-century Greece. Three important features of Herodotus’
work that have been recently stressed are the use of prose for his narrative; the
consciousness of the peculiar status that being written confers on the work; and
the peculiar form taken by the proem (an epistolary proem?).6 However, there is
no agreement on the implications of the authorial use of expressions such as 
ºªø and  ªæçø; more generally, the debate on the level of intentionality of
Herodotus’ Schriftlichkeit is still very open. The analysis of the peculiar character-
istics of written and spoken messages in Herodotus may contribute to the ongoing
discussion on the shape of the Histories.7
The wider cultural implications will emerge from the comparison between the
Herodotean use, the use in the other historians, and the enunciation of the ‘real’
letters on lead or on ceramic fragments of the archaic and classical period. If, as
I shall suggest, in Herodotus’ work there is no formal difference in terminology or
enunciation between written and oral messages, then this has implications for
identifying the moment when an ‘epistolary genre’ became codified, and also for
the enunciative stance of Herodotus himself.
Let us then see how mediated communication works in Herodotus.8 I will not
discuss here the mediators themselves. Various types of messenger-characters
(¼ªªºØ, æ Ø, Œ æıŒ) play a role in the Histories, and one might have
expected that the form and type of individual messages could be related to the
specific type of messenger; but the Herodotean nomenclature for them is not
consistent and precise enough to allow a distinction on the basis of the mediator
between the different types of diplomatic activities.9 We shall thus look at the
messages only.

6
Choice of prose: Goldhill 2002: 5–6. Writtenness: de Jong 1999, esp. 220–3. On the inscriptional
character of the proems of Hecataeus, Herodotus, and Thucydides, Moles 1999; for the notion of an
epistolary proem in Hecataeus and Herodotus, Corcella 1996b; Porciani 1997. Degree of ‘Schriftlich-
keit’: contrast the positions of Murray 2001a, Murray 2001b and Slings 2002 with those of Fowler 2001,
Rösler 2002.
7
See, however, Calame 1995: 12–15 for cautionary words on the danger of considering the
enunciative marks embedded in a text as a direct reflection of its situation of communication:
‘Semiotically speaking, the utterance of the enunciation creates its own world, just as the story creates
its own fiction’.
8
For the methodology see Calame 1995: 12–26 and for Herodotus especially 75–96; Darbo-
Peschanski 1987: 107–26. See also de Jong 1999 and Dewald 1999 (using a different terminology);
and de Bakker 2007: 49–66. I have taken Hude (1927) as the edition of reference for Herodotus; the
translations are adapted from Godley (1920–4).
9
Gazzano 2002: 24–31; Barrett 2002: 62. Definition of ‘messenger characters’ and overall descrip-
tion of diplomacy between cities: Longo 1981: 27–42; Piccirilli 2002; Russell 1999: 63–76 (and 143–89
for a sketch of the ways in which various types of messages can be transmitted); older important works
on Greek diplomacy include Kienast 1973b and Mosley 1973.
104 Greek Beginnings

The formal classification proposed by Deffner and further refined by van den
Hout will be our point of departure.10 But while they had limited their scope to
messages in oratio recta, I have included in my analysis both messages quoted in
direct speech and messages reported in indirect speech, for a total of 121 instances
of mediated communication, either oral or written;11 110 passages concern an oral
(or presumably oral) transmission of the information, while eleven pertain to the
sending of letters; non-verbal communication has not been taken into account. Of
course, only mediated communication in oratio recta is pertinent to a comparison
of the forms of enunciation of written and oral messages; but examining all
mediated communication, in oratio recta and in oratio obliqua, may help define
narrative strategies.12
In slightly more than half of the corpus, the oral message is reported in indirect
speech (forty-three instances of oral message reported by the narrator in indirect
speech, and twenty-one in which oratio obliqua is mixed with oratio recta;13
moreover, in seven instances the decision to send a letter is mentioned, with
more or less stress, in the narrative, while the content of the message is reported
only indirectly). In the remaining examples, the message is given in direct speech
(forty-six instances of oral messages and four of letters), following a procedure
already attested in the Homeric poems. When compared to archaic Greek epic,
however, some important, significant changes in the way the chain of communi-
cation is represented are noticeable.
In the Iliad, a message is generally repeated twice, the first time when the sender
instructs the messenger, and the second time when the messenger transmits it to
the addressee. In general, the message is transmitted verbatim or almost verbatim,
but each time the deictic markers change to adjust to the situation. Thus for
instance in Il. 24. 112–16, Zeus speaks to Thetis as follows:

10
Deffner 1933: 38–40; van den Hout 1949: 30–3. De Bakker 2007: 27–66 offers a nuanced analysis
of what he terms ‘transported discourse’.
11
On the importance of taking into account messages in oratio obliqua see e.g. Hohti 1976: 7–9. For
messages in oratio recta, I have tried to achieve completeness; those reported in oratio obliqua present
an objective margin of arbitrariness (see above on the difficulty of distinguishing between a message
reported in oratio obliqua and a summary by the narrator)—hence the discrepancy in numbers in
respect to de Bakker 2007 (his definition is wider than the one adopted here); but the number of
instances is high enough to allow projections. On the characteristics of characters’ speeches and of
narrator’s speech in Herodotus see Lang 1984: 90 and especially 132–41; Darbo-Peschanski 1987:
107–26; de Jong 1999: 251–61; as for the characteristics and specific uses of indirect speech in
Herodotus, see Heni 1977: 160–2. How to evaluate the alternation between oratio recta and oratio
obliqua is a difficult issue: Darbo-Peschanski 1987: 120 and Bers 1996: 270 and n. 6 have suggested that
in Herodotus the speeches in oratio recta may be even less close to reality than those in oratio obliqua
(opposite view in Waters 1966).
12
Narrative of events, reported speech (when the narrator refers what others say), transposed
speech (indirect speech purporting to relay the very words of the characters), and mimetic (direct)
speech are polyphonically combined in Herodotus’ Histories (Darbo-Peschanski 1987: 115–23); this of
course applies to most other ancient historians as well.
13
Numerous ‘mixed’ instances concern multiple exchanges. It is worth mentioning that out of the
about forty-three messages in indirect speech, thirteen concern oracles; out of the twenty-one instances
mixing oratio recta and oratio obliqua, fourteen concern oracles, and, in this group, the question is
always asked in oratio obliqua, while the oracle’s answer, the important part in a consultation, is quoted
in oratio recta. Cf. Barnabò 1977–8; Darbo-Peschanki 1987: 74–83.
 When a Letter and Why? 105
‘go then in all speed to the encampment and give to your son this message (K غ):
| tell him that the gods frown upon him, that beyond all other | immortals I myself am
angered that in his heart’s madness | he holds Hector beside the curved ships and did
not give him | back. Perhaps in fear of me he will give Hector back.’
(Trans. Lattimore)
A few verses later, Thetis transmits the message:
‘But listen hard to me, for I come from Zeus with a message (˜Øe  Ø ¼ªªº N Ø).
| He says that the gods frown upon you, that beyond all other | immortals he himself is
angered that in your heart’s madness | you hold Hector beside the curved ships and
did not redeem him. | Come, then, give him up and accept ransom for the body.’
(Il. 24. 133–7, trans. Lattimore)
While the message is rendered with remarkable fidelity, the adjustments in the
deictic markers make it clear that this is Thetis speaking.14 Thus in both cases (in
the instruction speech and in the delivery speech), whoever is talking is in charge.
This is obvious for the instruction speech; it might be less so for the delivery
speech. In fact, de Jong (1987: 183) stresses that there are ways in which messen-
gers can deliberately alter the message: for instance, by changing the order of
presentation when there are various points to be conveyed; by modifying the
formulation; by leaving out parts of the message, or conversely by adding some
comments of their own; or by changing its mode of presentation from direct to
indirect speech. This last strategy allows messengers to introduce the message with
a verb of speaking that may convey further information as to the tone of the
message, or serve to dissociate themselves from the statement they convey. In the
Iliad, then, notwithstanding the high level of verbatim repetition of messages,
messengers are not simply intermediaries repeating with precision the words of
someone else; they also are secondary focalizers, interpreting, selecting, modify-
ing, and organizing the information they must convey.
A very different strategy is adopted in the Homeric Hymn to Demeter. Here the
words of the sender are not expressed; the audience only knows that a messenger is
being sent. The first stage in communication between parties, the instruction speech
or Coding Time, is thus systematically omitted. The messenger herself conveys
the message in direct speech, making it clear that the message comes from someone
else (in this case Zeus). Thus, for instance, Iris is sent to fetch Demeter as follows:
First he dispatched golden Iris to fly to and summon his sister, | Fair-haired Demeter
whose form is admired and desired of so many. | Thus her commission (S çÆŁ’), and
Iris obeyed him. (Hy. Dem. 315–17, trans. Hine)
But once she reaches Demeter, she delivers a message from Zeus, in the third
person:

14
Compare Il. 24. 113: Œ Ç ŁÆ ƒ Nb Ł , K b  åÆ ø with Il. 24. 134: Œ Ç ŁÆØ 
çÅ Ø Ł , b  åÆ ø. On messages in the epos and on the principle of literal repetition (with
adaptation of personal and verbal forms to the new pragmatic situation) see Létoublon 1987; de Jong
1987: 180–5 (note at 181 three instances in which messengers speak apparently on their own initiative),
and list of instances at 241–3; on the evolution in the Odyssey towards a greater freedom in the
repetition of the message between ‘instruction’ and ‘delivery’ speech, de Jong 2002: 82–4; general
discussion in Barrett 2002: 56–69; Longo 1981: 46–7; Longo 1997: 661.
106 Greek Beginnings
‘Demeter, father Zeus, whose thoughts are unwithered and deathless, | Calls you to
come to your kind, the gods who are engendered forever. | Go, then, and let not my
message (K e ) from heaven remain unaccomplished!’
(Hy. Dem. 321–3, trans. Hine)
Thus, there is no mention of the first moment in the chain of communication;
however, the messenger’s self-presentation as an intermediary makes it clear that
there was a chain and that the message originated somewhere else.15 Richardson
(1974: 59 and 262) suggests that the reasons for the omission of the first moment
(Zeus’ words) may reside in the characterization of Zeus in the Hymn as a distant,
august figure whose remoteness is enhanced by the avoidance of direct speech; the
direct speech is delivered by the messenger, with adjusted deictic marks.
A similar omission of the instruction speech is observable in Herodotus: in the
Histories the message is formulated only once, and not repeated; and it is generally
revealed not at the moment in which the addresser confides it to the intermediary,
but rather at the moment in which it is transmitted to the addressee. Similarly, the
content of a letter is quoted not at the time of composition, but when it is read out
or at a point midway between when it was sent and its arrival.16 However, in
Herodotus the omission of the first stage in the transmission of the message (the
Coding Time, i.e. the instruction speech) is used to achieve a very different effect
from that aimed at in the Hymn to Demeter. In a number of instances, the message
is given to the recipient without any adjustments in the deictic markers, as if it
were declared directly by the originator to the addressee, without the messenger
acting as an intermediary.17 The presence of the messenger is almost suppressed;
the effect is that of a face-to-face exchange.

4.1.1. Oral Messages

Deffner, analysing only messages in oratio recta, separated the messages in which
the sender appeared to speak directly in the first person through the mouth of the
messenger from those in which the messenger had a degree of autonomy (i.e.
those in which he reported the message by transposing into the third person a text

15
Richardson 1974: 262; Létoublon 1987: 132–5. Létoublon 1987: 134 presents a scheme of eight
possible ways in which the communicative chain may be organized. In theory, a further distinction is
possible, between messenger speeches which repeat so closely the words of the sender that the presence
of the messenger disappears (first-person reporting), and messenger speeches where the messenger
presents himself or herself as an intermediary (third-person reporting). In this sense, the criticism
levelled at Létoublon by Bassi 1997: 336 n. 64 is justified, although ultimately unconvincing: Bassi can
cite only one example of Iliadic first-person reporting, Agamemnon’s reporting of the Dream, and the
text makes it very clear that we have a direct speech inserted inside another speech.
16
Partial exceptions are Hdt. 1. 53, where the message is summarized in indirect speech by Croesus
and repeated in direct speech by the messenger, and Hdt. 8. 114, where the message of the oracle,
formulated in indirect speech by the narrator, is transmitted in direct speech by the messenger.
17
Richardson 1974: 262 notes that the messenger only in rare instances adds his own comments to
a message in the epos; yet this happens with Hy. Dem. 323 and 467–9 (other instances are Il. 2. 33, 8.
423f., Od. 5. 100ff.). Cf. the pertinent reflections of Svenbro 1993: 26–7 on ‘what would happen if
Agamemnon’s message were to be written’, and more generally his chapter 2 (26–43).
 When a Letter and Why? 107

originally formulated in the first person). This distinction was further refined by
van den Hout, who arrived at the following scheme:18
(a) The sender speaks directly, through the messenger, to the far-away addressee.
From the standpoint of narrative logic, there is a rupture, since the moment of the
first encoding of the message (the Coding Time), the time spent in travel by the
messenger (in a sense the very presence of the messenger), and the Reception
Time coincide, with an effect of deictic simultaneity. This mode of presentation
allows the main characters to speak as it were ‘face to face’, notwithstanding the
physical distance. The formal structure of this type of communication is as
follows: the narrator announces the sending of a messenger by someone, whose
speech is then directly introduced: ‘A, sending a messenger . . . , pronounced the
following words: “ . . . ” ’: ( łÆ Œ æıŒÆ/¼ªªº . . . ºª ·).19 This intro-
duction by the narrator is followed by direct speech in the first person. The speech
may take various forms: (1) it may start in medias res, without an introductory
formula;20 (2) it may be introduced by a vocative, sometimes preceded by an ‘O’,
as in 1. 206. 1, the message of queen Tomyris to Cyrus ( łÆ Æ   ıæØ Œ æıŒÆ
ºª · t Æ ØºF  ø);21 (3) it may begin with the formal opening clause:
‘A to B says thus’, as in 3. 122. 3 ( Oæ Å —ºıŒæØ z ºªØ).22 In all three

18
Deffner 1933: 38–40; van den Hout 1949: 29–31. In what follows van den Hout’s data have been
integrated with mine. On forms of address in Greek (but not tied to epistolary communication) see
Dickey 1996.
19
The ellipsis inherent in this structure emerges clearly when there is discrepancy between the
number of messengers sent and the verbum dicendi, as for instance in Hdt. 9. 18. 3 ( łÆ ÆæØ
Œ æıŒÆ ºª · ŁÆæ , t øŒ·), or 8. 28. 1 (ƒ ¨ ƺd  łÆ Œ æıŒÆ Mªæı ·
 O øŒ); when a ‘speaking message’ is sent (Iªªº Å ºªı Æ ), as in 2. 114. 1 (Thonis to
Proteus) or in 3. 122. 3 (Oroetes to Polycrates); or when  Ø is used absolutely with a participle of a
verbum dicendi, as in 4. 80. 2,  ł غŒÅ Ææa e  OŒÆ Æ Å ºªø Ø. In Xerxes’ order
to insult the Hellespont (7. 35. 2: Kºº b t ÞÆ ÇÆ ºªØ æÆæ  ŒÆd I ŁÆºÆ· t ØŒæe
oøæ,  Å Ø  ŒÅ KØØŁE . . . ) the message is presented as an apposition of æÆæ  ŒÆd
I ŁÆºÆ. In one case only, 9. 98. 2 (¸ııå Å e Œ æıŒ æŪæı E Ø  ”ø Ø ºªø), the use of
a herald to express someone else’s words is totally explicit.
20
a1, six instances: 2. 114. 2 and 3 (twice, in the exchange between Thonis and Proteus); 3. 68. 4 and
3. 68. 5 (messages of Otanes to his daughter Phaidyme and her answer, in the context of a very complex
exchange, for which see below); 3. 119. 4 (the wife of Intaphernes to Darius, not in this group in van den
Hout 1949: 31); 4. 80, Sitalkes to Octamasades. Van den Hout has six instances too, because he
discounts 3. 119. 4, but adds 9. 18. 3 ( łÆ ÆæØ Œ æıŒÆ ºª · ŁÆæ , t øŒ· ¼æ
ªaæ KçÅ), an instance which I would rather put in group a2, because of the presence, even though in
second position, of an apostrophe with vocative.
21
a2, sixteen instances, of which five are apostrophes with simple vocative (1. 212, Tomyris to
Cyrus; 4. 126, Darius to the king of the Scythians Idanthyrsus; 6. 97. 2, Datis to the Delians; 9. 60,
Pausanias to the Athenians; 9. 98. 3, Leotychidas to the Ionians), while in the remaining eleven the
vocative is preceded by ‘o’ (1. 206, Tomyris to Cyrus; 3. 14. 10, Psammenitus’ answer to Cambyses; 3.
69. 2, Otanes to his daughter Phaidyme (not in van den Hout 1949: 31); 3. 119. 3 and 5, Darius to the
wife of Intaphernes; 3. 119. 6, answer of Intaphernes’ wife; 4. 127, Idanthyrsus to Darius’ messenger; 7.
35. 2, words of Xerxes to the Hellespont (not taken into account by van den Hout 1949: 31); 8. 29. 1, the
Thessalians to the Phocidians; 9. 18. 3, Mardonius to the Phocidians; 9. 48, Mardonius to the Spartans).
Van den Hout 1949: 31 does not consider 7. 35. 2—one might hesitate over its pertinence; forgets(?) 3.
69. 2; and puts in this category 1. 69 (discussed below, with other messages presenting characteristics
common to both groups).
22
a3, three instances: 3. 122. 3, Oroetes to Polycrates; 5. 24, Darius to Histiaeus; and 7. 150. 2,
message of Xerxes to the Argives (a special case discussed in more detail below, n. 34), with in all
108 Greek Beginnings

cases, an absent person speaks directly, as is made evident in the third type of
opening.
(b) Alternatively, the messenger may transmit the message in the third person,
adjusting the deictic markers. In this case, the structure used is as follows: ‘A sent
messengers, telling them what they should say; and they, once they arrived, said:
“A said/says” ’, followed by one of the following forms: (1) a direct speech in the
third person, without any introductory elements,23 or (2) a speech introduced by a
simple vocative (or ‘O’ plus vocative).24 The deictic markers are mainly oriented
with reference to the sender, but here too one can notice a tendency to organize
the deictic system by taking as temporal origo the moment when the delivery
starts. A good instance is the consultation of the oracles of Delphi and Amphiar-
aus by Croesus:
E Ø b ¼ªØ ººı Ø H ¸ıH ÆFÆ a HæÆ K a ƒæa Kºº › ˚æE 
KØæøA a åæÅ  æØÆ N æÆ ÅÆØ Kd —æ Æ [˚æE ] ŒÆd Y ØÆ æÆe
IæH æ ŁØ ç º.   b IØŒ Ø K a I çŁÅ Æ ƒ ¸ıd IŁ Æ a
IÆŁ ÆÆ, Kåæø E Ø åæÅ Åæ Ø Ø ºª· ˚æE  › ¸ıH  ŒÆd ¼ººø
KŁø Æ Øº ,  Æ  Æ ØÆ r ÆØ FÆ K IŁæ!Ø Ø,  E  ¼ØÆ HæÆ
øŒ H KıæÅ ø, ŒÆd F  Æ KØæøfi A N æÆ ÅÆØ Kd —æ Æ ŒÆd Y ØÆ
æÆe IæH æ ŁØ Æå. (Hdt. 1. 53. 1–2)
Croesus instructed those of the Lydians who were to bring these gifts to the temples to
inquire of the oracles whether he was to send an army against the Persians and
whether he was to acquire an army of allies. When the Lydians came to the places
where they were sent, they presented the offerings, and inquired of the oracles, in these
words: ‘Croesus, king of Lydia and other nations, believing that here are the only true
places of divination among men, has endowed you with such gifts as your wisdom
deserves, and now he asks you whether he is to send an army against the Persians, and
whether he is to add an army of allies.’
The Lydian messengers present thoughts formulated in the past, but with refer-
ence to the spatial place of the actual utterance (); and then advance a request
still formulated by the absent Croesus, but anchored in the here and now.
The formula ‘A to B says thus’ cannot be employed in the opening of this type
of message, since besides necessarily implying the absence of the speaking person,
it also involves a shift to the first person, which is impossible here, since the

instances a passage to the first person in the rest of the message. Van den Hout counted as instances of
this type also 3. 14. 8–9, 8. 140 and 9. 21; but see below.
23
b1, eight instances: 1. 53, messengers of Croesus consult oracles (not in van den Hout); 3. 21. 1
bis, the Ichthyophagi to the king of the Aethiopians, and the king of the Aethiopians to Cambyses; 7.
157, envoys of the Greeks in Syracuse; 8. 75. 2–3 and 110. 2–3, Sicinnus transmitting Themistocles’
message to the Persians and to Xerxes; 8. 142, envoys of Sparta in Athens; 9. 7, envoys of the Athenians,
and with them Megarians and Plataeans, ask the help of the Spartans. Van den Hout 1949: 30–1 also
cites eight instances, because he puts in this group 3. 119. 4, here assigned to a1 (see n. 20 above).
24
b2, eight instances: 1. 60. 5, messengers of Peisistratos; 5. 98, messenger of Aristagoras to the
Paeonians; 6. 105–6, the Athenians to the Spartans; 7. 135–136. 2, the Spartan heralds Sperthias and
Boulis in Susa; 7. 172, the Thessalians to the Greek council; 8. 24. 2, herald sent by Xerxes to the Persian
fleet; 8. 114, herald sent to Xerxes by the Lacedaemonians; 9. 12. 2, herald of the Argives to Mardonius.
Van den Hout 1949: 30–1 had nine, because of his—unjustified—inclusion of 7. 158, Gelon’s answer to
the envoys of the Greeks—but Gelon answers without using messengers.
 When a Letter and Why? 109

messenger relates the message in his own voice.25 This second category main-
tains—in fact, enhances—the distance between the sender of the message and the
addressee. But there is a remarkable degree of freedom, because the messengers’
speech can also include a message of the first type.26 This feature is well exempli-
fied in a message sent by the same Croesus to the Spartans:
› ˚æE    K æÅ Iªªºı Hæ  çæÆ ŒÆd Å  ı ı Æå Å,
Kغ   a ºªØ åæB. ƒ b KºŁ ºª·  0E ł  Æ ˚æE  › ¸ıH
 ŒÆd ¼ººø KŁø Æ Øº , ºªø · t ¸ÆŒÆØ ØØ, åæ Æ F ŁF e
 ‚ººÅÆ ç º æ Ł ŁÆØ,  Æ ªaæ ıŁ ÆØ æ ÆØ B  Eºº,  Æ t
ŒÆa e åæÅ  æØ æŒÆº ÆØ ç º  Łºø ª ŁÆØ ŒÆd Æå ¼ı  ºı
ŒÆd IÅ. ˚æE  b c ÆFÆ Ø Iªªºø KŒÅæıŒ  (Hdt. 1. 69. 1–2)
Croesus, then, sent messengers to Sparta carrying gifts and asking for an alliance,
having instructed them what to say. And they having arrived said: ‘Croesus, King of
Lydia and other nations, has sent us saying this: “Lacedaemonians, the god has
declared that I should make the Greek my friend; now, since I learn that you are the
leaders of Hellas, I invite you, as the oracle bids; I should like to be your friend and ally,
without deceit or guile.” ’ Croesus proposed this through messengers.
A closer look shows that the distinction between these various modes of
communication is not strongly marked. In the first place, often the enunciative
typology does not correspond strictly to the way in which the message has been
introduced. Thus normally messages of the first type (a) are composed in the first
person, since they are supposed to report the very words of the sender. This is
always the case of those messages beginning in medias res (a1). Similarly, those
messages opening with the prescript ‘A to B says thus’ (a3) regularly make an
enunciative shift from the third to the first person. As for the messages opening
with an address in the vocative (a2), they are also composed in the first person,
whether singular or plural,27 with three interesting exceptions: the words ad-
dressed by Xerxes to the Hellespont (7. 35, one of the rare instances in which
the content of the message is expressed at the moment of the instruction speech,
and not at the time of its delivery), and the two messages sent to Darius by the wife
of Intaphernes (3. 119. 3 and 5). In all these instances the intermediaries disap-
pear; the text of the message is, however, given in the third person. The strategy at

25
Van den Hout 1949: 31.
26
Five instances: 1. 69. 1–2, a mixture of b1 and a2 (placed by van den Hout 1949 in a2); 3. 14. 8–9,
7. 150. 2, 8. 140 (discussed below), and 9. 21. 1–2, put by Deffner 1933: 40 in the second (b) category,
but considered by van den Hout as an instance of official introductory formula (a3): æ ƺº Å t
B ¥ ı ƒ ªÆæ ØÇ Ø   Kd f æÆŪf H  Eºº ø Œ æıŒÆ, IØŒ  b ›
ŒBæı æe ÆPf ºª · ªÆæ ºªı Ø·  E, ¼æ ÆåØ, P ıÆ N  c
—æ ø ¥  Œ ŁÆØ FØ, å  Ø Æ Å K c  Å  Iæå · Iººa ŒÆd K  ºØÆæ fi Å
 ŒÆd IæBfi Iå  ŒÆ æ ØÇ Ø. F  N ØÆ ¼ººı  ł ØÆåı B Ø, Y 
 Æ KŒº łÆ c Ø. › b  çØ ÆFÆ I ªªºº . . . This begins as b1, and goes on as a2; the
narrative resumes underlining the presence of the herald, as b.
27
A meaningful alternation, as the messages of Tomyris to Cyrus show: the first (1. 206) is in the
first person plural, and the queen represents her people; the second (1. 212) is in the first person
singular, for the queen has been touched directly in the person of her son.
110 Greek Beginnings

work in these three passages obtains the result of conveying the impression of a
face-to-face conversation, while at the same time stressing the distance between
the king and everyone else.28
As for the messages of the second type, in which the messenger seems to have a
degree of autonomy, they present a wide range of possibilities. When the message
begins in medias res, without apostrophe in the vocative, it is generally in the third
person. Thus the messages sent by Croesus and Cambyses to Delphi and to the
king of the Aethiopians respectively (Hdt. 1. 53, 3. 21. 1); the second half of the
answer of the Aethiopians’ king to Cambyses (Hdt. 3. 21. 3); and the two messages
sent by Themistocles to the Great King through Sicinnus (Hdt. 8. 75. 2–3 and
8. 110. 3).29 A certain number of instances, however, present after an opening in
the third person an enunciative shift towards the first person, in general plural;
this move is particularly frequent in the case of long messages, comparable to
rheseis, as in the speech of the envoys of the Greeks in Syracuse, in the one made in
Athens by the envoys of the Spartans, and in the one with which the Athenians
request a Spartan intervention.30 In these instances the message is attributed by
the narrator to a group (¼ªªºØ) representing a wider community: the shift has
the function of bringing on the scene the voice of the collectivity.31
The tendency to move towards an enunciation in the first person is especially
strong when the message opens with a vocative, because the apostrophe in this
position points to the person speaking. The function and the reasons of this shift
may vary; the length of the message also may play a role. The message sent by
Aristagoras to the Paeonians, for instance, begins with a sentence (in the first
person) in which the messenger appears as an agent of the tyrant; this is followed
by a sentence describing the situation in Ionia; the message moves to a smooth

28
Respectively, Hdt. 7. 35: (Xerxes) Kºº b t ÞÆ ÇÆ ºªØ æÆæ  ŒÆd I ŁÆºÆ· t
ØŒæe oøæ,  Å Ø  ŒÅ KØØŁE  , ‹Ø Ø M ŒÅ Æ Pb æe KŒ ı ¼ØŒ ÆŁ. ŒÆd
Æ Øºf b ˛æÅ ØÆ Æ , X  ª  ºfiÅ X  · d b ŒÆa  ŒÅ ¼æÆ Pd IŁæ!ø
Ł Ø ‰ KØ ŒÆd ŁºæfiH ŒÆd ±º ıæfiH Æ fiH.    c ŁºÆ Æ Kºº  Ø Ø ÇÅ ØF . . . 36. ŒÆd
Q b ÆFÆ K ; Hdt. 3. 119. 3: (Darius)  łÆ b ¼ªªº ºª · t ª ÆØ, Æ Øº  Ø
˜ÆæE ØE Œº; and Hdt. 3. 119. 5: ıŁ  b ˜ÆæE ÆFÆ ŒÆd Łø  Æ e ºª  łÆ
Mªæı· t ª ÆØ, Næøfi A  Æ Øºf  Æ åı Æ ª! Å Œº.
29
The first three passages present a very spare structure, with name of the sender, verb, and object
of the message/request; the other two are construed with the name of the sender, verb of sending
( ł), name of messenger in the accusative with an agreed participle of a verbum dicendi, on which
the subordinate depends (çæ Æ ‹Ø).
30
Respectively, Hdt. 7. 157 and 8. 142 ( łÆ  Æ ¸ÆŒÆØ ØØ/#ŁÅÆEØ + participle
coordinated with the accusative—here, ÆæÆºÅł ı, Å  ı—followed by infinitives); and
9. 7Æ 1–2, where  łÆ  Æ #ŁÅÆEØ introduces a participle referring to the Athenians themselves,
followed by ‹Ø ( . . . ‰ b I Œ K c ¸ÆŒÆ Æ ƒ ¼ªªºØ ƒ I #ŁÅø, – Æ Iª Ø Œ 
ªæø Iªªºı ŒÆd KŒ —ºÆÆØø, ºª  KºŁ Kd f Kçæı·  łÆ  Æ
#ŁÅÆEØ ºª ‹Ø  E Æ Øºf ›  ø F b c å!æÅ IØE, F b ı åı
KŁºØ K Y fi Å  ŒÆd ›  fi Å Ø Æ ŁÆØ ¼ı  ºı ŒÆd IÅ, . . .  E b ˜ Æ   Eºº Ø
ÆN Ł . . . P ŒÆÆØ Æ  ƺº IØ ŁÆ). The first person plural reflects the words pro-
nounced by the Athenians when they instructed the messengers. It is interesting that in these situations
the instructions are never marked with forms of Kºº ÆØ.
31
So also in 7. 172. 2–3: IØŒ Ø b Kd  ı H ¨ ƺH ƒ ¼ªªºØ ºª· ¼æ  ‚ººÅ,
E çıº  ŁÆØ c K ºc c  Oºı ØŒ , . . .  E  ı Ø  N  ı çıº Ø,  Ø b
åæc ŒÆd  Æ æÆØc ºº , ‰ N c  ł, K Æ Ł  Æ › ºª Ø fiH —æ fi Å . . .  E
b ØæÅ  ŁÆ ÆP ØÆ øÅæ Å ÅåÆ! Ø. ÆFÆ ºª ƒ ¨ ƺ .
 When a Letter and Why? 111

close with a first person plural that includes the messenger himself in the
message.32
Conversely, in the narrative of the first return of Pisistratus, were it not for the
clear statement of the narrator specifying that the messengers are repeating what
they have been told to say, the messengers’ words might be considered as
depending on their initiative.33 The enunciative position of the messenger and
the enunciation of the message itself are thus more complex than would appear at
first sight, and convey a number of nuances.
Some flexibility is also occasioned by the contextualization of the message by
the narrator. Thus a message sent by Mardonius to the Greek army (9. 48)
presents the typical hallmarks of what we have defined as the first group (a2),
an apostrophe in the vocative, followed by the first person: Mardonius is here
speaking ( łÆ › ÆæØ Œ æıŒÆ K f ÆæØ Æ ºª · t
¸ÆŒÆØ ØØ . . . ). However, when the narrative resumes it is focused on the
herald, and no longer on Mardonius, bringing about retrospectively a situation
close to that of the messages of the second group (9. 49: › b ÆFÆ YÆ . . .
Iƺº  O ø, IºŁg b K ÆØ Ææ ø fi a ŒÆƺÆÆ).
Moreover, the two opening formulas may be combined. Thus in the message
sent by Xerxes to the Argives (Hdt. 7. 150), the messenger initially appears to be
speaking in his own voice (as in group b): he begins by addressing the Argives,
with a vocative—but then proceeds to formulate the message in a way strongly
reminiscent of the official introductory formula › EÆ fiH EØ  ºªØ.34 The
following shift to a first person (plural, to include all Persians, and then singular)
makes it clear that the words pronounced are those of Xerxes.35

32
5. 98: #æØ ƪæÅ . . .  ł K c æıª Å ¼æÆ Kd f —Æ Æ, . . . n KØc I Œ K
f —Æ Æ ºª · ¼æ —Æ ,  ł  #æØ ƪæÅ › غ ı  æÆ øÅæ Å
ŁÅ    E, X æ  º Ł  Ł ŁÆØ. F ªaæ $ ø Å A Æ I ÅŒ Ie Æ Øº, ŒÆd  E
ÆæåØ fi!Ç ŁÆØ Kd c  æÅ ÆPH· åæØ b ŁÆº Å ÆPE Ø  E, e b Ie  ı  E XÅ
º Ø. ÆFÆ b IŒ Æ ƒ —Æ  Œº. It may be that from F until º Ø the messenger is
referring the ipsissima verba of Aristagoras; but it is also possible that he remains active, and that the
first person plural should be taken to mean ‘us Ionians’.
33
Hdt. 1. 60. 5: Pisistratus and Megacles drive Phye to the city, ææ ı Œ æıŒÆ æ łÆ,
Q a Kƺ Æ Mªæı IØŒ Ø K e ¼ ı, ºª Ø· t #ŁÅÆEØ, Œ Ł IªÆŁfiH ø fi
—Ø æÆ, e ÆPc  #ŁÅÆ Å Ø Æ Æ IŁæ!ø ºØ Æ ŒÆªØ K c øıB IŒæºØ. ƒ
b c ÆFÆ ØÆçØ ºª Œº. The Ø here may reflect not uncertainty on Herodotus’ part
as to what was announced (for if this were the case, we should find Ø much more often, where
instead we find , see below), but rather the fact that the messengers, speaking independently and
autonomously, transmit the message each in his own words. It may be that in an Athenian context
messengers have to be presented as independent, even though they are actually not so—there would be
irony here on the part of the narrator.
34
Hdt. 7. 150. 1–2:  Ø b ¼ºº ºª ºª  Ia c  EººÆ, ‰ ˛æÅ  ł Œ æıŒÆ K
@æª ææ X æ ›æ B ÆØ æÆ  ŁÆØ Kd c  EººÆ. KºŁÆ b F ºªÆØ NE· ¼æ
#æªEØ, Æ Øºf ˛æÅ   E ºªØ·  E  Ç  —æ Å r ÆØ I y  E ªªÆ , ÆEÆ
—æ  F ˜ÆÅ, ªªÆ KŒ B ˚Åç ŁıªÆæe #æ Å . . . j ªaæ K d ªÅÆØ ŒÆa ,
PÆ f ÇÆ  ø ¼ø. Deffner 1933: 40 categorized this within the group b; van den Hout 1949:
31–2 and n. 47 within the group a3. The following narrative does not dissolve the ambiguity.
35
A combination of various elements is also 3. 14. 8–9: Łø  Æ b › ˚Æ  Å a Ø Æ
 łÆ ¼ªªº Næ!Æ ÆPe ºªø ·  Å  ˚Æ  Å, %Æ Ø, Næøfi A Ø ‹ Ø c . . . e
b øåe P Ø æ ŒÆ, ‰ ¼ººø ıŁÆØ, K Å Æ; The message begins with a modified
version of the official introductory formula; there is, however, no shift from the third to the first person.
Van den Hout had classified this as a case of an official introductory formula (a3).
112 Greek Beginnings

It is appropriate to close this discussion with the most complex instance of the
transmission of a message in the Histories, the message conveyed by Alexander of
Macedon to the Athenians (8. 140–1), meant to transmit a message of Mardonius,
in its turn transmitting a message (possibly a letter) from Xerxes.36 Alexander is
introduced as the intermediary chosen by Mardonius in 8. 136: the narrator
explains the reason for this choice, and then gives a detailed genealogy (8. 137–9),
before going back to the message. Thus the status of Alexander as intermediary is
highlighted, and he is presented as an independent messenger (8. 140a: ’¯ªª
b c z #ºÆæ › # ø· ‰ b I Œ K a #Ł Æ I çŁd e
Ææ ı, ºª ). The words with which Alexander begins his speech, an
address to the Athenians, and a formula that introduces the report of someone
else (¼æ #ŁÅÆEØ, ÆæØ  ºªØ), appear to confirm this. However, the
speech continues with the reporting of Mardonius’ own words, in the first person,
a shift possibly prepared by the closeness of the opening words to the ‘official’
introductory formula › EÆ fiH EØ  ºªØ: Mardonius is now directly
speaking through Alexander’s mouth. He begins by announcing, in the first
person, the reception of a message of the King (K d Iªªº Å lŒØ Ææa Æ Øº
ºªı Æ oø, whether written or oral is left unspecified);37 the message itself
follows, presented in the first person and addressed to Mardonius (#ŁÅÆ Ø Ø a
± ÆæÆ a K K b K KŒ ø ª Æ  Æ  Å Ø. F  z, ÆæØ,
 Ø· F b c ªB çØ I, F b ¼ººÅ æe Æ fi Å º Łø ÆP ,
lØÆ i KŁºø Ø, K ÆP Ø. ƒæ  Æ çØ, j c  ºøÆ ª K d
› ºªØ, IæŁø , ‹ Æ Kªg KæÅ Æ.) At this point, Mardonius (and Alex-
ander) are simply transparent transmitters of the King’s voice; but here the
movement backwards begins: Mardonius resumes his own speech, in the position
of the independent messenger ( ø b Iت ø IƪŒÆ ø åØ Ø ØØ
ÆFÆ, j c e  æ ÆYØ ªÅÆØ. ºªø b  E  . . . .) Next, Alexander
resumes his role of independent actor (ÆæØ b ÆFÆ, t #ŁÅÆEØ,
K ºÆ Ø NE æe  Æ. Kªg b æd b P Å B æe  Æ. . . ),
underlining that what preceded were Mardonius’ words, but that he shares his
point of view (he himself adduces further arguments in his final peroration).
Finally, in 8. 141 the narrative resumes, with #ºÆæ b ÆFÆ º·
¸ÆŒÆØ ØØ b ıŁ Ø . . . That the character’s speech is now over is restated
in 8. 142, after a short remark on the behaviour of the Spartans (‰ b KÆ Æ
ºªø #ºÆæ).
The encapsulation of three messages into each other causes a projection
backwards in time, from the performance time of the speech of Alexander, in
which coding time and receiving time coincide (he is speaking directly to the
Athenians), to the reported message of Mardonius, whose coding time is once
removed, and which enjoys two receiving times, the reception by Alexander
some time earlier and that taking place now, and then further back to the
message of Xerxes, placed squarely as the origin of everything, and presented, of

36
Asheri 2003: 357–8 stresses ‘l’abilità di Erodoto a utilizzare tre stili diversi, in gradazione
ascendente . . . : lo stile autoritario e paratattico del messaggio del Gran Re al suo satrapo, quello più
articolato e argomentativo del satrapo e la libera retorica persuasiva del re macedone’.
37
Mardonius’ message is for Asheri 2003: 357 a written communication (‘il discorso di Alessandro
comprende due parti: la lettura del bando di Mardonio . . . e il commento personale di Alessandro’).
 When a Letter and Why? 113

course, in the first person. This chain expresses well the complexity, oppressive-
ness, and ramifications of Persian control, in much the same way as do the
descriptions of the Royal road from Sardis to Susa in 5. 52–3, the angareion (the
race of the Persian messengers through the empire) in 8. 98, and the catalogues
interposed at various points in the narrative. At the same time, the fact that each
time the message is not ‘digested’ and then presented in the speaker’s own words,
but is given in the very words of the sender, enhances the distance—respectful
distance, in the case of Mardonius presenting the words of the king; astute
distance, in the case of Alexander presenting the words of Mardonius.
The other instances combining direct speech and oratio obliqua confirm that it
is not possible to confine within a simple scheme the multiplicity of the ways in
which information is transmitted in the Histories. One point, however, emerges:
while oral messages lacking apostrophe or introduced simply with a vocative are
pronounced by all sort of characters, the very few oral messages introduced with
the ‘official’ formula ‘A to B says thus’ appear only when the person speaking is a
non-Greek (whether speaking to another non-Greek or a Greek).

4.1.2. Written Messages (Letters)

How do the composition and transmission of written messages compare to the

two main ways of presenting of an oral message? There are eleven explicit
references to the sending of written messages in the Histories:
1. the letter with which Harpagus exhorts Cyrus to rebellion, inserted in the
belly of a hare to avoid interception (Hdt. 1. 123. 3–124);
2. the letter written by Cyrus with which he proclaimed himself strategos of
the Persians, presented by him to the Persians as if it had been sent by
Astyages (Hdt. 1. 125);
3. the long letter sent by Amasis to Polycrates (Hdt. 3. 40), and
4. Polycrates’ answer (Hdt. 3. 42. 4–43. 1);
5. the (false) royal letters read by Bagaeus to Oroetes, leading to the death of
the latter (Hdt. 3. 128);
6. the letter of Darius to Megabazus (Hdt. 5. 14);
7. the message tattooed by Histiaeus on the shorn head of a slave, later sent—
once the hair had grown back—to Aristagoras to enjoin revolt (Hdt. 5. 35);
8. the letters sent by Histiaeus to the Persians in Sardis, intercepted by
Artaphernes, and thus leading to the death of all involved (Hdt. 6. 4);
9. the message inscribed by Demaratus on the wood of a writing tablet, then
covered with wax and sent to Sparta, a message that can be deciphered only
thanks to the ingenuity of Gorgo (Hdt. 7. 239);
10. the message ostensibly addressed to the Ionians, but catering also for a
Persian audience, inscribed by Themistocles on the rocks of Artemision
(Hdt. 8. 22);
11. the treasonable exchanges, via letters wrapped around the shaft of an arrow
and then covered with feathers, between Timoxenus, the general of the
Scioneans, and the Persian general Artabazus during the siege of Potidaea
(Hdt. 8. 128).
114 Greek Beginnings

The message sent by Darius to Histiaeus to bring him back from Myrcinus, in
Hdt. 5. 24. 1, is sometimes considered to be a letter, and may have been perceived
as one by the audience of the Histories—but this is not stated explicitly; the same
applies to the message sent by Oroetes to Polycrates (Hdt. 3. 122. 3, discussed
below).
Some other instances in which writing is used to register and transmit a
message are linked to the consultation of oracles; they also represent a mediated
communication, since a message from a god to a mortal is first passed on through
the god’s servant, then written down by the messengers, and finally brought back
to the instigator of the quest. Thus, when Croesus sends messengers to put to the
test the oracles of the Greeks (1. 47–48. 1), when the Athenians consult Delphi
before Salamis (7. 140–2), and when Mardonius sends Mys the Carian to consult
the oracles (8. 135–6. 1), the answer of the god is to be preserved in writing and the
person who had initiated the inquiry is presented as reading its results.38 This may
reflect actual practice, and may be linked to a sense that the exactness of the words
is important (which is an interesting development in itself). The non-verbal
messages in the Histories, such as the exchange of gifts among the Ichthyophagi
and the king of the Aethiopians (Hdt. 3. 20–2), or the gift offered by the Scythians
to Darius (4. 131. 4), are also worth mentioning. Inasmuch as the information is
transmitted not through words but through objects, these messages are reified,
and thus close to written messages; but they differ from them in the same way that
they differ from oral messages, because of their symbolic character, susceptible of
multiple interpretations and verbalizations.39
If we apply to written messages the same grid used for oral messages, we obtain
the following distribution. Two long letters are quoted in extenso, making use of
direct speech: the letter sent by Harpagus to Cyrus and the letter of Amasis to
Polycrates. The first one presents an introductory formula with a vocative (a2), the
second one opens with the formal, official introductory formula ‘A to B says thus’
(a3);40 both are written in the first person of the sender (the second letter presents
the enunciative shift already discussed from the third to the first person singular),
while the recipient is addressed in the second person. Also in oratio recta are the
two short injunctions of the Great King falsified by Bagaeus, presenting variations
on the formal introduction, and in both cases a message in the third person
singular,41 and the address to the Ionians, inscribed by Themistocles on the rocks

38
In writing was also the oracle concerning Delos, 6. 98. 3: ŒÆd K åæÅ fiH q ªªæÆ  æd
ÆPB z· ŒØ ø ŒÆd ˜Bº IŒ Å æ KF Æ. On oracular ‘speeches’ in the Histories see de Bakker
2007: 60–3.
39
On non-verbal communication in Herodotus see Lateiner 1987; Gazzano 2002: 50–62.
40
Respectively, Hdt. 1. 124 and Hdt. 3. 40. The first letter, comprising fourteen lines of Hude’s text,
is introduced with a b ªæ ÆÆ ºª ·, followed by the address Ὠ ÆE ˚Æ  ø,  Œº; in it,
Harpagus presents to Cyrus their common grievances against Astyages, and explains how Cyrus might
easily take power. The second letter—again a long one: approximately fifteen lines of Hude’s text—is
composed in a moralizing style; it is introduced with ªæłÆ K ıº   K غ K  ·
«@ Æ Ø —ºıŒæØ z ºªØ·  Œº . . . ».) Neither has a formal conclusion.
41
Hdt. 3. 128. The two letters are mentioned after others whose content is simply summarized; they
are introduced by the narrator in a very similar way: ØE ¼ºº (ıº ), K fiH KB Æ · t
—æ ÆØ, Æ Øºf ˜ÆæE Iƪæ Ø  E c æıçæØ  Oæ Æ (thus a2/a3); and e ºıÆE
H ıº ø ØE fiH ªæÆ ÆØ B fi , K fiz KªªæÆ· BÆ Øºf ˜ÆæE —æ fi Å Ø E Ø K æØ Ø
KººÆØ Œ Ø  Oæ Æ (thus a1/a3).
 When a Letter and Why? 115

of the Artemision: it opens in the vocative, and is written in the first person
plural.42 As for the message sent by Histiaeus to Aristagoras to exhort him to
rebel, it might appear as the transposition in the narrative of a literal quote in
direct speech—but it is just one word, ‘revolt’.43 In the remaining six instances of
epistolary communication (Cyrus to himself/the Persians, Polycrates to Amasis,
Darius to Megabazus, Histiaeus to the Persians in Sardis, Demaratus to the
Spartans, and the exchanges between Timoxenus and Artabazus) the message is
summarized, with more or less detail, by the narrator of the Histories.44
The epistolary messages in oratio recta should, as regards their enunciation,
correspond to the first category of oral messages, those in which the sender speaks
through the mouth of the messenger: both situations offer the greatest promise of
fidelity to the words of the sender. This hypothesis is confirmed by the formal
analysis: the few epistolary messages of which Herodotus quotes the text are fully
comparable to the first category of oral messages, in the sense that they present
exactly the same variations of introductory formulas. Thus in the prescript of
letters it is possible to find the apostrophe with t + vocative (Harpagus’ letter,
Themistocles’ message to the Ionians); or the more formal › EÆ fiH EØ ºªØ
(Amasis’ letter); but also a combination of the various possibilities. The royal
orders falsified by Bagaeus, opening with a vocative and written in the third
person, are directly comparable with the messages of Darius to the wife of
Intaphernes (3. 119. 3 and 5) and with Xerxes’ insults to the Hellespont (7. 35).
The closeness between the enunciation of letters and that of the first group of
messages is enhanced by the fact that often Herodotus uses the metaphor that
makes writing speak when inserting in his Histories the writing of others, thus
rendering these texts effectively close to oral messages.45
As for those letters whose content is simply summarized without a break in the
diegesis (six instances out of eleven), they also are treated in a way that finds a
close parallel in Herodotus’ way of dealing with oral messages: the procedure by
which the narrator assumes the function of relating the message is directly

42
Hdt. 8. 22: a b ªæ ÆÆ  ºª· @æ  ”ø, P Ø  ŒÆØÆ Œº . . . (a2). The speech
is all built on the opposition ‘we’/‘you’: it closes with an exhortation to remember that IæåBŁ  åŁæÅ
æe e æÆæ I  ø  E ªª, ‘in the beginning the quarrel with the barbarian came to us
from you’.
43
Hdt. 5. 35. 2–3: the narrator gives the content of the message in indirect speech twice ( Å Æ Æ
I Æ ŁÆØ #æØ ƪæÅ and ıº  fiH #æØ ƪæfiÅ Å BÆØ I BÆØ); he then gives the
message itself in conclusion: a b  ª ÆÆ K ÆØ, ‰ ŒÆd ææ Ø YæÅÆØ, I Æ Ø. The
shortness is unsurprising, since the support is a human scalp.
44
These are not messages in oratio obliqua (transposed speech) but simple narrative (reported
speech), and for this reason they are not discussed in Lang 1984 (they correspond to what de Bakker
2007 terms RSA, ‘record of speech act’). On the authenticity of 7. 239 see Corcella 1985.
45
See Porciani 1997: 4. This metaphor is also frequently used for inscriptions, cf. 3. 88. 2 (ªæ ÆÆ
ºªÆ ), 4. 91, 7. 228. 1, 8. 22 (discussed above, n. 42); for letters, cf. 1. 124 (above, n. 40).
Similarly Å Æ ø, with the meaning of ‘signify, allude to, indicate’, is found for both oral and written
messages: see Histiaeus’ message, 5. 35 (above, n. 43); the oracle given by Delphi to the Athenians, and
transcribed by the ambassadors (theoroi), 8. 142. 2; but also a prodigy, 1. 78. 2. This same indecision
between orality and writtenness should then also be assumed for those occurrences, such as at 1. 5. 3, in
which Å Æ ø is used of the activity of the narrator.
116 Greek Beginnings

comparable to that by which he summarizes oral messages, or relates them in

indirect speech.46
The conclusion is evident: from the formal point of view there is no difference
between oral and written messages.47 The two modes of conveying information
are so close that when there are no external elements, it is impossible to decide
whether a message such as 3. 122 (Oroetes to Polycrates, discussed below) or 5. 24. 1
(Darius to Histiaeus) is a written or an oral message.
It is, moreover, possible to show that even from the point of view of the effect
aimed for, there may be no difference between written technology and oral
message. I shall limit myself to one instance only, and compare the stratagem
whereby Themistocles tries to push the Ionians in Xerxes’ army to rebellion—or at
least, to diminish the trust of the Persians in the Ionians—with the proclamation
made by Leotychidas just before the battle of Mycale. Herodotus presents the
events thus:
¨ Ø ŒºÅ Kæ  æd a Ø Æ oÆÆ, K ø K E Ø º ŁØ Ø ªæ ÆÆ,
a  ”ø KºŁ B
fi  æÆ fi Å  æfiÅ Kd e #æ Ø KºÆ. a b ªæ ÆÆ
 ºª· @æ  ”ø, P Ø  ŒÆØÆ Kd f ÆæÆ æÆı Ø ŒÆd c
 EººÆ ŒÆÆıº Ø. #ººa ºØ Æ b æe  ø ª  Ł· N b  E K Ø F
c ıÆe ØB ÆØ,  E b Ø ŒÆd F KŒ F  ı  E Ç Ł ŒÆd ÆPd ŒÆd H
˚ÆæH  Ł a ÆPa  E ØØ· N b Åæ  ø x   ª  ŁÆØ, Iºº’ ’
IƪŒÆ Å Ç ŒÆÇıåŁ j u  I Æ ŁÆØ,  E b K fiH æªø fi , Ka
ı ªø , KŁºŒÆŒ,  Å Ø ‹Ø I’  ø ªªÆ ŒÆd ‹Ø IæåBŁ 
åŁæÅ æe e æÆæ I’  ø  E ªª. (Hdt. 8. 22. 1–2).
Themistocles made his way to the places where drinking water could be found, and
engraved on the rocks writing which the Ionians read the next day when they came to
Artemisium. This was what the writing said: ‘Men of Ionia, you do wrong to fight
against the land of your fathers and bring slavery upon Hellas. It would be best for you
to join us, but if that is impossible for you, then at least now withdraw from the war
yourselves, and entreat the Carians to do the same as you. If neither of these things
may be and you are fast bound by such constraint that you cannot rebel, yet we ask
you not to use your full strength in the day of battle. Remember that you are our sons
and that our quarrel with the barbarian was of your making in the beginning.’48
As Herodotus remarks right afterwards, Themistocles had a double intent, to
induce the Ionians to change sides and join with the Greeks, or, if the writing were
reported to Xerxes, to create mistrust towards the Ionians and keep them out of
the sea-fights.49 The plurality of addressees of this message, as well as the

46
The oral messages in oratio recta of the second type (those in which the heralds speak with a voice
different from that of the sender) do not find a parallel here. Because the herald speaks with a voice
different from that of the sender, the messages cannot be compared to letters; these messages are
mediated communication, but not scripted speech.
47
So already van den Hout 1949: 23: ‘Externally and internally the oldest Greek letter is entirely a
verbal message with relation to one subject only’.
48
On the stratagem, whose historicity is discussed, see S. West 1985: 285–6, who finds the message
far too long for an inscription; Asheri 2003: 223, with bibliography.
49
Hdt. 8. 22. 3: ¨ Ø ŒºÅ b ÆFÆ ªæÆł, ŒØ K  , K’ I çæÆ ø, ¥Æ j ºÆŁÆ a
ªæ ÆÆ Æ ØºÆ  ”øÆ Ø fi Å ÆƺE ŒÆd ª ŁÆØ æe øıH, j K  IØåŁB fi ŒÆd
ØÆºÅŁBfi æe ˛æÅ, I ı Ø fi Å f  ”øÆ ŒÆd H Æı ÆåØø ÆPf I åfiÅ. ¨ Ø ŒºÅ
b ÆFÆ KªæÆł.
 When a Letter and Why? 117

anonymity in the text itself of the sender, makes this a peculiar case: the text is
entirely construed on the dichotomy ‘we’/‘you’. One may thus wonder whether
‘letter’ is the proper term for it. In fact, this text belongs to the communicative
mode defined by Longo as ‘communication from one to many’, whose main
vehicles are the oral Œ æıª Æ (‘proclamation’) and the public inscription.50 In
this specific case a person, Themistocles, is representing the many, the Athenians
and all the Greeks of the Hellenic alliance.
Strikingly, a very similar story, namely, the sending of a message having exactly
the same double function of exhorting the Ionians to revolt or of exciting mistrust
in the Persians, is narrated in book 9 just before the battle of Mycale; this time, the
stratagem is attributed to the Spartan general Leotychidas, who, through a herald,
addresses the Ionians, asking them to leave the ranks of the Persians. Leotychidas
insists in his message on the fact that the Persians cannot understand him, and
closes by encouraging the Ionians to spread the message through word of mouth:
¸ııå Å e Œ æıŒ æŪæı E Ø  ”ø Ø ºªø· ¼æ  ”ø, ‹ Ø  ø
ıªåı Ø KÆŒ , Ł a ºªø. ø ªaæ Pb ı ı Ø —æ ÆØ H
Kªg  E Kºº ÆØ. Ka ı ªø ,  B ŁÆ ØÆ åæc KºıŁæ Å b ø
æH, a b F ıŁ Æ  „Å. ŒÆd  Y ø ŒÆd › c KÆŒ Æ  ø æe
F KÆŒ Æ. ‰ıe b y Kg ıªåØ  F æ ª Æ ŒÆd ›
¨ Ø Œº › K #æ Ø fiø: j ªaæ c ºÆŁÆ a Þ ÆÆ f Æææı  ºº
f  ”øÆ  Ø, j Ø IØåŁÆ K f Æææı Ø Ø I ı E Ø
 ‚ººÅ Ø. (Hdt. 9. 98. 3–4)
Leotychidas through a herald addressed the Ionians saying: ‘Men of Ionia, those of you
who hear, understand what I say. For by no means will the Persians understand
anything of what I tell you. When we join battle, first of all each man must remember
everyone’s freedom, and next the battle-cry “Hebe”: and let those of you who do
not hear me know this from those of you who have heard.’ The purpose of this act was
the same as Themistocles’ purpose at Artemisium; either the message would be
unknown to the barbarians and would prevail with the Ionians, or if it were thereafter
reported to the barbarians, it would cause them to mistrust their Greek allies.
The remark of the narrator, comparing Leotychidas’ proclamation to Themis-
tocles’ stratagem at Artemision, highlights the doubleness of both messages.51 But
even without this explicit signal, the identity in function and closeness in termin-
ology of the two messages is evident, and certainly not accidental. Thus two
communicative strategies, both applied in the realm of the ‘public’, but one relying
on (inscriptional) writing, the other on oral announcement, are used for the same
purpose in exactly the same way (ºÆŁÆ a ªæ ÆÆ/ºÆŁÆ a Þ ÆÆ).

50
Longo 1997: 676–9. Just as writing in Herodotus may be private (the Persian letters) or public
(not only this message, but also other inscriptions), so also oral messages may be private or public.
A good instance is offered by Themistocles’ communications with the Persian generals: eminently
private, these negotiations are undertaken orally by a trusted man, Sicinnus (8. 75 and 110), cf. nn. 23
and 29 above.
51
Hude as well as Flower and Marincola 2002: 273 and 275 consider the remark as Herodotean;
Macan 1908: 796–7, How and Wells 1928: ii, 330, and Legrand consider this an ancient gloss. It is true
that the Greek is rather problematic. Mistrust of the Ionians there must have been, for as Macan 1908:
796 further remarks, the Persians at Mycale were in a position to dispute a Greek attempt at landing,
and their failure to do so is best explained by their mistrust of the Ionians.
118 Greek Beginnings

Let us close this section by underlining those features that are common to both
oral and written messages. Very often messages, both oral and written, are
introduced by a cataphoric  (ºª ), while the narrative restarts with
an anaphoric ÆFÆ, implying that what has been quoted is the very text of the
message. However, some messages, again both oral and written, are introduced by
Ø, while the narrative, when it resumes, refers back with ØÆFÆ. Whatever
the importance one may want to attribute to the alternation in the use of these
pronouns, it is important to note that in Herodotus they are not distributed
according to a written/oral divide.52 Moreover, the terminology used is the
same: Kºº ÆØ for instance is used both for written and oral messages; five
times KØ ººø has the meaning ‘to order’, twice it means ‘to send a letter’.
Kغª ÆØ is used eleven times for ‘to read’, but in nineteen instances it means ‘to
think, to ponder’, and in five other instances it means ‘to choose’ (soldiers, or fast
ships): its use is clearly not limited to written communication.53 Finally, the
introductory formula ‘A to B says thus’ is used for both oral and written messages;
it is moreover exclusively reserved, in both situations, for oriental kings and
tyrants.54

4.1.3. Herodotean Strategies: Functions and Uses

of Epistolary Writing

If oral and written messages are so similar from the point of view of the form, if
moreover the same effects can be obtained with the one or the other, is it possible
to explain Herodotus’ choices? And is it possible in the first place to speak of
choices? Is the insertion of letters to be explained as desire for variatio? As a
consequence of narrative constraints? As a reflection of the traditions received by
Herodotus? The two first contexts presenting epistolary exchanges are the story of
the ascent of Cyrus to the Persian throne and the narrative of the end of Poly-
crates. These are the only instances in which the text of a relatively long letter is

52
See Gazzano 2002: 20–3 (as well as Lang 1984: 132–3) for a discussion of the use of  and
Ø (and of the alternation between oratio recta and oratio obliqua) in the work of Herodotus: they
both conclude that the difference is not significant, and point to the great difference in respect to
Thucydidean usage: famously, after having explained himself in 1. 22 on his method in respect to
speeches, Thucydides will systematically introduce them with Ø. There is, however, consistency in
Herodotus’ systematic use of  to introduce metrical oracles, inscriptions, or quotations from
Homer: for these forms of speech, Ø (or Ø) is never used (de Bakker 2007: 44).
53
Cf. Powell 1938, s.v.; for Kغª ÆØ used of letters, see 1. 124. 1 and 125. 2 (in both instances
Cyrus sending a letter, ıº ); 3. 41. 1, 43. 1 (Polycrates and Amasis) and 128. 3 (the grammatistes of
Oroetes); 5. 14. 2 and 8. 128. 3 (respectively the ıº  of Darius to Megabazus and that of Artabazus
to Timoxenus); 7. 239. 4 (a ªæ ÆÆ of Demaratus). But the term is found also for inscriptions
(compare 2. 125. 6, where the verb is used of an Egyptian interpreting for the narrator a ªæ ÆÆ)
and for oracles (particularly interesting is 8. 136. 1: ÆæØ b Kغ  ‹ Ø c ºªÆ q a
åæÅ  æØÆ, since these oracles are first transmitted to Mardonius in writing, but then ‘speak’:
Kغª ÆØ might here have the generic meaning of ‘learning’).
54
I am not implying with this a derivation of the Greek epistolary prescript from a Persian formula;
van den Hout 1949: 29–30 and Porciani 1997: 34–6, have rightly rejected the notion of a Persian origin
of the Greek epistolary prescript. But Herodotus may have wanted to characterize through the use of
this formula the oriental style of communication.
 When a Letter and Why? 119

given back integrally; they also prove conclusively that these letters owe their
presence in the Histories to a decision of the author, linked to specific reasons.

4.1.3.1. The Place of Letters in the Story of Cyrus’ Ascent

Having related Cyrus’ victory over Astyages, Herodotus makes a pause in the
narrative of events, introducing a dialogue between victor and loser. Such dia-
logues are typical of the Histories: the best known instance is the one having as
protagonists Croesus and Cyrus.55 Here, the characters are the defeated king of
the Medes, Astyages, and Harpagus, the general who, in order to avenge himself,
first sent a letter to Cyrus inciting him to revolt and then convinced the army of
the Medes to withdraw. This is what Herodotus has to say:
When Astyages was a captive, Harpagus came and exulted over him and taunted him
(ŒÆåÆØæ  ŒÆd ŒÆŒæ ), and besides many other bitter words (ŒÆd ¼ººÆ ºªø
K ÆPe Łı ƺªÆ Æ) he recalled his banquet, when Astyages had fed Harpagus his
son’s flesh, and asked Astyages what it was like to be a slave after having been a king.
[2] Fixing his gaze on him, Astyages asked back whether he considered Cyrus’ deed as
his own. (›  Ø æ Øg I æ N øıF ØÆØ e ˚ æı æª.) Harpagus
answered that he had written, and thus the action rightly was his. (& Aæƪ b çÅ,
ÆPe ªaæ ªæłÆØ, e æBª Æ øıF c ØŒÆ ø r ÆØ) [3] Then Astyages showed him
(with the logos) that he was the most foolish and most unjust man of all; most foolish,
in giving another the throne which he might have had for himself, if the present
business was indeed his doing (# ıªÅ  Ø IçÆØ fiH ºªø fi ŒÆØÆ  ŒÆd
IØŒ!Æ KÆ ø IŁæ!ø, ŒÆØÆ  ª, N Ææe ÆPfiH Æ ØºÆ
ª ŁÆØ, N c Ø øıF ª Kæ åŁÅ a ÆæÆ, ¼ººø fi æØŁÅŒ e Œæ); most
unjust, in enslaving the Medes because of that banquet. (Hdt. 1. 129. 1–3)
Harpagus and Astyages refer to events that have already been narrated: Astyages
had served to Harpagus his own child as a meal, because Harpagus had failed to
kill the new-born Cyrus, the son of Astyages’ daughter and a Persian prince. Once
Cyrus had been recognized and had been accepted back into his family, Harpagus
sent him a letter hidden inside the belly of a hare in order to avoid the message’s
being intercepted, with the purpose of avenging himself. Herodotus reports the
text of this relatively long letter in its entirety: this is rather surprising, both
because the content of the letter is fairly straightforward and because elsewhere
in the Histories letters are summarized. In fact, what is striking about this letter is
its mode of conveyance rather than its content. At any rate: Harpagus, having
promised his help, convinced the young prince to rebel against Astyages.
The dialogical coda to the narrative of Cyrus’ ascent to power is clearly part of
an elaborate construction: the dialogue between Harpagus and Astyages is struc-
turally symmetrical to the first dialogue of this same narrative, also taking place
between Astyages and Harpagus (1. 108. 4–5  1. 129).56 This dialogue has the
purpose of promoting a reflection on two themes that will be central in the three
final books of the Histories: the duties of the individual towards his own people,

55
Another instance appears at 3. 14. 9–10, the dialogue between Cambyses’ messenger and
Psammenitus. On the sudden wisdom of defeated kings see the terse comment of Lattimore 1939: 31
‘Fallen kings drop their delusions with their power’.
56
See Heni 1977: 143 n. 161.
120 Greek Beginnings

and the consequences that the private affairs of a few individuals may have over
the destiny of entire nations.57 But the mention of the earlier epistolary exchange
gives Herodotus the opening needed to introduce within the context of the
traditional opposition of ºª and æª, word and action, a third element,
writing, here represented by the letter in the hare’s belly; the overall purpose of
all this is, I suggest, to advance here a first reflection on the purpose of the
Histories themselves.58 Harpagus stresses the primary role of his writing in respect
to the ergon, the action: it is the fact of having written, and not, for instance, his
role in convincing the army of the Medes to withdraw, that allows him to claim
Cyrus’ victory as his own.59 As for the defeated Astyages, he ably uses the logos to
demonstrate the enormity of Harpagus’ error (nor does he seem altogether
convinced of the correctness of Harpagus’ premise).60 The terminology used
and the rather complex articulation of the dialogue—related in oratio obliqua by
the narrator—show that behind this is an Herodotean reflection that goes beyond
the problem of the justness of Harpagus’ revenge; particularly revealing is the
apposition of fiH ºªø fi to IçÆØ, in itself not strictly necessary, and actually
rather difficult to explain.61 Here, the stress on the logos is intended to create a
contrast with the written character of Harpagus’ action; but why the determina-
tive? It is difficult to interpret this other than as a reference by Astyages to ‘his’
logos, the (written) logos composed by Herodotus, in which the story of the fall of

57
Von Fritz 1967, i, Text: 290 rightly draws attention to the extremely tight connection between
received information and Herodotean reworking in this part of the Histories; he also points out where
the seams are still visible. For the Herodotean reworking of traditional themes in the context of the
story of Astyages see also Pelling 1996; Ceccarelli 2005c; and Murray 2001a [1987]: 38 for the
hypothesis of a Median source.
58
On the opposition logos/ergon, which will be particularly developed by Thucydides, see Parry
1981, and for the earlier period, Heinimann 1945: 43–58; Parry 1981: 15–61 (49–50 on Herodotus).
Parry considers that in Herodotus, as in Homer, logos and ergon are complementary; he mentions as
representative of this complementarity Hdt. 5. 24. 1 (in the context of a direct speech: this is the
message sent by Darius to Histiaeus): F b P ºªØ Ø Iºº æªØ Ø r Æ ÆŁ!. Our passage seems
to me closer to the Gorgianic notion of the ºª ıÆ  , which can accomplish (create) ŁØÆÆ
æªÆ (cf. Parry 1981: 44–5).
59
Astyages had initially spoken of æª, but Harpagus speaks of æBª Æ (a term taken up by
Astyages in his final comment, N c Ø øıF ª Kæ åŁÅ a ÆæÆ). æBª Æ, a term with more
abstract connotations, is possibly closer to Harpagus’ way of seeing things—it also does not have the
positive connotations of ergon. A similar striking interaction of ‘saying’, ‘writing’, and ‘action’ is in
Critias fr. B 5 D.–K. (= Plut. Alc. 33. 1), on the return of Alcibiades: ª! Å  l  ŒÆ ªÆª , Kªg Æ Å
K –Æ Ø | r , ŒÆd ªæłÆ sæª æÆ Æ . | çæƪd   æÅ ªº!Å Kd E  Ø ŒEÆØ.
(Analysis of this passage in Iannucci 1998; Iannucci 2002: 35–77).
60
Erbse 1992: 38–9 considers the episode of the letter sent by Harpagus to Cyrus an invention of
Herodotus. In this case, the dialogic coda must all the more appear to reflect Herodotean (and not
traditional) ideas. A similar constellation is in 4. 36. 2, where the act of graphein (‘to draw, to write’) a
geographical map is opposed to its interpretation, in what is clearly a Herodotean comment: ªºH b
›æH ªB æØı ªæłÆÆ ººf XÅ ŒÆd PÆ  Kåø KŪŠ , Q ὨŒÆ 
ÞÆ ªæçı Ø æØ c ªB . . .
61
A verbum demonstrationis with ºªø fi (but always without the article) is also found in 5. 84. 1
(IçÆØ ºªø fi ‰ PŒ IØŒE), in 5. 94. 2 (IØŒ   ºªø fi ), and in 8. 61. 2 (K ºı ºªøfi );
compare moreover 1. 37. 3: ºªø fi IØ . These are all contexts in which there is a dispute among
two parties; but in none of these instances is there any connection with writing. The article is, however,
present in 2. 18. 1, where the narrator speaks in the first person, and where fiH ºªø fi clearly refers to
Herodotus’ own work: ÆæıæØ  Ø B fi ª! fi Å, ‹Ø  Æ Å K d `Yªı ‹ Å Øa Kªg
I Œı Ø fiH ºªø fi , ŒÆd e @ ø åæÅ  æØ ª .
 When a Letter and Why? 121

Media is narrated.62 It is also remarkable that ŒÆÆåÆ æø, the verb with which
Herodotus describes Harpagus’ behaviour towards Astyages, is used only one
more time in the Histories, in 7. 239. 2, in a context linked to the sending of secret
messages, to connote the possibly negative intentions (Herodotus here hesitates
between two versions) of Demaratus’ decision to warn the Spartans by means of
an engraved wooden tablet covered with wax.63 Finally, the emphasis on a letter as
the cause of the ruin of the power of the Medes points back to the determining
action of Deioces, the founder of the empire: the refusal of face-to-face communi-
cation and the imposition of a filter through the use of writing as the main
instrument of communication.64 All this shows that, as with Cyrus’ stratagem of
writing to himself a letter supposedly sent by Astyages in which the latter appears
to give him supreme command (a very similar strategy will be deployed in 3. 128.
2–5), the epistolary exchange between Harpagus and Cyrus is an integral part of
the Herodotean version of the ascent of Cyrus to the throne, and not just a
marginal addition.

4.1.3.2. Polycrates: A Greek Tyrant between Egypt and Persia

The other long letter of the Histories (approximately fifteen lines of Hude’s text) is
the one sent by Amasis to the tyrant of Samos Polycrates. It is likely that
Herodotus had a particularly important role in the ‘writing’ of this letter. Firstly,
in this exchange the Egyptian pharaoh assumes the role that had been Solon’s in
the first book; there is a close link, at the level of both terminology and content,
between this part of the Histories and the conversation between Croesus and
Solon.65 Moreover, the sending of letters signposts important articulations of this
part of Herodotus’ work: letters play a role in the end of Polycrates, and letters will
bring about the death of Oroetes.
Polycrates’ story opens with a short narrative, meant to illustrate the tyrant’s
extraordinary good fortune. This fortune preoccupies the pharaoh, who decides to
write to Polycrates, to warn him of the envy of the gods. The letter is first

62
See Pucci 1993: 13–14 for discussion of a passage (1. 90. 3), in which the verb KƺغºªÅ  in the
mouth of a character reflects the retelling of Herodotus’ own story, with an effect of mise en abyme.
63
On the primary sense of åÆ æØ (to rejoice over the offer of a sacrifice: rather appropriate in both
cases) see 55–6 and 90–8 above; Wachter 1998; Day 2010: 234–54.
64
The story of Deioces is told at Hdt. 1. 95–101 (for his introduction of messengers and writing see
Hdt. 1. 99–100, and above, 88–9). On the despotic character of Deioces’ writing see Steiner 1994: 130–
2, who moreover stresses the exemplary character of this narrative, the first one to thematize the
conquest of absolute power over an entire people by an individual; Ceccarelli 2002. Von Fritz 1967: 282
remarks that Deioces’ desire for power is expressed through the Greek term tyrannis.
65
For Heni 1977: 65 and n. 208 the letter is a typically Herodotean instrument, and does not come
from the tradition to which Herodotus owes the more ancient story of the ring. Amasis’ words in 3. 40. 2:
KØ Æ ø fi e ŁE ‰  Ø çŁæ can be directly compared to Solon’s in 1. 32. 1: KØ    e
ŁE A Ke çŁæ  ŒÆd ÆæÆåH; in both cases this opens the discussion, giving the grounds for
the warning. `N! and æææØÇ, used infrequently by Herodotus and exclusively in direct speech
(here respectively in 40. 2 and 40. 3), appear also in the logos of Croesus (respectively in 1. 32. 5 and
1. 32. 9). ÆN! appears also in 7. 46. 4 (words of Artabanus, where also çŁæ is referred to the god),
and in the words of the commanders of Phoceans and Athenians in 9. 17. 4 and 27. 3; æææØÇ
appears in the speech with which Leotychidas retells the story of Glaucus, in 6. 86d (another character
ruined by his avidity and desire for riches; see Nenci 1998: 248–50).
122 Greek Beginnings

introduced by the narrator; it opens with a formal prescript in the third person,
and then continues in the first person:
ªæłÆ K ıº   K غ K  · @ Æ Ø —ºıŒæØ z ºªØ· f b
ıŁ ŁÆØ ¼æÆ ç º ŒÆd E s æ Æ, K d b ƃ Æd ªºÆØ Pıå ÆØ PŒ
Iæ Œı Ø . . . (Hdt. 3. 40. 1–2)
Having written these thoughts in a letter he sent it to Samos: ‘Amasis to Polycrates
thus speaks. [2] It is pleasant to learn that a friend and ally is doing well. But I do not
like these great successes of yours . . . ’
The letter ends abruptly, without any final wishes; the narrative goes on with
ÆFÆ Kغ  › —ºıŒæÅ ŒÆd ø
fi ºÆg u ƒ s  Ł › @ Æ Ø . . .
(Hdt. 3. 41. 1)
Polycrates, reading this and understanding that Amasis’ advice was good . . .
The physical time needed for the message to be brought to Samos and delivered
to the tyrant disappears here completely: , the cataphoric deictic referring to
the content of the letter, corresponds to the ÆFÆ that, when the diegesis resumes,
refers back to the content of the letter, but this time from the point of view of
Polycrates and not Amasis. The letter is clearly a substitute, an expedient to
replace a conversation that is made impossible by the distance between the two
protagonists.
The tyrant follows the pharaoh’s counsel, throws a ring in the sea, but miracu-
lously finds it again.
Polycrates saw the hand of heaven in this matter; he wrote a letter and sent it to Egypt,
telling all that he had done, and what had happened to him (e b ‰ K BºŁ ŁE
r ÆØ e æBª Æ, ªæçØ K ıº  Æ a Ø Æ Ø x Æ ŒÆƺºÅŒ, ªæłÆ
b K `Yªı KŁÅŒ).66 When Amasis read Polycrates’ letter (Kغ  b ›
@ Æ Ø e ıº  e Ææa F —ºıŒæ wŒ), he perceived that no man could
save another from his destiny, and that Polycrates, being so continually fortunate that
he even found what he cast away, must come to an evil end. [2] So he sent a herald to
Samos to renounce his friendship. ( łÆ  ƒ Œ æıŒÆ K   Øƺ  ŁÆØ çÅ c
Ø Å) (Hdt. 3. 42. 4–43. 2)
The conclusion of this part of the story shows that the epistolary exchange is
reserved for private affairs; when it becomes necessary to break a friendship,
Amasis sends a herald.67 The peculiarity of the Herodotean narrative emerges
vividly when compared with the way in which Diodorus Siculus (writing in the
first century bc) relates the events. In Diodorus, Polycrates behaves unjustly and
violently (ØÆ ø); for this reason, Amasis sends him envoys (æ ıÆ ), to
convince him to mend his ways. But Polycrates persists, and as a result the

66
The content of this second letter is not given: the repetition would have been pointless. Moreover,
in this way the contrast between Amasis and Polycrates is stressed: as pointed out by Heni 1977: 68,
while Amasis has the ability to interpret events, seeing them in the larger context, Polycrates can only
tell what happens to him.
67
Whose personality is absorbed into the personality of the pharaoh, according to one of the
possible modalities for the transmission of oral messages (here, however, the exchange is summarized
by the narrator).
 When a Letter and Why? 123

pharaoh denounces the friendship through a letter.68 The distribution of oral and
written messages is the exact opposite of what we find in Herodotus; it corres-
ponds well to the changed diplomatic habits of Diodorus’ times.
Polycrates’ downfall, already announced in the letter sent by the tyrant to
Amasis, is brought about by the satrap of Sardis Oroetes.69 This part of the
story too involves heralds and exchanges of messages: according to one of the
versions related by Herodotus, one of the reasons for Oroetes’ hatred of Polycrates
was the lack of respect shown by the tyrant towards a messenger of the satrap—
and thus towards the satrap himself;70 moreover, the immediate cause of Poly-
crates’ ruin is a deceitful message sent by Oroetes to Polycrates (3. 122. 3–4)—it is
unclear whether written or oral. The text is as follows:
› b t  Oæ Å ƒÇ  K ƪÅ fi Å B fi bæ ÆØæı Æ F NŒÅ fiÅ  
 æ  e ˆ ªø ¼æÆ ¸ıe K   Iªªº Å çæÆ, ÆŁg F —ºıŒæ
e  . . . ÆŁg t ÆF Ø ØÆ  ›  Oæ Å  łÆ Iªªº Å ºª ·
 Oæ Å —ºıŒæœ z ºªØ. ıŁ Æ  KØıº Ø b æ ª Æ Ø ªºØ Ø,
åæ ÆÆ  Ø PŒ r ÆØ ŒÆa a çæ ÆÆ. ı z Ø Æ OæŁ! Ø b
øı, ! Ø b ŒÆd K : K d ªaæ Æ Øºf ˚Æ  Å KØıº Ø ŁÆ ŒÆ
Ø F KƪªººÆØ ÆçÅø. ı K b KŒŒ Æ ÆPe ŒÆd åæ ÆÆ, a b
ÆPH ÆPe å, a b K b Æ åØ· ¥ Œ  åæÅ ø ¼æØ B ± Å  Eºº.
N  Ø IØ Ø a æd H åæÅ ø,  ł ‹ Ø Ø Ø Æ ıªåØ K!,
fiH Kªg Iø. (Hdt. 3. 122. 1–4)
Oroetes, then at Magnesia which is above the river Maeander, sent Myrsus son of
Gyges, a Lydian, with a message to Samos, having learned Polycrates’ intention . . .
Learning then that he had this intention, Oroetes sent him this message: ‘Oroetes to
Polycrates thus speaks: I find that you aim at great things, but that you have not
sufficient riches for your purpose. Do then as I direct, and you yourself will succeed
and will save me. King Cambyses aims at my death; of this I have clear intelligence.
Now if you will transport me and my riches, you may take some yourself and let me
keep the rest; thus you shall have wealth enough to rule all Hellas. If you mistrust what
I tell you about the riches, send someone who is most trusted by you and I will prove it
to him.’
A messenger, Myrsus son of Gyges, is sent to Samos to hand over the message—
but the term used, Iªªº Å, here and two sentences later, can be used indifferently
for oral or written messages; the messenger is then subsumed in the personality of
the sender, as in the messages of the first type.71 The message itself is comparable

68
D. S. 1. 95: because Polycrates behaved violently and unjustly, e b æH ºªÆØ æ ıa
I  ºÆÆ ÆæƌƺE ÆPe Kd c æØÅÆ· P æ å ’ ÆPF E ºªØ KØ ºc
ªæłÆØ c çغ Æ ŒÆd c  Æ c æe ÆPe Øƺı . Cf. Heni 1977: 61ff., 76 ff.
69
The exchange between Amasis and Polycrates makes sense only in connection with the end of
Polycrates (and Oroetes later): if this is not (yet) an epistolary novel, the action definitely revolves
around letters.
70
Hdt. 3. 121. 1–2: according to a minority of informants, Polycrates was lying with Anacreon,
turned towards the wall; when the envoy of the satrap arrived, he did not turn himself towards the
envoy, nor deigned to give him an answer. The other version (for both we have to be content, in terms
of source-indication, with a ºªı Ø) sees the reason for Oroetes’ hatred in the insults addressed by
another Persian to Oroetes because of Polycrates.
71
Rosén 1987 (followed by Medaglia, in Asheri and Medaglia 1990) prints  łÆ K ƪªº Å at
3. 122. 3, on the basis of the codex D and of a comparison with 3. 84. 2, where an K ƪªº  is
mentioned; I prefer to follow Hude and Legrand. Iªªº Å may refer to a written message, a letter,
124 Greek Beginnings

to other Herodotean letters: a prescript of the formal type, followed by a first

person singular, and no greeting formula. The reaction of Polycrates to the
message (ÆFÆ IŒ Æ —ºıŒæÅ l ŁÅ  ŒÆd K º, ‘Hearing this, Poly-
crates was pleased and accepted’) does not allow a decision, because IŒ ø is of
course used for oral messages, but also for a written one in 1. 125. 1, where the
verb describes the reaction of Cyrus at the arrival of Amasis’ letter. Asheri
considers this message an epistle—of course, an invented one, be it by Herodotus
or by an earlier tradition; van den Hout, and with him the majority of scholars,
prefer to see in it an oral message.72 The anecdote about the beginnings of
Oroetes’ hatred for Polycrates, relating it to Polycrates’ offensive behaviour in
the context of the transmission of an oral message, may well have influenced the
reading of this message as oral too. But Asheri’s position on the whole allows a
richer understanding of the story of Polycrates, highlighting the inverted parallel-
ism between the letter of Amasis and the message of Oroetes (the first aiming at
convincing Polycrates that he should deprive himself of something precious, the
other stimulating Polycrates’ greed so as to bring about his downfall). At any rate
the message, written or oral, marks the end of Polycrates, who falls for the trap of
the satrap and, blinded by his desire to acquire riches, goes to Sardis.
Appropriately, letters play an important role in the story of the fall of Oroetes.
The connection between these two events (the end of Polycrates and that of
Oroetes) is made explicit by the narrator in two statements, strategically located
at the beginning and at the end of his account of the end of Oroetes: ‘But not long
after, atonement for Polycrates overtook Oroetes’, and ‘Thus atonement for
Polycrates the Samian overtook Oroetes the Persian.’73 This is an authorial
interpretation; the narrator is, however, also aware (and makes his readers
aware) of motivations specific to the Persian world. Among these, pride of place
is given to the murder by Oroetes of a messenger sent by Darius, who had brought
him unpleasant news: again a communication problem, and one that creates a
further parallel between the end of Polycrates and that of Oroetes—disregard for a
messenger had been mentioned by the narrator as one of the reasons of Oroetes’
hatred. In both instances, we should not take these remarks as marginal side
issues: there, the narrator had taken pains to report the opinion of a minority;
here, the killing of one or more royal messengers is mentioned twice, once in the
singular, as one of the crimes listed by the narrator in 3. 126. 2, and once again,

especially in those cases where one finds the iunctura Iªªº Å ºªı Æ oø, as in 8. 140 (cf. above
112), in 3. 69. 1 (Otanes and his daughter Phaidime), and 2. 114. 1 (Thonis sending to Proteus an
Iªªº Å ºªı Æ ). Only in one instance is it possible, however, to be certain of this—the message
sent by Darius to Megabazus, Hdt. 5. 14. 2 (ÆP ŒÆ b ƒf Ł çæø c Iªªº Å Kd e
 Eºº , æÆØøŁd b ØE e ıº  fiH ªÆÇø)—because the written character of the
Iªªº Å is proved by the use of ıº  (and of ªæ ÆÆ a few lines earlier, at 5. 14. 1) for this same
message. In many other cases it is clear that Iªªº ÆØ are oral messages, often simple voices.
72
Asheri 1990: 338; van den Hout 1949. Heni 1977: 71 n. 240, in discussing this message, which he
considers to be oral, remarks: ‘Ein Unterschied zwischen dem Stil des Briefs und der Botschaft kennt
Herodot wohl noch nicht’. See the discussion above.
73
3. 126. 1: åæø
fi b P ººfiH o æ ŒÆd  Oæ Æ —ºıŒæ  Ø BºŁ, cf. 128. 5: oø
c  Oæ Æ e —æ Å —ºıŒæ F Æ ı  Ø BºŁ. That this remarkable narrative
unity is due to Herodotus’ art is beyond discussion—Jacoby 1913: 488 speaks of ‘Kunst der
Zusammensetzung’.
 When a Letter and Why? 125

this time in the plural—as if by then killing royal messengers had become a habit
of Oroetes’—by Darius himself when he lists (in direct speech, addressing the
other Persians) his grievances against the satrap (3. 127. 3).
For all these reasons, Darius’ decision to get rid of the powerful satrap of Sardis
does not come as a surprise. Bagaeus is the man designed by lot to accomplish the
mission—and he will acquit himself of it through letters.
Bagaeus, having drawn the lot, did as follows: having written many letters concerning
many things he put the seal of Darius on them (ıº Æ ªæÆł  ººa ŒÆd æd
ººH åÆ æŪ ø çæŪE çØ Kƺ c ˜Ææ ı), and then went with
them to Sardis. [3] When he got there and came into Oroetes’ presence, he took out
each letter in turn and gave it to one of the royal scribes to read—all the governors of
the King have scribes (IØŒ  b ŒÆd  Oæ ø K ZłØ KºŁg H ıº ø £
(ŒÆ  æØÆØæ  K ı fiH ªæÆ ÆØ B fi fiH Æ ØºÅ øfi Kغª ŁÆØ—
ªæÆ ÆØ a b Æ ØºÅ ı ƒ  oÆæåØ åı Ø); Bagaeus gave the letters
(ıº Æ) to test the spearmen as to whether they would consent to revolt against
Oroetes. [4] Seeing that they were greatly affected by the rolls and yet more by what
was written in them, he gave another, in which were these words: ‘Persians! King
Darius forbids you to be Oroetes’ guard.’ (›æø  çÆ   ıº Æ  ı
ªºø ŒÆd a ºª Æ KŒ H ıº ø Ø Çø, ØE ¼ºº K fiH KB Æ ·
t —æ ÆØ, Æ Øºf ˜ÆæE Iƪæ Ø  E c æıçæØ  Oæ Æ.) Hearing this,
they lowered their spears for him. [5] When Bagaeus saw that they obeyed the letter so
far, he was encouraged and gave the last roll to the scribe, in which was written: ‘King
Darius instructs the Persians in Sardis to kill Oroetes’ (Ng b F çÆ › BƪÆE
ØŁ ı fiH ıº ø fi , KŁÆFÆ c ŁÆæ Æ e ºıÆE H ıº ø ØE fiH
ªæÆ ÆØ B fi , K fiz KªªæÆ· BÆ Øºf ˜ÆæE —æ fi Å Ø E Ø K æØ Ø KººÆØ
Œ Ø  Oæ Æ). Hearing this the spearmen drew their scimitars and killed him at
once. Thus atonement for Polycrates the Samian overtook Oroetes the Persian.
(Hdt. 3. 128. 2–5)
This story attests to the enormous power of the king’s writing in Persia; it also
illustrates the way in which the King’s written decisions were proclaimed. It is now
possible to better understand the motivation of Cyrus’ decision to read in front of
the Persian army a (false) letter from Astyages containing his nomination as a
general (Hdt. 1. 125. 2). This power derives from the complete homology between
King’s words and King’s writing: the guards show religious respect for the rolls,
and even more for what is written in them, for these are the very words of the king
(Æ ; the apostrophe in the vocative points in the same direction, conveying
the feeling that the King is speaking through the roll) brought to the knowledge of
the people through the use of a staging specific to their diffusion.74 The power of
writing is also made clear by the power attributed to the grammatistes, the
secretary: this scene presupposes that the only person having the right (and the
ability?) to read the King’s letter is the secretary. If grammatistai were nominated
and attached to the satraps by the kings, then the royal administration had in
them a powerful means of controlling the activities of the satraps.75

74
For a fascinating contrastive account of the interplay between the written text and its authorita-
tive oral (or written) diffusion within the Persian empire and in the world of the Greek poleis see
Bertrand 1999: 96–106.
75
See on the Persian royal messengers Briant 1996: 354–7 and 382–4. Secretaries are mentioned
also in Hdt. 7. 100. 1 (the ªæÆ ÆØ Æ at Doriscus note the answers given to Xerxes by the various
126 Greek Beginnings

4.1.3.3. Letters and Narrative Structure

The detailed analysis of these two epistolary exchanges has shown why, in these
instances, communication through letters rather than oral messages is chosen. In
the narrative of Cyrus’ ascent to power, the use of writing for communication had
first marked the beginning of the Median kingdom, with Deioces; it also marks,
with Harpagus’ letter, its end. Cyrus demonstrates his ability to exploit this
instrument with the fabrication of a false letter that will give him royal power:
he is well equipped to supplant the Medes. The more complex sequence of events
narrated in the third book helps give precision to the connotations of epistolary
communication. Amasis’ warning could certainly also have been given in the
course of a conversation, such as the one between Solon and Croesus; a ‘wise
adviser’ was anyway needed to introduce the story of the ring. But a meeting
between Amasis and Polycrates could have been dramatized only at the cost of
a number of implausibilities; a letter was an ideal solution, a solution moreover
that allowed the construction of a tissue of letters around Polycrates’ story.
The exchange between the tyrant and the pharaoh prepares for the exchange
between Oroetes and Polycrates, and the death of Polycrates—lured by an untrue
message—parallels that of Oroetes—destroyed by the power of the King’s writing
and by his disrespect for royal messengers. The stress on writing in this part of the
Samian logos further prepares the reader for the role that will be played by
Maeandrius, the ‘scribe’ (ªæÆ ÆØ  ) of Samos, whose exemplary role of
‘master of writing’ between tyranny and democracy has been well described by
Detienne.76 It is worthwhile pointing out that Maeandrius himself plays a role in
the death of Polycrates: it is he who is sent by the tyrant to check that the riches
promised by Oroetes do indeed exist; it is he, a ‘master of writing’, who falls for
the satrap’s trick of covering eight coffers full of stones with a thin layer of gold
(Hdt. 3. 123. 1–2).

4.1.4. Connotations of Letter Writing

The two aspects of Herodotean letters most often highlighted are their relatively
extraordinary character, and their oriental, tyrannical connotations. To begin with
the first point: the majority of written messages are composed or sent in

contingents of the army) and 8. 90. 4 (ªæÆ ÆØ Æ accompany Xerxes in Greece; at Salamis, they
assist to the battle and take note of the name, the patronymic, and the city of origin of those trierarchs
who distinguish themselves in the battle). The title of çØØŒØ c Æ ºØ in Xenophon (Anab. 1. 2. 20)
probably refers to one of these functionaries.
76
Detienne 1988b: 73–81. Herodotus (3. 142–3) narrates how Maeandrius first attempts (publicly)
to restore freedom to the Samians, judging that Polycrates had been wrong in acting as master of men
‘not inferior to himself ’ ( Çø IæH ›  ø øıfiH), but then, when forced to give accounts,
retires to the acropolis, ‘sends for each man separately’ ( Æ   (Æ (ŒÆ ) as if he were to
render his accounts individually, and throws them all in bonds. The story presents numerous
elements in common with that of Deioces (analysed above, 89): cf. Steiner 1994: 173–4; Bertrand
1999: 93–5.
 When a Letter and Why? 127

exceptional, rather adventurous, circumstances.77 The letters of Harpagus, of

Histiaeus, of Demaratus, and of Timoxenus are secret messages, inscribed on
unconventional materials (the head of a slave, the wood of a waxed writing tablet)
or conveyed to their addressees through stratagems (the belly of a hare, an
arrow).78 Cyrus writes himself a false letter; just as false are the letters read on
the orders of Bagaeus in front of Oroetes. The message addressed by Themistocles
to the Ionians has a double purpose, and occupies pride of place in a gallery of
astute stratagems. These stories, if they were traditional, would have been deemed
by Herodotus worthy of preservation and insertion in the Histories because they
are good stories, indeed, extraordinary and fascinating ones—and of course it is
no surprise that they should be part of narratives of deceit, since the overall frame
is that of a great war, and, as pointed out by Rosenmeyer, ‘political and military
contexts invite the use and abuse of letters in precisely this way’.79 As for Cyrus,
we know that Herodotus made a choice, because he tells us himself that the story
of Cyrus’ rise to power could have been told in three other different ways;
evidently the one he chose appealed to him because it allowed him to construe a
coherent and significant ensemble—and attention to communication is an im-
portant part of this construction. Similarly, the decision to bestow attention on the
otherwise extremely short message sent by Histiaeus—a message that could have
been conveyed orally (are we to believe that the man remained unaware of the
import of the text tattooed on his head?)—is probably because this way of
conveying a message is squarely located at the intersection of writing and mutila-
tion, showing how tyrannical uses of writing may enter the Greek world.80
There are, however, a few epistolary exchanges that lack connotations of
duplicity and astuteness: that between Amasis and Polycrates, and Darius’ letter
to Megabazus. The latter represents the ‘zero degree’ of epistolary writing, Persian
normality, and as such may have been part of a tradition received unproblematic-
ally by Herodotus, or may have been his interpretation of how things would have
happened, on the basis of his knowledge of Persian affairs.
In general, the decision to insert in the Histories a number of written messages
may be explained by the relatively exceptional aspect of epistolary communication
in the Greek world and by Herodotus’ own strong interest in communicative
processes and their manipulation, evidenced not only by the instances of written
and oral communication discussed, but also, for example, by his description of the
Persian royal road (5. 52–3) and the Persian messenger system (8. 98), and

77
Harris 1989: 88–9; Lewis 1996: 143–8; Rosenmeyer 2001: 45–55. More general discussion of the
manipulation of signs in Herodotus in Hollmann 2005, with ample bibliography.
78
For this reason they have all been suspected of being ‘inventions’ by Fehling (1981: 198). Of
course, no one thinks that Herodotus saw any of these letters, nor that the texts of the (Herodotean)
messages correspond exactly to whatever may have been sent—this applies in exactly the same way to
oral messages. However, Herodotus may have relied on traditions whose preservation (or indeed
invention) may have been due to the extraordinary character of the event narrated.
79
Rosenmeyer 2001: 47. For the use of letters in war see e.g. Polyaenus 4. 2. 8 (Philip II of
Macedon), 4. 3. 19 (Alexander), 4. 11. 2 and 4. 11. 3 (Cassander).
80
On this, Steiner 1994: 154–9; Munson 2007: 165–7. Entrance of tyrannical writing into the Greek
world: e.g. the tattooing, on the orders of Pericles, of the surviving Samians after their revolt, Duris
FGrH/BNJ 76 F 66 and Plut. Per. 26. 4, with Ar. fr. 71 K.–A (but see below, 240 n. 190, for doubts on the
story).
128 Greek Beginnings

conversely by his narratives of the deaths of messengers, such as are recounted in

the story of the wrath of Talthybius (7. 133–7).
The oriental connotation of letter writing in Herodotus has also attracted
attention. Out of the eleven instances of epistolary exchanges in the Histories
none reflect intra-Greek exchanges. The ‘open letter’ of Themistocles to the
Ionians is addressed by a Greek to Greeks—but Themistocles has in mind also
other readers, the Persians; similarly, if Demaratus has to send a letter to the
Lacedaemonians (and a peculiar one, cunningly inscribed below the waxed
writing surface, a message that only a woman will be able to decipher), it is
because he is at the moment in Susa with the Great King. It may not be an
accident that the only two Greeks writing letters to Greeks, Themistocles and
Demaratus, operate in their letter writing (and at least in the case of Themistocles
also in other matters) with cunning and trickery; and just as Demaratus has found
an asylum in Persia, so will Themistocles.
Four other letters, those of Harpagus and Cyrus, that of Bagaeus, and the letter
of Darius to Megabazus, concern Persians only, while two testify to exchanges
between Greeks and Persians (the letters of Histiaeus to the Persians in Sardis, and
the correspondence exchanged between Timoxenus and Artabazus in the course
of the Persian siege of Potidaea). The remaining epistolary messages are the letters
exchanged between Amasis and Polycrates, an Egyptian pharaoh and a tyrant; and
the word ‘apostasis’ tattooed on a head and sent by Histiaeus, a tyrant owing his
tyranny to Persia, to Aristagoras, another tyrant. Thus the fact of sending written
messages appears indeed characteristic of an oriental and/or despotic milieu.81
A comparison with the distribution of oral messages confirms this slant: Hero-
dotean character-speeches appear divided almost equally between Greeks and
non-Greeks;82 as we have just seen, this is emphatically not the case with written
messages. The latter are in Herodotus clearly marked as the chosen means of
communication between the Great King and his satraps.83
It is, however, necessary to qualify on at least two counts these statements on
the oriental, tyrannical, and deceptive connotations of written communication in
the Histories. In the first place, the Persian habit of sending written messages has
to be viewed against the fact that Persians tend anyway to make use of intermedi-
aries for communicative purposes, even when there is no actual physical distance
between sender and addressee, in order to create a distance between sender and
addressee. Thus, for instance, just before Salamis, Xerxes relies on Mardonius to
consult the Persians on whether he should engage in a sea-battle (8. 67. 2–68. 1);
this way of presenting Persian interaction has the purpose of emphasizing the
aloofness and greatness of the king. As suggested by de Bakker, Herodotus may
have modelled his portrayal of the elevated position of the Persian king in relation
to his subjects on that of Zeus in the Iliad, the only god who does not communi-
cate with mortals directly but always through a messenger.84 In this context, de

81
A reading developed by Steiner 1994: 127–39 and especially 149–54.
82
De Bakker 2007: 11 ascribes 812 character-speeches to Greeks and 815 to non-Greeks.
83
Cf. Hartog 1980: 286–8: ‘la lettre a cours surtout dans le monde barbare et non dans le monde
grec, à quelques exceptions près . . . la lettre transmet des informations ou des instructions, elle est un
mode secret de communiquer et, au total, un mode de l’exercice du pouvoir . . . ’
84
De Bakker 2007: 54–5, with further examples, and Létoublon 1987: 24. [Arist]. De Mundo 398b4–7
does indeed compare Xerxes to a god dispatching messengers.
 When a Letter and Why? 129

Bakker’s further remark that there is only one instance in the Histories of the use
of an intermediary for communication at close range between two Greeks is
significant.85 Even more significant is the fact that this unique instance is the
insulting message sent during the festival of the Gymnopaediae by Leotychidas,
through the intermediary of a slave, to Demaratus, after the latter’s deposition
from kingship (Hdt. 6. 67. 2). The implication is that the behaviour of Spartan
kings may at times be close to that of the Persians.
Secondly, it is important not to forget that there is almost no difference, either
in form or in function, between written and oral messages, and that in the fifth
century it is not yet possible to speak of an ‘epistolary genre’ recognized as such.
The remarkable closeness in the introductory formulae used by Herodotus for
both oral and written messages shows that he (and his audience) did not feel the
need for a specific language for the latter. Both this closeness and the existence of
numerous Greek letters on lead going back to the sixth century bc renders the
hypothesis of a Persian origin for the Greek epistolary prescript untenable.86
When compared to the relative diversity of the forms of address used between
Greek communities, however, the formula ‘A to B says thus’, used for oral and
written messages, seems to convey oriental connotations.

4.1.5. Letter Writing and Other Forms of Writing in Herodotus

The comparison between written and oral messages in the Histories has shown
that from a qualitative view-point the narrator does not distinguish markedly
between the two modes of conveying information at a distance. In Herodotus’
work there is no strong divide between written and oral communication; the
distinction is rather between public, polis-oriented communication and private
communication. The latter presents the deceitful, secretive, and despotic conno-
tations so well brought out by Steiner (1994) because it functions in the interest of
a small number of specific individuals. What is ultimately important is not the
medium of the communicative act, but its context (public or private).
A short review of the way other forms of writing function in the Histories will
confirm the above. Let us start with the inscriptions. Herodotus mentions twelve
Greek inscriptions, eleven non-Greek inscriptions, and a bilingual one (Greek and
cuneiform), dedicated by Darius. A comparison between this form of writing and
the letters brings out a number of differences, but also some similarities. To begin
with the differences: even if Herodotus may not have seen all the inscriptions he
mentions, he certainly saw most of them; and he could suppose that some of them
would be known to his public too. The inscriptions are not messages addressed
individually by one sender to one addressee; their function, eminently public-
oriented, is to perpetuate kleos or fame. Thus as individual objects (and texts) they
accomplish the same function as the Histories as a whole. This definitely does not
apply to the letters Herodotus mentions.87 Moreover, epigraphical documents are

85
De Bakker 2007: 55.
86
A conclusion reached already by van den Hout (1949: 29) and restated by Porciani 1997.
87
On the Histories as a monument see e.g. Moles 1999; the above makes it difficult to look at the
Histories as a letter.
130 Greek Beginnings

much more evenly distributed between Greeks and non-Greeks; for none of them
is there any question of secrecy, and for the Greek inscriptions at any rate, there
can be no question of tyrannical overtones.
And yet, for all these distinctions, inscriptions do not function very differently
from letters. Their writtenness is not a central element, just as the writtenness of
letters does not make them very different from oral messages. The best example of
this is offered by the first inscription mentioned in the Histories: the text engraved
on a golden perirrhanterion (a vessel for lustral water) originally dedicated at
Delphi by Croesus, an inscription proclaiming (falsely) that it was offered by the
Lacedaemonians (Hdt. 1. 51. 3). Herodotus’ brief comment (which makes for a
very awkward sentence, it must be said) puts this false document on exactly the
same level as a false oral statement: ‘it is inscribed (with letters) saying that it is of
the Lacedaemonians, not speaking correctly’ (KتªæÆÆØ ¸ÆŒÆØ  ø
çÆ ø r ÆØ IŁÅ Æ, PŒ OæŁH ºª)88: the error is in the statement, not
in its writtenness. However, the public inscription perpetuates a wrong statement,
just as the preservation of Herodotus’ work perpetuates his counter-statement.
Thus, even though inscriptions and letters are written documents, they both
present very close connections to the oral worlds of spoken kleos and oral
messages; the difference between these two types of writing is in the public status
of the first, as opposed to the private aspect of the second.
Another class of document that sits squarely between written and oral, public
and private, may be of interest: the oracular responses. The work of Herodotus is
evidence for the bringing back of an oral oracular answer (so in Athens, the
oracles of the Pythia), but also for the writing down by messengers of an answer,
which is then brought back to the initiator of the request (e.g. Croesus). The oracle
in itself functions always in the same way; it will be written down or transmitted
orally depending on the society, the group, or the individual who requires it. In
some extreme instances, such as in Athens in 480 bc, there will be a public
discussion of the very meaning of the oracle; at the other extreme are the
Pisistratids, possessors of written oracles that they keep in boxes. Again, a
‘message’ can be written or oral, and independently of whether it is written or
oral, it can be discussed publicly or kept secret.

4.2. MEDIATED LONG-DISTANCE

C O M M U N I C A T I O N IN T H U C Y D I D E S

While the proem of Herodotus’ Histories positions the work at the intersection
between writtenness and orality, Thucydides’ work presents itself, from the very
opening sentence, as written: ¨ıŒı Å #ŁÅÆE ıªæÆł e º ,

88
This is a problematic passage, and translators and commentators mostly gloss over the issue.
Madvig, followed by Legrand, proposed the following attractive correction, which would put the
perirrhanterion in the role of the speaking object, as in so many archaic inscriptions: KتªæÆÆØ
¸ÆŒÆØ  ø ç  r ÆØ IŁÅ Æ, PŒ OæŁH ºª (for full references see the apparatus and
commentary of Asheri 1988, ad loc.; and Flower 1991).
 When a Letter and Why? 131

‘Thucydides the Athenian wrote the war’.89 Here, we have an author who chooses
to communicate through writing. How does he present communication—and in
particular mediated communication—in his work? The War offers a number of
instances of mediated communication: besides the letters and the messages or
requests transmitted by heralds, a number of speeches are made by presbeis,
ambassadors representing their cities. Thus out of some 171 acts of communi-
cation, some sixty-nine represent mediated communication.90 Of these, thirty-
nine, in other words the majority, are speeches made by ambassadors; then, there
are short communications made by heralds (numbering thirteen, to which may be
added four instances in which messengers of unspecified status are sent); and
some thirteen instances of letters.

4.2.1. Ambassadors’ Speeches

We may wonder whether ambassadors’ speeches should be considered instances

of mediated communication: of course technically they are, in the sense that these
men are representing their city in another context; but in most instances the
æ Ø speak freely, as they think most appropriate, and using a first person
plural that encompasses themselves and their cities, not in any way giving the
impression that they are simply repeating a message. Thucydides’ narrative in the
first book offers one of the best examples of ambassadors’ independence: an
Athenian deputation who happened to be at Sparta for other business decided
to deliver a speech in the meeting convened by the allies to push the Spartans to
war.91 The formulae with which the speeches are introduced and closed tend to
reinforce this impression: in most cases, even though in the run-up to the speech
Thucydides had stated that an embassy would speak, the speakers in the para-
graphs introducing and resuming the speech are defined simply through their

89
On the importance of this opening sentence (and of writing in the War), see Loraux 1986; Moles
1999 (and Moles 1999: 47–8 for the Herodotean proem); Bakker 2006. On Thucydidean writing in the
context of the development of Athenian (and Greek) literacy, Crane 1996: 1–22; Goldhill 2002. I have
taken Alberti as the edition of reference for Thucydides; the translations are adapted from Crawley’s.
90
Numbers reached through integrating the list of West 1973: 7–15 with Scardino 2007: 387–94
and with those other references to communication that seemed pertinent (I have for instance, following
Bearzot 2003: 302–3, included kerygmata, as well as simple references to the sending of letters), and
then sorting out of the total the cases pertinent to mediated communication. Thus this is not exactly a
list of speeches, nor a complete list of all acts of communications. But the overall proportion of the
various types of ‘communicative acts’ is important; disagreement over details will not change the
overall picture. I have not included treaties; but I agree with the argument of Luschnat (1970: 1127–30)
that documents (i.e. letters, but also treaties) were considered by Thucydides as ºªØ, in the same
measure as speeches, and thus that documents are ‘nicht juristisches Beweismittel . . . sondern als
historisch-politischen Faktor, als reales dynamisches Element der Geschichte selbst betrachtet’. See
also the detailed argument of Porciani 2003: Thucydides considered the documents as logoi, and so
cited them intentionally, for reasons to be determined in each specific case.
91
Thuc. 1. 72. The Athenians ask the Spartans whether they can speak, and receiving assent, they
speak qua Athenians (ŒÆd ÆæºŁ ƒ #ŁÅÆEØ ºª Ø, 1. 72. 1); their speech, given in the
first person plural, closes with: ØÆFÆ b ƒ #ŁÅÆEØ r  (1. 79). Similarly, Themistocles in Sparta
(1. 91) acts as an individual and what he says is not mediated speech; at the same time he figures as part
of an embassy, and what is more, he delays presenting himself to the magistrates by saying that he must
wait for his colleagues (1. 90. 5).
132 Greek Beginnings

being citizens of a polis: the Corinthians, for instance, or the Corcyraeans.92 With
very few exceptions (one, Thuc. 1. 139. 3, is discussed below), in the context of
embassies the names of the individual speakers are not given. The one important
exception is the speeches of Hermocrates and Euphemus in Camarina (the latter
introduced as ›  ¯hçÅ  › H #ŁÅÆ ø æ ı , Thuc. 6. 81).93 In internal
debates, however, be they internal to Athens or to any other Greek city, the names
are mostly given, since the opinion expressed represents that of the individual
speaking (so e.g. also in the case of Teutiaplus’ speech in 3. 29). Now if Thucydides
thought he could keep close to the  Æ Æ ª! Å (the main thesis) of what had
actually been said (1. 22. 1, whatever the exact nature of that closeness), he
certainly could have given the name of the actual speaker in most instances: the
systematic omission in the case of embassies is, as so many other omissions,
significant—here, bespeaking the non-importance of the individual speakers: to
all intents and purposes the entire citizenship spoke.
The main task of the ambassadors consisted in persuading their audience on
behalf of the city; if they were successful, the polis would, retrospectively as it were,
acknowledge their words as its own.94 In some specific instances they did have full
powers, although of course always within limits, as is shown by an instance in
which negotiations fail, when the impression imposes itself that the Spartan
envoys do not have full powers (Thuc. 5. 45, where the technical expression
ÆPŒææ lŒØ, ‘to have come with full powers’, is used three times; Alcibi-
ades and private discussions play a major role in this failure).95
Whatever their exact powers in any given instance, the speech of envoys,
whether we consider that they aimed mainly at persuading with general argu-
ments preliminary to the drafting of an agreement or assume that they had full
powers, is not really—I suggest—mediated speech: rather, it is presented as the
independent speech of a collectivity. When individuals speak independently, on
their own account, they are named.
When, however, Thucydides wants to stress the fact that ambassadors are
simply conveying a decision already taken elsewhere, a different type of

92
Cf. 1. 36. 4 and 1. 44 (speeches of Corinthians and Corcyraeans); 1. 67 and 72 (Corinthians in
Sparta); 1. 72–9 (above, n. 91); 1. 90. 1–2 (indirect report of words of Spartan embassy); 1. 119–24
(Corinthians in Sparta); 2. 71. 1–72. 2 (Plataean embassy to Archidamus); 2. 72. 2 (reported speech of
the ambassadors of Plataea—and of the Plataeans—to Archidamus). Cf. Bers 1997: 8 on the normality
in Greek of couching statements about the activities of city-states in the form ‘the Athenians,
Corcyraeans, etc. did such and such’. And yet the question asked by Kennedy 1973: xi ‘Why does
the historian sometimes attribute speeches to specific individuals, sometimes to groups such as “the
Athenians”, “the Corinthians”?’ needs to be taken seriously.
93
Possibly by giving the names of individuals, Thucydies meant to imply that this ought to have
been an internal affair among the Siceliots. Alcibiades too speaks in Sparta, but by then he is speaking
on his own behalf (6. 88. 10–93. 1).
94
See Mosley 1973: 21–9 for a nuanced discussion of the margins of manoeuvre of the ambassadors.
The Plutarchan anecdote with which he opens his discussion (Plut. Lyc. 25. 4: the Spartan envoy
Polycratidas, when asked by the Persian generals ‘whether the embassy was there in a private or public
capacity, replied: “If we succeed, in a public capacity; if we fail, in a private” ’) may have actually been
applicable to many of the sixth- and fifth-century diplomatic exchanges.
95
On ÆPŒæøæ and its difference from Iı Łı see the excellent note in Gomme 1959: 426,
ad 1. 126. 8; on the powers of presbeis, ‘full’ only within understood limits, see Mosley 1973: 30–8 (for
this specific incident 31–2), and Hornblower 2008: 105. For the expression, compare Ar. Av. 1595.
 When a Letter and Why? 133

enunciation is used, as in the description of the last embassy of the Lacedaemo-

nians before the outbreak of the war reported in the first book, a solemn occasion
in which the names of the presbeis are indeed given:
º b IçØŒ ø H ºıÆ ø æ ø KŒ ¸ÆŒÆ ,  ) Æ ç ı  ŒÆd
ºÅ ı ŒÆd #ªÅ æı, ŒÆd ºªø ¼ºº b Pb z ææ N!Ł Æ,
ÆPa b  ‹Ø ¸ÆŒÆØ ØØ  ºÆØ c Næ Å r ÆØ, YÅ  i N f  ‚ººÅÆ
ÆP ı IçE, Ø Æ KŒŒºÅ Æ ƒ #ŁÅÆEØ ª! Æ ç Ø ÆPE
æı Ł Æ . . . (1. 139. 3)
Finally, as the last ambassadors from Lacedaemon arrived, namely Rhamphias, Me-
lesippus, and Hegesander, and as they said nothing of what they had used to say
earlier, but only this, that ‘The Lacedaemonians desire that there be peace; and there
may be, if you restore independence to the Hellenes’, the Athenians called an assembly
and held a discussion . . .
The lack of an address in the opening of the message, the avoidance throughout
of a first person plural, and the stress on the origin of the message (the Lacedae-
monians) makes it clear that here the presbeis are simply conveying the final
words of the Spartans, without leaving any space for personal interpretations or
additions (something already stated actually by the narrator in introducing them).
The possibilities outlined here are, however, only the extremities of a range that
offers possibilities for highlighting key points in the communicative (or non-
communicative) process. An interesting instance is offered by the negotiations of
the Lacedaemonian ambassadors sent to Athens to propose a peace and the
recovery of their men on the island (Thuc. 4. 16. 3–22. 3). The ‘preamble’ with
which the speech is introduced, as well as the opening of the speech itself, make it
clear that the envoys (presbeis) are speaking as negotiators on behalf of the
Lacedaemonians (‘the envoys were sent off. Having arrived at Athens, they
spoke as follows: “Athenians, the Lacedaemonians have sent us to negotiate for
the recovery of the men in the island.” ’, Thuc. 4. 16. 3–17. 1).96 The envoys keep
speaking in the first person plural, and referring to either past events or notions of
common wisdom, until 4. 19; at this point, there is a sudden switch to the third
person plural (the Lacedaemonians), concomitant with plans for peace and the
future:
‘The Lacedaemonians invite you (¸ÆŒÆØ ØØ b  A æŒÆºFÆØ) to make terms
with them and to finish the war, offering (Ø) peace and an alliance, and that
there be a general friendly and happy relationship, and requesting in return
(IÆØF) the men from the island. They think (ª Ø) it better that neither
should run any further risk, whether by some chance of safety they escape by force, or
being taken by siege they find themselves even more in your power. [2] We think
( Ç  ) that great enmities . . . ’ (Thuc. 4. 19. 1)
The switch to the third person is followed by a quick return to the first person
plural and to generalities on enmities, moderation, and advantageous policies.
This first person plural lasts until almost the very end of the speech, when there is
one final switch to the Lacedemonians as actors, followed by a return to the first

96
4. 16. 3: ƒ æ Ø I ºÅ Æ. IçØŒ Ø b K a #Ł Æ ºÆ Ø. 17.  łÆ  A
¸ÆŒÆØ ØØ, t #ŁÅÆEØ, æd H K B
fi  fi ø IæH æÆ . . .
134 Greek Beginnings

person plural, paired with a second person plural to indicate the Athenians (‘If
you realize this, it is possible for you to become firm friends with the Lacedae-
monians at their own invitation, which you do not force from them, but oblige
them by accepting. [4] And from this friendship consider the advantages that are
likely to follow: for if we and you speak the same, the rest of Hellas, you know, will
have the greatest respect, being inferior’, Thuc. 4. 20. 3–4);97 the narrative restarts
with the narrator’s comment ‘so spoke the Lacedaemonians’ (ƒ b s
¸ÆŒÆØ ØØ  ÆFÆ r ). When the Athenians, persuaded by Cleon, refuse
to accept the proposals, the envoys (so the narrator reports) ‘asked that commis-
sioners might be chosen with whom they might confer on each point, and quietly
talk the matter over and try to come to some agreement’—and were attacked for
this by Cleon, who suggested that the refusal to speak in front of the people and
the request for private discussion betrayed their bad intentions. Clearly the envoys
had some freedom of action; both this latitude and the very real backing of the
Lacedaemonians are conveyed during their speech to the assembly through the
switches between first and third person plural.
Once, in 5. 61. 2–3, a speech is presented by the narrator as given by the city
through one specific person:
ŒÆd ºª ƒ #ŁÅÆEØ #ºŒØØı æ ıF Ææ   E #æª Ø ŒÆd E
ı åØ ÆP, ‹Ø PŒ OæŁH ƃ Æd ¼ı H ¼ººø ı åø ŒÆd ªØ
. . . [3] ŒÆd  Æ . . .
The Athenians, through Alcibiades their ambassador there present, told the Argives
and the allies the same, that they had no right to make a truce at all without the
consent of their fellow-confederates . . . [3] These arguments proving successful . . .
But this is a unique instance, and it may actually point to the fact that
Alcibiades is here exceeding his instructions;98 in most cases, it is the citizenry
that speaks. If we discount the examples of what is effectively the city speaking as
not being instances of mediated communication stricto sensu, we are left with a
few cases of envoys relating words of another city or group, heralds transmitting
oral messages, and with letters. The Plataean embassy to Athens offers a good
example of the first possibility:
KºŁ b ƒ —ºÆÆØB æ Ø ‰ f #ŁÅÆ ı ŒÆd ıºı  Ø  ÆPH
ºØ qºŁ Iªªºº E K B
fi ºØ Ø: [3] h K fiH æe F åæø
fi , t ¼æ
—ºÆÆØB, Iç y  ÆåØ Kª ŁÆ, #ŁÅÆE çÆ Ø K Pd  A æ ŁÆØ
IØŒı ı h F æØł ŁÆØ, ÅŁ Ø b ŒÆa  Æ Ø. KØ Œ ı   E
æe H ‹æŒø o ƒ Ææ þ  Æ Åb øæ ÇØ æd c ı Æå Æ. 74.
ØÆFÆ H æ ø Iƪªغø ƒ —ºÆÆØB Kıº Æ #ŁÅÆ ı c
æØÆØ . . . (Thuc. 2. 73. 2–74. 1)
The Plataean envoys came to Athens, and after deliberating with the Athenians they
brought back the following message to their fellow-citizens: ‘Plataeans, the Athenians

97
X  ªH, ¸ÆŒÆØ  Ø  Ø  E ç ºı ª ŁÆØ Æ ø, ÆPH  æŒÆº Æ ø
åÆæØ Æ Ø  Aºº j ØÆ Æ Ø. ŒÆd K  øfi a KÆ IªÆŁa ŒE ‹ Æ NŒe r ÆØ:  H ªaæ
ŒÆd  H ÆPa ºªø  ª ¼ºº  EººÅØŒe Y  ‹Ø  æ k a ªØ Æ Ø Ø.
98
For Fowler 1888: 192, this shows that Alcibiades is now acting as a private individual; he
compares with the explicit D. S. 12. 79: NØ!Å þ.
 When a Letter and Why? 135
say that neither in the previous time since we became allies have they allowed you to be
wronged, nor will they now, but will assist you to the utmost of their power; and they
adjure you, by the oaths which your fathers swore, not to forsake the Athenian
alliance.’ 74. When the answer came, the Plataeans resolved not to desert the
Athenians.
The message is reported here throughout in the third person: the Plataean
envoys are stating what the Athenians said, without adding anything of their own.
As for the heralds, they are, unsurprisingly, mostly portrayed as literally reporting
the words of the group they represent. In quite a few instances, the ‘transparency’
of the heralds is such that the messenger disappears, as in Herodotus: so for
instance in 1. 53. 1–3, a passage in which the Corinthians at Sybota
put some men on board a boat, and sent them without a herald’s wand (¼ı ŒÅæıŒ ı)
to the Athenians, to test them. Having sent them, they spoke as follows ( łÆ 
ºª Ø): [2] ‘You do wrong, Athenians (IØŒE, t ¼æ #ŁÅÆEØ), to begin
war and break the treaty. Engaged in chastizing our enemies, we find you placing
yourselves in our path in arms against us. Now if your intentions are to prevent our
sailing to Corcyra, or anywhere else that we may wish, and if you are in favour of
breaking the treaty, first take us that are here, and treat us as enemies’ ( A  
æ!ı ºÆ åæ Æ Ł ‰ º Ø). [3] Such was what they said (ƒ b c
ØÆFÆ r ), and as much of the Corcyraeans’ army as were within hearing immedi-
ately called out to take them and kill them. But the Athenians answered as follows . . .
So also more than once in the discussions between Plataeans and Thebans at
the very beginning of book 2, where the same strategy is applied in two slightly
different ways within the space of a few paragraphs. In one instance, the herald
emerges out of the collective, is given specific individual words, and then quickly
(with an anacoluthon) disappears again into the collectivity:
ª! Å  KØF ŒÅæ ª Æ  åæ Æ ŁÆØ KØÅ Ø ŒÆd K  Æ Ø Aºº ŒÆd
çغ Æ c ºØ IªÆªE, ŒÆd IE › ŒBæı, Y Ø  ºÆØ ŒÆa a æØÆ H
ø BØøH ı ÆåE,  Ł ŁÆØ Ææ Æf a ‹ºÆ,  Ç ç Ø ÞÆ fi  ø
 ø
fi fiH æø
fi æ åøæ Ø c ºØ (Thuc. 2. 2. 4)
(The Thebans) determined to make a conciliatory proclamation, and if possible to
bring the city to a friendly understanding; and the herald proclaimed, that if any
wished according to the ancient custom of the Boeotians to be their ally, they should
put their arms with them, thinking that in this way the city would readily join them.
In another instance, the herald is mentioned only to be ‘cancelled’ (thus
avoiding the anacoluthon):
ƒ b —ºÆÆØB . . .  Æ æd E ø Œ æıŒÆ K łÆ Ææa f ¨ÅÆ ı,
ºª ‹Ø h a ØÅ Æ ‹ ØÆ æ ØÆ K ÆE çH ØæÆŁ
ŒÆƺÆE c ºØ,   ø ºª ÆPE c IØŒE: N b , ŒÆd ÆPd çÆ Æ
ÆPH f ¼æÆ IŒE o åı Ø ÇHÆ. (Thuc. 2. 5. 5)
But the Plataeans fearing for their fellow-citizens outside the town sent a herald to the
Thebans, saying that they had acted outrageously in trying to seize their city in time of
peace, and warning them against any outrage on those outside; otherwise, they
themselves said that they would to put to death the men they had in their hands.
Ironically, in a world where writing has become much more important,
‘scripted’ speech seems overall to have shrunk. The norm in the Peloponnesian
136 Greek Beginnings

War, especially in its first part, is either face-to-face open speech, in which a
named citizen speaks to his fellow citizens or entire (anonymous) citizen bodies
engage in discussion with each other; or, less frequently, formalized communi-
cation through heralds. Divergent attitudes (mostly concentrated in the second
part of the work) are flagged up as dysfunctional.99 Instances of dysfunctional
attitudes towards communication of course abound: they comprise all those cases
in which someone suggests that discussion take place among a restricted group
(e.g. 4. 22. 1–3; 5. 27. 2; 5. 36; 5. 37; 5. 43. 3; the Melian dialogue)—as well as
letters. This dysfunctional attitude is most notably present in the epistolary
exchanges of the first book, which go back to the time immediately following
the Persian wars, and then again markedly in the fifth and eighth books—that is,
in the second part of Thucydides’ work—possibly to highlight the contradictions
and changes in the normal diplomatic practice brought about by the war.

4.2.2. Letters in Thucydides

For all the differences between Herodotus and Thucydides in what concerns the
formal aspects in the presentation of oral messages, the conclusions reached on
the uses of letter writing in Herodotus are supported by the picture of epistolary
communication offered by Thucydides’ work. Thematically, quite a few of these
letters have to do with intrigue and death. If we look at senders and addressees, out
of thirteen probable references to letters, five concern exchanges between a Greek
and a Persian, a very high proportion for a work much more focused than the
Histories on the Greek world.100 The remaining eight letters reflect communi-
cation internal to the Greek world, and more precisely, communication between
Athenians, between Spartans, or between Spartans and Athenians. An epistolary
thread thus links the three powers fighting for the hegemony.
As for the relation between written and oral communication, many—but
definitely not all—of the Thucydidean instances of epistolary exchange could
have been replaced without problems by an oral exchange. A comparison between
letters and the other speech-acts in the War highlights the following features.101

99
I develop here in another direction the Thucydidean theme of the comparison between cities and
the individual, sketched by Morrison 1994: if Thucydides is fond of comparing cities and individuals, it
is indeed because cities do behave as individuals; but when individuals and not the city predominate,
disaster looms. Although made from a different angle and aiming at a different conclusion (an
evolution in Thucydides’ style and in his approach to the question of the importance of individuals
in history), observations on Thucydidean diplomacy close to those sketched above are in Westlake
1970–1, who emphasizes the fact that in the second part of the War Thucydides includes ‘far more and
much fuller information about diplomacy conducted by very small groups, often in secret and
sometimes with wholly negative results’ (228); see esp. 232–3 for anonymous ambassadors’ speeches,
and 235–9 for negotiations in which individuals play an important role.
100
Momigliano 1966: 817 remarks that it is curious that out of the eleven documents cited by
Thucydides five concern Persian affairs (the letters between Xerxes and Pausanias, and the three
versions of the agreement between Sparta and Persia of 411 bc), and wonders whether Thucydides
may here have followed the example of an Ionian historian, who might have derived his attention to
documents from the Persians themselves.
101
What follows is based on the first half of the table of all speech-acts in the War drafted by
Luschnat (1970: 1163–6). The other half is not pertinent, since a letter is usually a document sent by an
individual and not by a multitude (although in theory a polis could have sent a letter).
 When a Letter and Why? 137

In general, in Thucydides addresses by individuals to individuals are much rarer

than addresses to large numbers; within this overall frame, letters may appear as
indirect speech addressed by one person to another (for example, 8. 50. 2,
Phrynichus to Astyochus, comparable to the exchange in indirect speech between
an Athenian ally and one of the Spartiate prisoners from Sphakteria in 4. 40. 2),
and as indirect speech addressed by one to many (4. 50. 2, letter of Artaxerxes to
the Spartans, comparable to a number of addresses in indirect speech by individ-
uals, be they heralds, generals, or politicians, to groups). Letters also appear as
direct speech addressed by one person to another individual (1. 128. 7 and 1. 129. 3,
the letter sent by Pausanias to Xerxes and Xerxes’ answer), while instances of
direct, face-to-face speech between two individuals are much more difficult to
find.102 Finally, both letters and ‘speeches’ can combine direct and indirect speech:
the letter of Themistocles to Artaxerxes (1. 137. 4) is comparable to the dialogue
between the Ambraciote herald and an Athenian (3. 113. 3). On the whole, then,
and from a formal point of view, in Thucydides’ work letters function as oral
exchanges would.
Before discussing the letters themselves, it is necessary to address the question
of the dislocation of these documents across the work. Four of the five references
to epistolary communication between Greeks and Persians come from the digres-
sions on Pausanias and Themistocles in Thucydides’ book 1, digressions narrating
events of the early seventies; it has been often felt that these digressions are in a
distinct, fairly different narrative style from that of the rest of the work.103
Another group is formed by the seven letters mentioned in book 8: this is
notoriously an unfinished book (unfinished at any rate in the sense that it ends
in mid-sentence), a book moreover including, besides these references to letters, a
number of written documents—and no direct speeches. These features of the
eighth book have given rise to unending discussions as to the shape it might have
had if polished to completion.104 The concentration of letters in some specific
parts of Thucydides’ work may mean that we should reassess the importance
attributed by the historian to this type of document. And yet, while keeping in

102
Luschnat suggests the address of Brasidas to Clearidas in the context of his speech at Amphi-
polis, 5. 9. 7 (and, rather oddly, the epigram of Pausanias to Apollo, 1. 132. 2), but in both cases there is
a larger implied audience, and this is emphatically not—or should emphatically not—be the case with
letters, although they do become public.
103
The high number of letters in book 1 is one of the reasons for Westlake’s hypothesis (1977: 102–3)
that there was a written source for the excursuses on Themistocles and Pausanias. The reason
highlighted by an ancient scholiast, commenting on 1. 126–7: ‘Some, admiring the perspicuity of the
narrative about Cylon, have said: “here the lion smiled” ’, i.e. a different narrative style, is possibly the
most convincing argument. See Gomme 1959: 446–7; Hornblower 1991: 211–12.
104
Overview of the history of the question in Hornblower 2008: 1-4 and 883–6; see also on the
status of documents Porciani 2003 and Bearzot 2003 (ample review of all types of documents present in
Thucydides). Luschnat 1970: 1112–32 offers the best, most sensible, and at the same time most concise
discussion of the question of the completeness of the work, with all that this involves concerning
speeches and documents; he shows very clearly that homogeneity was not necessarily a Thucydidean
concern, and hints at the fact that the different narrative styles are adjusted to the requirement of
narrating different historical periods and events (a point further developed by Rood 1998). Lane Fox
2010 makes a plausible case for Lichas being the channel through which many documents reached
Thucydides; but he does not explain satisfactorily the distribution of the documents within the work,
and dismisses too lightly issues of composition and meaning.
138 Greek Beginnings

mind that there are some difficulties, the best way of proceeding must be to start
from the War as we have it, and to take these differences in narrative style not as
marks of incompleteness, but as potentially significant.105
Let us then start from the epistolary exchanges between Persians and Greeks.
Not only are these closer in time to the Herodotean letters (both in their historical,
extradiegetic time, since the events narrated belong to the early seventies, and in
their narrative time, since in this part of the War Thucydides is bridging the
interval between the end of the Histories and the beginning of the Peloponnesian
war—moreover, this might be an early part of Thucydides’ work); they also mirror
some of their main themes: deception, despotism, deadliness. In his excursus on
Pausanias, Thucydides quotes the text of a letter (KØ º , 1. 128. 6) sent by
Pausanias, the Spartan regent, to Xerxes (Thuc. 1. 128. 7); he gives a verbatim
translation of Xerxes’ answer (1. 129. 3); and he mentions the interception by the
ephors of another letter sent by Pausanias to Xerxes, of which he gives a para-
phrase (1. 132. 5).106 Discussion of these documents has tended to concentrate on
their authenticity, and on their origin; I suggest looking rather at their ‘shape’ and
their function, in the first book and in the work as a whole.
The context of the first letter is as follows: at the time of the capture of
Byzantium, Pausanias sent the king some relatives whom he had taken prisoner,
saying to the rest of the Greeks that they had escaped; the captives were accom-
panied by an Eretrian, Gongylus, to whom Pausanias entrusted a letter:
 ł b ŒÆd KØ ºc e ˆªªØº çæÆ ÆPfiH· KªªæÆ b  K ÆPB
fi , ‰
o æ IÅıæŁÅ: [7] —Æı Æ Æ › ª g B æÅ    Ø åÆæ Ç ŁÆØ
ıº  I Ø æd º!, ŒÆd ª! Å ØF ÆØ, N ŒÆd d ŒE, ŁıªÆæÆ 
c c ªB ÆØ ŒÆ Ø æÅ  ŒÆd c ¼ººÅ  EººÆ å æØ ØB ÆØ. ıÆe
b ŒH r ÆØ ÆFÆ æAÆØ a F ıºı . N s    ø Iæ ŒØ,  
¼æÆ Ø e Kd ŁºÆ Æ Ø y e ºØe f ºªı ØÅ  ŁÆ. 129.  ÆFÆ b 
ªæÆçc K ºı, ˛æÅ b l ŁÅ  Bfi KØ ºBfi . . . (Thuc. 1. 128. 6–129. 1)
He also sent to him Gongylos carrying a letter; this was written in it, as was afterwards
discovered: [7] ‘Pausanias, the general of Sparta, desiring to do you a favour, sends you
these, his prisoners of war. I propose also, if it is agreeable to you, to marry your
daughter, and to make Sparta and the rest of Hellas subject to you. I think I am able to
do this, with your co-operation. Accordingly if any of this pleases you, send a
trustworthy man to the sea through whom we may in future conduct our correspond-
ence.’ 129. The writing revealed as much, and Xerxes was pleased with the letter.
The letter begins in the third person, and goes straight to the point, omitting
any formal address or salutation. Already in the second sentence there is a shift to

105
Rood’s (1998) position: to see whether it is possible to make sense of the different narrative styles
in terms of different narrative objectives, and thus to first try to account for what we have on its own
terms. In a number of instances this works.
106
On the questions raised by these letters, and by the Thucydidean tradition on Pausanias, see
Nafissi 2004, with further bibliography. Clearly, these documents are fakes, possibly created in Sparta to
justify a posteriori the condemnation of Pausanias (so Nafissi); alternatively, they could have been
forged in Sparta before the death of Pausanias, or they might be an Athenian falsification going back to
the years of the foundation of the Delian league. Gauger’s hypothesis (2000: 263–4), that these are
authentic letters, rendered public by the Spartans immediately before the outbreak of the Pelopon-
nesian war, or also after 404, is unconvincing.
 When a Letter and Why? 139

the first person, with the rather direct request to marry Xerxes’ daughter, and the
hyperbolic offer of making Sparta and all of Greece subservient to Persia. The
letter closes on a first person plural, encompassing sender and addressee, Pausan-
ias and Xerxes (rhetorically, an astute move). The letter’s inauthenticity can hardly
be doubted: Pausanias might have kept a copy of Xerxes’ letter, but his own letters
would have been impossible to retrieve; moreover, the tone of the letter is unlikely
to reflect what Pausanias might possibly have written; the ‰ o æ IÅıæŁÅ
must refer to later oral reports of the presumed content, or to the alleged discovery
of an (inauthentic) letter; in the same direction goes the narrator’s summary.
Finally, the reaction of Xerxes (l ŁÅ) is, as pointed out by Gomme, typically
Herodotean—and of course an inference.107 To this, Xerxes answers by sending a
trusted man, Artabazus, with a royal signet and a letter:
c KØ ºc Ø ł· IªªæÆ b : z ºªØ Æ Øºf ˛æÅ
—Æı Æ fi Æ. ŒÆd H IæH o Ø æÆ ŁÆº Å KŒ BıÇÆ ı  ø Æ Œ Æ
Ø Pæª Æ K fiH  æøfi YŒø
fi K ÆNd IªæÆ, ŒÆd E ºªØ E Ie F
Iæ Œ ÆØ. ŒÆ   f Ł  æÆ KØ åø u  IEÆØ æ Ø Ø z K d
Ø åE, Åb åæı F ŒÆd Iæª æı ÆfiÅ ŒŒøº Łø Åb æÆØA º ŁØ, Y Ø E
Ææƪ ª ŁÆØ, Iººa  #æÆÇı Iæe IªÆŁF, ‹ Ø  łÆ, æA  ŁÆæ H ŒÆd
a K a ŒÆd a a ‹fiÅ ŒººØ Æ ŒÆd ¼æØ Æ (Ø I çæØ. (Thuc. 1. 129. 3)
He sent over the letter, which contained the following answer: ‘Thus says King Xerxes
to Pausanias. For the men whom you have saved for me across the sea from Byzantium,
an obligation will be laid up for you in our house, recorded forever; and with your
proposals I am well pleased. Let neither night nor day stop you from diligently
performing any of your promises to me, neither for cost of gold nor of silver let them
be hindered, nor yet for number of troops, wherever it may be that their presence is
needed; but with Artabazus, an honourable man whom I send to you, boldly advance
my objects and yours, as may be most for the honour and interest of us both.’
The letter opens with the type of address that had been used in Herodotus’ work
to introduce letters by kings or tyrants, and that reflects the actual formula used by
Persian kings; some of the expressions used in the body of the letter find parallels
in oriental expressions.108 This in itself cannot be used as a proof of authenticity:
Herodotus’ work shows that the form of address and the language typical of an
oriental king were common knowledge. If the letter is authentic, it shows that
exchanges in epistolary format took place between Persians and Greeks; if it is a
false document produced to incriminate Pausanias, its appropriateness in terms of
language demonstrates a remarkable familiarity with Persian epistolary practices,
a familiarity necessarily based on other letters and exchanges. As a result, the
question of the authenticity of this specific letter loses much of its importance

107
Gomme 1959: 432. Gomme highlights the difficulty of procuring the original of Pausanias’ letter
to Xerxes (and of Themistocles’ to Artaxerxes); but he also adds that, by using  rather than Ø,
Thucydides might mean that he is quoting word for word. One could object that the  ÆFÆ (rather
than ÆFÆ, and moreover in conjunction with źF) summarizing the letter speaks for the opposite
assumption. Hdt. 5. 32 reports (doubtfully: N c IºÅŁ  ª K Ø › ºª) that Pausanias was betrothed
to a daughter of the noble Persian Megabates.
108
See Gomme 1959: 432, who stresses the parallelism with both the beginning and the conclusion
(ll. 15–17) of the letter of Darius to Gadatas (M.–L. no. 12), and with Hdt. 8. 85. 3; also Hornblower
1991: 215–16.
140 Greek Beginnings

from the point of view of the evolution of the epistolary form (of course not for
what concerns internal Spartan politics and the fate of Pausanias).
At any rate, the letter (this time, a ªæ ÆÆ, 1. 130. 1) has a dire effect on
Pausanias, who begins to live in a grand style, to dress with a Median dress, to be
attended by a bodyguard of Medes and Persians . . . and to make himself difficult
of access.109 As a result, the Lacedaemonians recall him a first time; Pausanias is
acquitted, but the Lacedaemonians do not give him any further public office.
At this point, on his own initiative Pausanias sails again towards the Hellespont
(Å  fiÆ b PŒØ K çŁÅ, N fiÆ b . . . , 1. 128. 3), and once again lays himself
open to suspicion. As a result, the ephors send a herald with a skytale, ordering
him to follow the herald, or else the Spartans would declare war against him
( łÆ Œ æıŒÆ ƒ çæØ ŒÆd ŒıºÅ r  F Œ æıŒ c º  ŁÆØ, N b
, º  ÆPfiH ÆæØÆ æƪæ Ø, 1. 131. 1). Pausanias returns and is
thrown in prison, but after a while manages to get out. For the Spartans, Thucydi-
des says, had no clear evidence (çÆæe . . . Å E) against him, although his
behaviour both in the past (here Thucydides mentions the epigram on the tripod
at Delphi) and in the present continued to cause offence, and although there were
rumours of his conspiring with the helots.110 It is finally a stratagem devised by
Pausanias to guarantee his own security that causes his undoing: a messenger, a
man of Argylus who had once been a favourite of Pausanias, chosen to bring a
letter of his to Artabazus, noticed that none of the previous messengers had ever
come back; after preparing a counterfeit seal, so that he could reseal the letter if
necessary, he opened it, found that it contained the order to put him to death, and
showed it to the ephors. Interestingly, this last part is introduced with a qualifica-
tion (‰ ºªÆØ, 1. 132. 5, reappearing, always apropos of Pausanias, in 1. 134. 1);
and the text of the letter is not given (the only part of the content summarized
refers to the order of putting the messenger to death). And yet, this is the one letter
that could most easily have been produced . . . 111
In the excursus on Themistocles, closely linked to the preceding, since the
ephors are prompted to accuse Themistocles in front of the Athenians because of
facts discovered during their enquiry on Pausanias, Thucydides partly quotes,
partly reports in indirect speech, the text of another letter (ªæ ÆÆ, 1. 137. 3),
allegedly sent by Themistocles to Artaxerxes. This is once again a very interesting
document, both in itself, and fr the way it is introduced in the War:

109
ı æ   Æe ÆæEå, reinforced by the following sentence: he treats everyone with
such temper that no one dares to come near, B fi OæªBfi oø åƺB fi KåæB K Æ ›  ø u 
ÅÆ  Æ ŁÆØ æ ØÆØ, 1. 130. 2. One is here strongly reminded of the invention by Atossa of
epistolary writing—aimed at imposing a distance—and of the rise to the Median throne of Deioces,
Hdt. 1. 99–100 (see above, 87–9, as well as below, 233). The signet ring does not play any role in the
story; possibly it was simply felt to be one of the necessary ingredients.
110
The lack of proof at this point (Thuc. 1. 132, after Pausanias has been for a time in prison) is
another element for the inauthenticity of the previous two letters.
111
What follows is also not without problems: in the conversation arranged by the ephors, who
wish to overhear, the man of Argylus affirms to have served Pausanias well in his previous negotiations
with the king—and yet all previous messengers had been put to death. Gomme 1959: 436 attempts to
solve the problem by suggesting that earlier negotiations in the Troad are meant, but he also mentions
Steup’s solution, to bracket æe Æ ØºÆ and to understand negotiations with the helots.
 When a Letter and Why? 141
[4] K ºı b  ªæÆçc ‹Ø ¨ Ø ŒºB lŒø Ææa , n ŒÆŒa b ºE Æ  Eºº ø
YæªÆ ÆØ e  æ r Œ, ‹  åæ e e ÆæÆ KØÆ K d IªŒfiÅ
M ı Å, ºf  Ø º ø IªÆŁ, KØc K fiH I çƺE b K  , KŒ ø fi b K
KØŒØ ø
fi ºØ  IŒ Øc Kª ª. ŒÆ Ø Pæª Æ Oç ºÆØ (ªæłÆ   
KŒ ÆºÆ E 檪º Ø B IÆåøæ ø ŒÆd c H ªçıæH, m łıH
æ Ø Æ,  Ø Æe P غı Ø), ŒÆd F åø  ªºÆ IªÆŁa æA ÆØ
æØ Ø ØøŒ  e H  Eºº ø Øa c c çغ Æ.  º ÆØ  KØÆıe KØ åg
ÆP Ø æd z lŒø źH ÆØ. (Thuc. 1. 137. 4)
The writing stated that ‘I, Themistocles, am come to you, who did your house more
harm than any of the Hellenes when I was compelled to defend myself against your
father’s invasion, but far more good during his retreat, which brought no danger for
me but much for him. For the past, you are a good turn in my debt,’—here he
mentioned the warning sent to Xerxes from Salamis to retreat, as well as his finding
the bridges unbroken, which, as he falsely pretended, was due to him—‘for the present,
able to do you great service, I am here, pursued by the Hellenes for my friendship for
you. I desire, having waited one year, to declare in person the reasons of my coming.’
The letter is throughout written in the first person, without any form of
greeting, and it mixes direct and indirect speech, a fact that might betray the
reworking of a bare summary.112 From the point of view of the structure of the
passage, it is interesting to note that the same terminology recurs in the narrator’s
introduction and in the conclusion of the letter: K ºı b  ªæÆçc ‹Ø lŒø  lŒø
źH ÆØ: the writing does the initial ‘showing’, to be completed orally by Themis-
tocles himself. Of course, the same problem arises here as for the letter of
Pausanias to Xerxes: how could a letter addressed to the King of Persia have
entered the public domain? Unsurprisingly, when the narrative resumes, we find
another qualifying ‰ ºªÆØ, ‘it is said’ (1. 138. 1). More interesting is the fact
that the letter refers to the two (oral) messages that, according to Herodotus, were
sent to Xerxes by Themistocles through the intermediary of Sicinnus (Hdt. 8. 75
and Hdt. 8. 110. 3).113 Stories concerning Themistocles will have begun to
circulate early on; they will have given a large place to his communicative skills,
whether in giving speeches, in sending a messenger with a deceitful oral message,
or in sending a letter. There is a certain symmetry in having the two heroes of the
Persian wars end in parallel disgrace, both writing letters to the Persian king—a
symmetry that may have played a role in the formation of the tradition.114
To these four instances of epistolary exchanges, all involving Persia, should be
added the reference to letters (KØ º) written ‘in Assyrian characters’ (i.e., in
this context, in Aramaic), sent by Artaxerxes through the mediation of the Persian
Artaphernes to the Spartans, but intercepted by the Athenians in the winter of

112
Cf. Gomme 1959: 440–1; Gomme’s further comment about the move in the later romanced
stories on Themistocles from a letter to a speech points to something potentially important; but his
conclusion (that the speech cannot have been authentic, but the letter, in substance, might be [my
italics]) seems to me off the mark.
113
Gomme 1959: 440–1.
114
For the diffusion of sympotic songs concerning heroes of the Persian war, and Themistocles in
particular, see e.g. the poem of Timocreon (727 PMG) with Stehle 1994. A literary source, and in
particular the names of Charon and of Stesimbrotus, has been suggested as the source for this part of
Thucydides’ narrative—they would explain what is felt to be the ‘Ionian’, Herodotean style of this part.
See Westlake 1977; Bearzot 2003: 285; Ceccarelli forthcoming (c), BNJ 262 (Charon of Lampsakos) T1,
F1, and Biographical Essay.
142 Greek Beginnings

424 bc (Thuc. 4. 50. 2: ƒ #ŁÅÆEØ a b KØ ºa ƪæÆł Ø KŒ H
# ıæ ø ªæÆ ø Iªø Æ).115 The content of the letters is simply sum-
marized, as is reasonable, since this exchange and its interception did not have any
impact on this part of the war; the main point was, according to Thucydides, that
the King, having received a number of Spartan envoys, none saying the same
thing, now asked them to send someone with definite proposals in the company of
Artaphernes. The Athenian reaction was themselves to send an embassy to Persia,
which, however, turned round after reaching Ephesus on the news of the death of
Artaxerxes. Athenians and Spartans may have had with them letters or not; but
the presence of letters, together with an envoy, is stressed only for the Persian side.
As for the remaining letters, they include the famous letter sent by Nicias to the
Athenians, reported in its entirety in direct speech (Thuc. 7. 11–15); and a group
of seven letters, all written in the winter of 412/411, and all mentioned in the
eighth book.
Nicias’ letter is a fascinating document; it has been discussed more than once.
This is probably not ‘the’ letter written by Nicias; it was composed by Thucydides
in the form of a speech. The epistolary status of this message was discussed already
by ancient critics: Dionysius of Halicarnassus (Thuc. 42) calls it KØ º , a letter,
but mentions it in the course of a discussion of Nicias’ assembly speeches; the
author of the treatise De elocutione includes it in its condemnation of those letters
that, because of their excessive length, end up being treatises rather than letters.116
Ancient theorists insist that letters should be short and to the point; however, their
parameters, including the norm about length, may not have applied to the fifth-
century notion of an KØ º : the long, detailed, and impersonal report of a
general, sent over a distance and giving advice, might well have been called
epistole. Some indications may come from the syntagms źF Æ Ø (7.
10), with which the letter is introduced, and  b F ˝ØŒ ı KØ ºc  ÆFÆ
K ºı (7. 16), indicating the return to the narrative. In presenting the letter,
Thucydides combines the pronoun used to introduce speeches, Ø (a pronoun
not used to refer to letters or written texts, inscriptions, or treaties, which are
usually introduced with a form of ‹), and a form of źø, the verb used for the
letters of Pausanias and Themistocles, but usually not for speeches; when resum-
ing the narrative, źø appears again, and with  ÆFÆ.117 Thus it is reasonable
to assume that Nicias’ own letter will have had a different shape; the one given in
the War is a Thucydidean creation.118

115
See Hornblower 1996: 205–7 for an updated discussion of the letter itself, and of the chrono-
logical problems linked to the following Athenian embassy and the death of Artaxerxes. This is one of
the rare references of Thucydides to Persia (discounting book 8), and it is difficult to contextualize it:
should we imagine that such exchanges were a relatively frequent thing? Would Thucydides, once he
realized the importance of Persia in the later course of the war, have added precisions?
116
Demetr. Eloc. 228. Cf. Van den Hout 1949: 37–8; Rosenmeyer 2001: 58–9.
117
: cf. 1. 128. 6, 1. 129. 3 (letters), 3. 104. 5 (poetry; cf. 4. 104. 4, Thucydides’ own work), 5. 17. 2, 5.
77. 1, 5. 78 (negotiations and treaties); see 1. 139. 3, the embassy of the three named Spartan heralds,
discussed above, 133; and 3. 29. 2, the speech of Teutiaplos, also exceptionally named. Note, however, that
when resuming the narrative after a letter,  ÆFÆ (used for speeches as an alternative to Ø) is found:
128. 7:  ÆFÆ b  ªæÆçc K ºı. For źø with a letter, see also 137. 4: K ºı b  ªæÆçc ‹Ø.
118
See Dover, in Gomme, Andrewes, and Dover 1970: 385: ‘the real letter may have contained many
more detailed facts and figures’. Hornblower 2008: 557–68 calls it a ‘remarkable piece of generic
 When a Letter and Why? 143

This letter may have been an early instance of a communicative typology that
will be better attested later: the reports sent by campaigning strategoi to the
polis.119 Nicias himself refers, in his ‘letter’, to previous communications, al-
though not necessarily written ones (a b ææ æÆåŁÆ, t #ŁÅÆEØ, K
¼ººÆØ ººÆE KØ ºÆE ¥ , 7. 11. 1).120 And yet, how frequent will the
sending of letters back home have been? And how did Thucydides look upon it?
Thucydides stresses at length Nicias’ special reasons for writing: ‘he feared that
the messengers, either through inability to speak, or through failure of memory, or
from a wish to please the multitude, might not report the truth’, in other words,
they might not be able to describe the situation with the precision and accuracy
required; he thus ‘thought it best to write a letter, to ensure that the Athenians
should know his own opinion without its being lost in transmission, and be able to
decide upon the real facts of the case (æd B IºÅŁ Æ)’.121 The detailed
presentation of Nicias’ thoughts (note the extraordinary expression ª! Å . . . K
fiH Iªªºøfi IçÆØ ŁE Æ, ‘opinion . . . lost in transmission’, but literally ‘lost in the
messenger’) tells something about the respective status of public speech and
written communication: for Nicias, an assembly speech is a dialogue between
messengers and demos; it may afford a better understanding of the situation, but it
may also lead astray. A letter instead represents the point of view of the sender,
fixed forever with precision; it is—Nicias thinks—closer to the truth. The very
presence of such a lengthy explanation underlines the peculiarity of the choice of
sending a written message.122 Hence the impression that Nicias and Thucydides
may here have just made a discovery, and that they may be explaining it to the
public.123
But at this point it is necessary to ask: What is the ‘sign’ of this discovery? Nicias
of course thinks that writing a letter will allow a decision based on a better, more

experimentation. Nowhere else does Thucydides try to present both a narrative and an expression of a
sustained argument, both of them focalized through an individual’ (557). For Canfora (1990: 198)
Thucydides may have paraphrased or reworked a document that he had personally consulted.
119
Instances of communication between generals and the polis are discussed below (for Sparta),
and in chs. 6 (266–7, 275–6, 280) and 7 (317, 326; cf. Appendix 3 no. 27). The tradition on the letter
sent by Cleon after the victory at Sphacteria (discussed above, 94–8) offers one possible earlier instance
of an official written communication sent to the Athenian assembly by a general.
120
Dover (in Gomme, Andrewes, and Dover 1970: 385) points out that in view of Thucydides’
stress on the exceptionality of Nicias’ decision, the allusion to earlier exchanges (K ¼ººÆØ ººÆE
KØ ºÆE) should not be interpreted in connection with ‘letters’: ‘It follows necessarily that by
KØ º Thucydides means “message”.’ See also, in the same sense, Hornblower 2008: 560.
121
Thuc. 7. 8. 2:   c ƒ   Ø j ŒÆa c F ºªØ IıÆ Æ j ŒÆd  Å
KººØE ªØª Ø j fiH Zåºø fi æe åæØ Ø ºª P a ZÆ Iƪªººø Ø, ªæÆł KØ º ,
 Çø oø i ºØ Æ c ÆF ª! Å Åb K fiH Iªªºø fi IçÆØ ŁE Æ ÆŁÆ f
#ŁÅÆ ı ıº Æ ŁÆØ æd B IºÅŁ Æ.
122
So Dover, in Gomme, Andrewes, and Dover 1970: 385: ‘Thucydides makes it sound as if it were
unusual for a general to write home’; even more radical is Harris 1989: 78: ‘he had probably not written
often if at all before, in spite of 7. 11. 1, where KØ º probably means ‘message’’; Hornblower 2008:
555. A number of scholars have, however, taken the opposite view, that Thucydides is here thinking of
numerous real letters sent by Nicias: so Longo 1978: 540 and n. 7; Stéfanis 1997: 201–2; Rosenmeyer
2001: 57–60.
123
Harris 1989: 78. See also the short, but illuminating, remarks of Bertrand 1985: 110 and Nicolai
2004: 119 (it was not normal practice at the time for a general to send a letter; written messages were
generally accompanied by oral presentations).
144 Greek Beginnings

precise knowledge of the events. He thus sees in writing a positive solution to the
difficulties of communication at a distance; but he also acknowledges that a letter
cannot entirely convey reality, so much so that he gives explicit instructions to the
messengers to enable them to answer eventual questions. The narrator does not
express himself either way—he simply reports in detail the reasons for Nicias’
decision. Most interpreters follow Nicias in his assessment, and assume that
Thucydides agreed with him in recognizing the greater reliability of a written
document against oral information.124 Of course, there is something true here;
after all, Thucydides wrote his work, and could present it as a ŒB Æ K I , a
permanent possession, because it was in writing. But should the same criteria be
applied to a historiographical work (Thucydides’ ŒB Æ K I , whose value will be
grasped in time and which is explicitly contrasted with a showpiece for a single
hearing, Iª!Ø Æ K e ÆæÆåæB Æ IŒ Ø) and to an official letter addressed on
a specific occasion to a large body (the ıº first, and then possibly the assembly)?
A comparison with the position upheld by the rhetor Alcidamas in his work
Against Those Who Write Written Speeches, or, On the Sophists may be helpful.
For the sake of the argument, Alcidamas takes in his speech an extreme, almost
paradoxical position, arguing in a written speech against writing speeches; but he
does not altogether deny the value of writing. His point is that elaborate speeches,
in which the choice of words is fixed, because the speech has been written in
advance, insinuate in the audience the feeling of being in front of something
artificial; they breed incredulity and antipathy (IØ  Æ ŒÆd çŁı a H
IŒıø ª! Æ K Ø ºA Ø, 12). Moreover in improvising the speaker can
react to the public (22–3), exactly Nicias’ point, but here seen as a positive rather
than a negative quality.
The most important sentence for our purpose comes from the close of the
speech: Alcidamas asserts there that ‘the precision that characterizes written
speeches does not bring the same advantage as the opportunity that declamations
of speeches composed on the spur of the moment offer’.125 His conclusion is that
‘whoever wants to become a brilliant orator and not simply a good writer of
written speeches, whoever wants to be able to make the most of occasions rather
than speak with precision, whoever wants to have on his side the goodwill of the
assembly rather than its diffidence as antagonist’, this person will do well to
dedicate himself to the art of improvising speeches.126 Throughout his pamphlet,
Alcidamas has in mind success with the public and an ability to grasp the kairon,
the right opportunity; he is of course aware of the potentialities of writing, and
says as much. But these potentialities lie in a different area from that covered by
deliberative or judicial speeches: writing is useful because by looking closely at
written texts, as in a mirror, it is possible to observe the progress of intellect; and it

124
So e.g. Bearzot 2003: 287, who suggests that the thoughts of Nicias here may give us a clue to
understanding Thucydides’ documentary choices; Longo 1978: 520–1.
125
Alcid. 1 (On Those Who Write Written Speeches), 33: P ªaæ  Æ Å TçºØÆ ƃ H ªæÆH
ºªø IŒæ ØÆØ ÆæÆØÆ Ø, ‹ Å PŒÆØæ Æ ƃ H KŒ F ÆæÆåæB Æ ºª ø ź! Ø åı Ø.
126
‹ Ø s KØŁı E Þ øæ ª ŁÆØ Øe Iººa c ØÅc ºªø ƒŒÆ, ŒÆd  ºÆØ Aºº
E ŒÆØæE åæB ŁÆØ ŒÆºH j E O Æ Ø ºªØ IŒæØH, ŒÆd c hØÆ H IŒæø ø K Œıæ
åØ ıÇØ Aºº j e çŁ IƪøØ   . . . , Alcid. 1. 34. See in general the remarks of
Avezzù 1982: 72–8.
 When a Letter and Why? 145

is useful in order to leave a memory.127 Alcidamas’ pamphlet is dated approxi-

mately to the 390s, and any direct contact between his work and Thucydides’ is
unlikely; yet Alcidamas is part of that debate on writing and its uses that is present
in tragedy, and that was linked, in the fifth and early fourth century, to the
character of Palamedes (an Odysseus, or, Against Palamedes for Treason is also
attributed to Alcidamas). His ideas were certainly in the air. The terminology, in
the passages quoted and throughout the pamphlet, is strikingly close to that
employed by Thucydides in 1. 22, and more generally in the methodological
chapters and the ‘archaeology’—of course, the connotations it has in Alcidamas
are exactly opposite to those in Thucydides. But then they have two very different
goals in mind: on the one hand, a work for reading that will not please the
majority but will provide the few who will tackle it with instruction and a way
of understanding human behaviour; on the other, a speech that will please, that
will rise to the kairos, and thus win the approval of the majority.128
Thus the fact that Nicias, in this specific context, settles for the truth, aletheia,
rather than for being able to grasp the kairon (or, more precisely: rather than
trusting his envoy’s ability to grasp the kairon), should not be uncritically assessed
as positive; it is just one more trait that fits his character, a character that will be
one of the causes of the catastrophe of the Sicilian expedition.
Looking at the larger context may help assess the specific impact attributed to
the decision of writing a letter. The act of sending for reinforcements is presented
as happening in parallel in both camps. In 7. 7, the narrative focuses on Gylippus’
travelling across to Sicily to collect reinforcements and win over new supporters,
as well as on the despatching of a group of envoys to Sparta and Corinth to ask for
further troops. One of the arguments for the request is that the Athenians too are
sending home for fresh troops (7. 7. 3). This prepares the shift of focus to the
Athenians’ activities in 7. 8, where Nicias notices the movements of the others and
decides to send to Athens for reinforcements. The Corinthian and Lacedemonian
requests, entrusted to a group of presbeis, are successful; just as successful is the
recruiting campaign of Gylippus: we learn in 7. 17–19 that Corinthians, Sicyo-
nians, Lacedaemonians, and the Peloponnesians in general send troops, while
Gylippus comes back with ‘as large a body of troops as he could secure’ from each
of the cities he had convinced to join in the fight (7. 21). Nicias’ request is basically
unsuccessful: the Athenians refuse to recall the army (his first alternative), and
they refuse to relieve him from the command of the expedition (7. 16. 1); they do
send reinforcements, but do not seem to fully grasp the gravity of the situation,
and delay (7. 16. 2–17. 1; 7. 20. 2–3; 7. 31), against Nicias’ explicit recommenda-
tion (7. 15. 2), causing for instance the loss of Plemmyrion. Strikingly, the
possibility of a failure in convincing the Athenians had been anticipated by Nicias:

127
Alcid. 1. 32: N b a ªªæÆ Æ ŒÆØÆ u æ K ŒÆæø fi ŁøæB ÆØ a B łıåB
KØ Ø ÞᾴØ K Ø. Ø b ŒÆd Å EÆ ŒÆƺØE  H ÆPH ıÇ ŒÆd B fi çغØ fi Æ
åÆæØÇ Ø ºªı ªæçØ KØåØæF . The image of the mirror appears also in discussions of
epistolary writing, adapted to the context.
128
Akroasis (public lecture, hearing), akribeia (precision), terpsis (pleasure), akoe (oral report),
parachrema (immediate, present, as opposed to lasting) are all present in both Alcidamas and Thuc.
1. 22; in a sense, Thucydides’ work aims to be that ‘mirror’ of which Alcidamas speaks (Thucydides
uses skopein, ‘examine’, ‘look into,’ whereas Alcidamas has katidein or theoresai, with similar meaning).
146 Greek Beginnings

Dover is certainly right that ‘the whole point of his writing a letter was to disarm any
suspicion on the part of the assembly that the messengers were misrepresenting
him’.129 Yet the solution did not work; and all this ends in emphasizing once again
Nicias’ difficult relationship with the soldiers and with the Athenian assembly.130
To conclude: the Thucydidean strategy of highlighting Nicias’ decision of
sending a letter as a peculiarity, and of omitting any mention of a discussion of
the contents of that same letter in the assembly,131 aims at bringing to the fore a
complex set of meanings. Technically, the device of the letter offers Thucydides
the possibility of giving Nicias one last speech in front of the Athenian assembly.
At an overt level, the choice of a letter allows for an explicit reflection on the
advantages and disadvantages of this new form of official communication (the
‘discovery’); Nicias’ decision to settle for the—ultimately unsuccessful—aletheia
also serves to characterize his generalship. Implicitly, the juxtaposition of the
unsuccessful choice by Nicias of sending a letter and of the Syracusans’ trad-
itional—and successful—practice of sending envoys may hint at the unsuitability
of letters for official communication, when decisions by the assembly are involved.
The numerous references to letters in book 8 mostly reinforce the notion of
the unsuitability of letters for official communication. In no case is the text of the
letter quoted; in most instances Thucydides briefly summarizes the content of
the message, with words implying the use of writing (although this is not always
clear when the term used is a form of KØ ººø). The series begins with an
instance of communication among two Spartan generals on the war front. The
letter (? KºŁ Å . . . e  ŒÆ KØ ºB, where epistole might also mean simply
‘message’) of the commander Pedaritus to the navarch Astyochus (8. 33. 3)
fittingly arrives during the night and contains information about the vague
possibility that some Erythraeans who have escaped from Samos might be
intending to betray Erythrae to the Athenians; upon investigation, the Spartan
generals find that the story of the Erythraeans’ willingness to betray the city had
been invented by them to get away from Samos in the first place.132 The second
letter is an instance of communication between the general and the Spartan
authorities at home (Thuc. 8. 38. 4, KØ ººØ æd ÆPF K c ¸ÆŒÆ Æ ›
—æØ): the same Pedaritus now accuses his colleague Astyochus; we learn
slightly later that the letter accomplished its purpose, and that because of Pedar-
itus’ letter the Spartans held Astyochus in suspicion (8. 39. 2: æe ªaæ a F
—Ææ ı KØ ºa !ı ÆP).133 A third letter, whose content is briefly

129
Dover in Gomme, Andrewes, and Dover 1970: 385.
130
Corcella 1996: 354. This aspect of Nicias’ personality is also highlighted by Plutarch Nic 2.3–6.2
(esp. 5.2 for the connection with writing).
131
Excellent discussion in Hornblower 2008: 568–9. Actually, the proceedings are extremely
compressed: the first meeting with the boule all but disappears.
132
Good analysis in Hornblower 2008: 844–6, who highlights the fact that it is impossible to know
whether those duped are finally the Spartans or the Athenians. But his comment (Hornblower 2008:
844) that ‘Th. uses the letter which conveyed information to Astyochos as a technique for informing us,
the readers, as well’ is slightly off the mark: an oral message would have functioned just as well. I would
rather suggest that the choice of a letter (or a messenger) arriving at night, rather than a messenger
reporting in face-to-face speech in full daylight, ties in well with the content and outcome of the affair.
133
See Hornblower 2008: 844 and 861–3 for comments on Pedaritus’ activity as letter writer. The
plural KØ º need not imply more than one letter, although, as Hornblower notes, in 8. 33. 3, 7. 8. 2
and 7. 10 Thucydides had used the singular.
 When a Letter and Why? 147

summarized by Thucydides, contains an order to Astyochus from the Spartan

authorities at home (KØ ºc æe #  å KŒ ¸ÆŒÆ ) to kill Alcibiades
(Thuc. 8. 45. 1).134
These exchanges between Spartans are followed by a very close, dramatic
succession of four letters, crossing the Aegean between Athenian Samos, the
neighbourhood of Miletus, where the Spartan fleet is anchored, and Magnesia,
the seat of Tissaphernes. These letters mark a series of astute moves by Phryni-
chus. The moment is the winter of 412/411, and the Athenian oligarch Phryni-
chus, ready to do anything to get rid of Alcibiades, hopes to find an ally in the
Spartan admiral Astyochus; to him he sends two letters. Astyochus betrays him,
however, and reveals the intrigue to Alcibiades, who in turn sends a letter to the
Athenians in Samos exposing Phrynichus’ dealings with Sparta. But in a final
twist, Phrynichus, having learnt of the betrayal, manages to pre-empt the accus-
ation and turns the tables on Alcibiades:135
He [Phrynichus] sent a secret letter to the Lacedaemonian admiral, Astyochus, who
was still in the neighbourhood of Miletus, to tell him that Alcibiades was ruining their
cause by making Tissaphernes the friend of the Athenians, and containing an express
revelation of the rest of the intrigue ( Ø . . . Œæ çÆ KØ  ºÆ ‹Ø #ºŒØØÅ . . .
ŒÆd pººÆ Æ ÆçH KªªæłÆ), desiring to be excused if he sought to harm his
enemy, even at the expense of the interests of his country. [3] However, Astyochus,
instead of thinking of punishing Alcibiades, who, besides, no longer ventured within
his reach as formerly, went up to him and Tissaphernes at Magnesia, communicated to
them the letter from Samos, and turned informer (ºªØ  ÆPE a KØ ƺÆ KŒ
B  ı ŒÆd ª ªÆØ ÆPe Åı ), and if report may be trusted, became the paid
creature of Tissaphernes, undertaking to inform him as to this and all other matters;
which was also the reason why he did not remonstrate more strongly against the pay
not being given in full. [4] Upon this Alcibiades instantly sent to the authorities at
Samos a letter against Phrynichus (› b #ºŒØØÅ PŁf  Ø ŒÆa æı åı
ªæ ÆÆ K c  ), stating what he had done, and requiring that he should be
put to death. [5] Phrynichus, distracted and placed in the utmost peril by the denunci-
ation, sent again to Astyochus (I ººØ ÆsŁØ æe e #  å), reproaching
him with having kept the secret of his previous letter so badly, and saying that he was
now prepared to give them an opportunity of destroying the whole Athenian arma-
ment at Samos. He provided a detailed account (ªæłÆ ŒÆŁ (ŒÆ Æ) of the means
which he should employ, Samos being unfortified, and pleaded that being in danger of
his life on their account, he could not now be blamed for doing this or anything else to

134
This is a rather problematic document: cf. Andrewes, in Gomme, Andrewes, and Dover 1981:
95–6. This letter and the ones mentioned in 8. 50–1 are part of a narrative section (probably extending
from 8. 36 to 8. 54, although other views have been held) that jumps back in time; more importantly, it
is placed at the beginning of a new style of narration that gives more place to authorial statements on
motives and detailed psychological analysis. For guidance on the problems posed by this part of
Thucydides’ work see Hornblower 2008: 883–6; Rood 1998: 262–8.
135
See Hornblower 2008: 901–7 for a brilliant interpretation of these chapters, labelled by him ‘the
game of chess’. In Hornblower’s view, Phrynichus would have been one step ahead throughout the
affair; his profession of treachery in the first letter would only have been a trick to deceive Astyochus
into believing the allegations against Alcibiades. This may have been so. And yet, this might not be
what Thucydides meant to convey: the wording in 8. 50. 5 points to something more than just ‘plan
A having gone awry, Phrynichus moved to plan B’. But definitely we are in the narrative category of the
‘clever trick’, for whose language in Thucydides see the excellent remarks of Hornblower 2008: 105–7
and 901–3.
148 Greek Beginnings
escape being destroyed by his mortal enemies. This Astyochus also revealed to
Alcibiades. 51. Meanwhile Phrynichus, having had timely notice that he was playing
him false and that a letter on the subject was on the point of arriving from Alcibiades
(ŒÆd ‹  P ÆæF Æ Ie F #ºŒØØı æd  ø KØ º ), himself antici-
pated the news, and told the army that the enemy, seeing that Samos was unfortified
and the fleet not all stationed within the harbour, meant to attack the camp; that he
could be certain of this intelligence; and that they must fortify Samos as quickly as
possible, and generally look to their defences. It will be remembered that he was
general, and had himself authority to carry out these measures. [2] Accordingly they
addressed themselves to the work of fortification, and Samos was thus fortified sooner
than it would otherwise have been. Not long afterwards came the letter from Alcibi-
ades, saying that the army was betrayed by Phrynichus, and the enemy about to attack
it (ƃ b Ææa F #ºŒØØı KØ ºÆd P ºf o æ wŒ ‹Ø æ Æ  e
æı Æ e æı åı ŒÆd ƒ º ØØ ººı Ø KØŁ  ŁÆØ). [3] Alcibiades,
however, gained no credit, it being thought that he was privy to the enemy’s designs,
and had tried to fasten them on Phrynichus and because of hatred was making out that
he was their accomplice; and consequently, far from hurting him, he rather bore
witness to what he had said by this intelligence. (Thuc. 8. 50–1)
These are clearly not official letters. They (and their carriers) seem to cross the
Aegean, from one army to the other, with an astonishing ease.136 As befits secret
letters, their exact content is not given. And the necessity of a letter, instead of
entrusting the message to the carrier, is not immediately clear.137 One cannot help
being struck by the sudden intensification, here as in the first book, of epistolary
materials: is it an accident that this happens once the war has moved to Ionia, and
once Persia is again one of the main players? It is also intriguing that the group of
letters of the eighth book appears (with the exception of the one letter exchanged
between Pedaritus and Astyochus) in a section of book 8 that goes backwards in
time, as if the intricacies of the narrative structure matched the intricacies of the
uses of communication.138 Finally, the eighth book presents an ensemble of
narrative characteristics that can be related to each other: it lacks direct speeches;
it is among those books that make most use of ‘documents’, transcribing them
directly into the text; and epistolary communication is mentioned rather fre-
quently, more than anywhere else in the War, exception being made for the
excursus in book 1.139 These aspects have often been interpreted along the
analytic-versus-unitarian divide; but these three peculiarities may also be viewed
as strategies tending to highlight, towards the end of the war, the move away from

136
Andrewes (in Gomme, Andrewes, and Dover 1981: 120) notes: ‘it must be assumed that secret
communication of this kind was common (cf. Nicias at 7. 48. 2 and elsewhere) and that commanders
were felt to have some discretion in what they did or did not reveal’—but the point is that these are
secret communications, and that while this kind of contact must have been indeed relatively frequent, it
could not be brought into the open.
137
Hornblower 2008: 906 refers to Schindel 1970: 289 for a comparison of this letter with the letter
[sic] of Themistocles in Hdt. 8. 75. 2–3. Schindel offers an excellent treatment of the affair—and makes it
clear that he is comparing Phrynichus’ letters (whose written status is signalled by the use of various
forms of ªæçø) to the oral message sent by Themistocles to Xerxes before Salamis, in Hdt. 8. 75, which
was a trick to force the Persians into battle. Once again, oral and written messages can share characteris-
tics of secrecy and intrigue: it is epistolary writing, and not writing as such, that is problematic.
138
On the structuring of the narrative in this part of book 8, see Rood 1998: 262–71.
139
See Andrewes and Dover (in Gomme, Andrewes, and Dover 1981: 382–3 and 389–93).
 When a Letter and Why? 149

face-to-face discussion, the reliance on treaties and agreements that appear solid
(and are ‘solidly’ quoted) but that systematically fail and have to be redrafted, and
the dependence on means of communication that favour secrecy and trouble-
making over direct interaction.140 In this case, the excursus of book 1 would take
on an added meaning: the narrative of the fall of the two leading politicians of
Athens and Sparta, just after the Persian wars, involved private communications
through letters, prefiguring the ample use of letters for irregular communications
made by the leading new politicians of Athens and Sparta when the war will
escalate—and prefiguring their end as well.
However this may be, the only cases of official letters in Thucydides are the
intercepted letter of the Great King to the Spartans, the letter of Nicias to the
Athenians, and the three messages exchanged between Spartan officials. Letters
were the normal means of communication employed by the Persian kings. The
reasons for having Nicias send a letter have been discussed. The existence of letters
from Spartans is not surprising (as we shall see better later on), nor is it isolated:
Sparta seems to have made a relatively abundant use of epistolary communication,
and the ancient tradition bears witness to this. The other letters mentioned by
Thucydides are all instances of personal correspondence by which the sender
passes on sensitive material in a secretive fashion. Of course, if the existence of
these letters is known to the audience of the War, it is because they failed in their
purpose and became public—or because they succeeded in making false infor-
mation public.141
Clearly, Thucydides is interested in this type of document. The reason for his
interest may be that the letters (in theory) offered the possibility of reaching
directly into the thoughts of the protagonists, without any mediation;142 but
letters are definitely also one of the means by which he characterizes individuals,
as well as the changing policies developing in the course of the Peloponnesian war.

4 . 3 . I N T O T H E F O U RT H C E N T U R Y:
CONTINUITY AND CHANGE

The activity of the two historians to be discussed next spans the first half of the
fourth century: if there is any truth in the claim that Ctesias was present at Cunaxa
in the entourage of Artaxerxes II Mnemon (Plut. Artax. 11, cf. Xen. Anab. 1. 8. 25),
then Xenophon and Ctesias were contemporaries. In Xenophon’s Hellenica,
which continues Thucydides’ War, letters seem to follow the Thucydidean pat-
tern; but the works on Persia and the Agesilaus offer a glimpse of a new, different
vision of letter writing. Similarly, Ctesias mentioned in his work epistolary

140
For this function of documents in the period of the uneasy peace (as well as later), cf. Connor
1984: 144–7; Hornblower 1996: 116–17 and 360. Bearzot 2003: 278 considers that the difficulty of
explaining a complex and changing situation would have pushed Thucydides into inserting documents.
The two explanations are not mutually exclusive.
141
On this see Longo 1978: 530–1; Harris 1989: 88 (and 128 for the presence of similar aspects in
Polybius); Steiner 1994: 221–2; Rosenmeyer 2001: 56–7.
142
So Bearzot 2003: 288.
150 Greek Beginnings

exchanges between Persians and Greeks, Athenians and Spartans, in the context
of the power struggle in the Aegean; but he also presents us with the first instance
of a suicide-note-cum-love-letter.

4.3.1. Xenophon

Xenophon’s work preserves some examples of official letters from generals to their
polis. Their relatively small number (altogether, eleven instances of epistolary
exchanges—although of course the term KØ º appears more often) has to be
seen against the background of a general scarcity of official documents.143 As had
been the case with Thucydides, these letters concern only Sparta, Athens, and
Persia. I shall discuss first the references to letters in the Hellenica and then those
present in other works. I shall close by comparing my findings concerning letters
with the presence of (and references to) other documents and written texts in
Xenophon’s oeuvre.144
The first letter in the Hellenica, a famous example of epistolary Laconism, was
sent by the second in command Hippocrates to the ephors after the naval disaster
at Cyzicus in 410; it was intercepted by the Athenians, which is why Xenophon
can quote its text:
 0EææØ a ŒAºÆ.  Ææ I Æ. ØHØ þæ. Iæ    åæc æA.
(Hell. 1. 1. 23)
The ships are gone. Mindarus is dead. The men are starving. We know not what to do.
It does not add much to the narrative, and its inclusion may have been due to
rhetorical factors: besides being a nice example of the Laconian style, it expresses
the Spartans’ feelings after the disaster at Cyzicus as no other comment could
possibly have done.
Writing also plays a role in the way the ephors deal with Cinadon’s conspiracy
(Hell. 3. 3. 8–11) in what appears to be an interesting variant of the death-bringing
letter: when they learn about the plans for a revolt, the ephors call Cinadon and
send him to Aulon with orders to arrest and bring back all those of the Aulonians
and Helots whose names are written on the official letter (¼ªÆ H `PºØH
 ØÆ ŒÆd H ƒº!ø f K B fi Œıºfi Å ªªæÆ ı, Hell. 3. 3. 8). To
accomplish this duty, some neoi (‘young men’) have been selected to accompany
him. However, the neoi have been previously informed that it is Cinadon they
must arrest, once out of the city, and that after learning from him the names of his
associates in the revolt, they must write them down and send them back to the
ephors (f b ıØÆ ıŁ Ø ÆPF ªæłÆ I Ø c Æå Å

143
Only two instances of inscriptions: the stele at Scillus, Anab. 5. 3. 13, on the plot dedicated to
Artemis by Xenophon himself, bearing a text typical of sacred regulations; and Xen. Cyr. 7. 3. 15, the
stele of the eunuchs (ŒÆd F e B Æ åæØ F F H P åø ŒåH ŁÆØ ºªÆØ· ŒÆd Kd b B fi ¼ø
 ºfiÅ F Iæe ŒÆd B ªıÆØŒe KتªæçŁÆØ çÆ d a O ÆÆ,  æØÆ ªæ ÆÆ, Œø b r ÆØ æE
ºªı Ø  ºÆ ŒÆd KتªæçŁÆØ ˚˙—ˇ-. ˝). Neither of these can be considered an ‘official
document’.
144
More details in Orsi 2002, who offers an extremely detailed discussion of the diplomatic
vocabulary of Xenophon, in particular in the area of international communications.
 When a Letter and Why? 151

E KçæØ, Hell. 3. 3. 10). The first official message is thus a fictive one, but the
second one, with the names of the accomplices, is a very real one: a horseman
returns with the names listed by Cinadon (wŒ ƒf çæø a O ÆÆ z ›
˚Øø IªæÆł) and the most influential associates are immediately arrested.
Writing here serves the purpose of credibly sending Cinadon away from Sparta,
allowing for the arrest to succeed, and the names of the associates to be extorted
without any danger of their learning that Cinadon has been arrested; the real
negotiations (how to best deal with the threat, and which men to send with
Cinadon) are conducted through private conversations.
As for Athens, Xenophon records the message sent to the boule and the people
by the strategoi after the battle of the Arginusae in 406 bc: it plays an important
role in Xenophon’s narrative of their trial, both in the accusations levelled at them
and in their defence. Theramenes and others produced their dispatch (KØ º )
in the assembly, and made much of the fact that the only reason given in it for not
collecting the shipwrecked had been the storm; they took this to mean that the
responsibility had been that of the generals only.145 In his turn, Euryptolemus
affirmed in his speech for the defence that Pericles and Diomedon ‘persuaded
their colleagues to change their mind when they wanted to send a letter to the
boule and to you, stating that they had assigned to Theramenes and Thrasybulus
the duty of picking up the shipwrecked with forty-seven triremes, and that they
failed to pick them up’.146 The dispatch had thus been discussed among the
generals and a first version drafted, which was subsequently discarded in favour
of the one blaming the failure simply on the storm. This is the only reference to an
Athenian letter in Xenophon’s oeuvre. One should note that it is successfully used
in the assembly for a political purpose: the destruction of those who had sent it.
Other reports of letters include some Persian ones. Xenophon ‘quotes’ a part of
the text of the letter bearing the royal seal that Cyrus brought along to prove his
nomination to karanos (‘supreme military commander’) in 407 bc:
. . . ŒÆd ˚Fæ, ¼æø ø H Kd ŁÆºfiÅ ŒÆd ı º ø ¸ÆŒÆØ  Ø,
KØ º   çæ E Œø A Ø e Æ ºØ çæªØ Æ åı Æ, K fi w KB ŒÆd
: ˚ÆÆ ø ˚Fæ ŒæÆ H N ˚Æ øºe ±ŁæØÇ ø. e b ŒæÆ  Ø
Œ æØ. (Xen. Hell. 1. 4. 3).
(the ambassadors met also with) Cyrus, who was to take command of all the peoples
on the coast and to support the Lacedaemonians in the war; he carried a letter,
addressed to all the inhabitants of the coastal area and bearing the King’s seal,
which contained among other things these words: ‘I send Cyrus as karanos of those
who are mustering in Castolus.’ Karanos means ‘lord’.
The term epistole is not used elsewhere in the Hellenica, but there are a number
of references to grammata, ‘writings’ of the King, spelling out the terms of the
Peace of 387/386 and the later Peace of 367/366. The text of the King’s Peace of
387/386 (referred to as a ªªæÆ Æ by the narrator in 5. 1. 30, and as a

145
Xen. Hell. 1. 7. 4: KØ ºc K Œı Ææ æØ m  łÆ ƒ æÆŪd N c ıºc ŒÆd N
e B , ¼ºº Pb ÆNØ! Ø j e åØ HÆ.
146
Xen. Hell. 1. 7. 17: ŒÆŪæH b s ÆPH ‹Ø Ø Æ f ıæåÆ ıº ı  Ø
ªæ ÆÆ B fi  ıºB fi ŒÆd  E ‹Ø KÆÆ fiH ¨ÅæÆ Ø ŒÆd ¨æÆ ı ºøfi ÆæŒÆ ŒÆd a
æØ æ Ø Iº ŁÆØ f Æıƪ , ƒ b PŒ I º.
152 Greek Beginnings

Æ Øºø ªæ ÆÆ in the words attributed to Agesilaus in 5. 1. 32), begins in

Xenophon’s version as a rescript from the King, in the third person and without
any specific addressees, but then moves to the first person, in other words it
‘behaves’ exactly as a letter would:
#æÆæÅ Æ Øºf  ÇØ  ŒÆØ a b K Bfi # fi Æ ºØ ÆıF r ÆØ ŒÆd H
 ø ˚ºÆÇ a ŒÆd ˚ æ, a b ¼ººÆ  EººÅ Æ ºØ ŒÆd ØŒæa ŒÆd ªºÆ
ÆP ı IçEÆØ ºc ¸ ı ŒÆd  ” æı ŒÆd Œ æı: Æ Æ b u æ e IæåÆE
r ÆØ #ŁÅÆ ø. ›æØ b Æ Å c Næ Å c åÆØ,  Ø Kªg º ø
a H ÆFÆ ıº ø ŒÆd ÇBfi ŒÆd ŒÆa ŁºÆÆ ŒÆd Æı d ŒÆd åæ Æ Ø.
(Xen. Hell. 5. 1. 31)
‘King Artaxerxes thinks it just that the cities in Asia should belong to him, as well as
Clazomenae and Cyprus among the islands, and that the other Greek cities, both small
and great, should be left independent, except Lemnos, Imbros, and Scyros; and these
should belong, as of old, to the Athenians. But whichever of the two parties does not
accept this peace, upon them I will make war, together with those who desire this
arrangement, both by land and by sea with ships and with money.’
(Trans. Brownson 1918)
Importantly, before reading the text to the assembled Greeks Tiribazus had
shown them the King’s seal (a Æ Øºø Å EÆ, 5. 1. 30), just as the King’s seal
had been an essential part of the letter giving authority to Cyrus. The language
attributed by Xenophon to Tiribazus is also noteworthy: the satrap orders that
those who are interested in hearing ‘the peace that the king has sent’ should be
present (ÆæEÆØ f ıº ı ÆŒF ÆØ m Æ Øºf Næ Å ŒÆÆ Ø,
Xen Hell. 5. 1. 30): the ‘peace’ is sent as a letter. As Dillery notes, this is basically
‘the only document that Xenophon purports to quote in the Hellenica’.147 Xeno-
phon’s is the fullest surviving treatment of the peace, yet this is probably not a
literal quotation. He may have preserved the preamble, or the gist of the edict, but
the full official text of the document must have contained further specific
clauses.148 However, Badian has convincingly argued that the wording reflects
very closely Persian ideology on the one hand (e.g. the use of  Çø  ŒÆØ—this
is an edict, and it rests on the ‘justice’ of the Persian king; note  Çø rather than
ºªø or indeed the Œø found in Athenian decrees) and on the other Spartan
phraseology (e.g. the formula ºØ ŒÆd ØŒæa ŒÆd ªºÆ, part of the trad-
itional formulary of Spartan diplomacy, at least in the versions preserved by
Xenophon and Thucydides).149 The framing of the document, which is halfway

147
Dillery 1995: 201. There are some exceptions though: the letter sent by Hippocrates at Sparta,
Xen. Hell. 1. 1. 23 (above); the one nominating Cyrus (above); other exceptions are listed below.
148
Isocrates states in Paneg. 180 and Panath. 107 that the text was inscribed throughout Greece. It
is unclear to me whether what Xenophon gives is the Greek translation of an official text itself written
according to specific indications of the Spartans, or whether there ever only was a Greek text, resulting
of the collaboration between Tiribazus and the Spartans; Badian’s point (1987, and 1991: 37–9) on the
unlikelihood that the King could swear a peace with a Greek city seems to me convincing. See also
Dillery 1995: 202 (and Rhodes and Osborne 2003, no. 20, an alliance between Athens and Chios dated
to 384/383, for a mention of the oath of the King, with their commentary); more generally Dillery 1995:
199–207 for the context of the Peace.
149
Persian formulae: Badian 1987: 27–8; Spartan formulae: Badian 1991: 35–6, with reference to
Xen. 5. 2. 35; Thuc. 5. 77. 5, and also 1. 125. 1 (for Badian, Thucydides’ reformulation of a Spartan
formula).
 When a Letter and Why? 153

between an edict and a letter, could perfectly well be the result of the cooperation
between the Spartans and the Persians around Tiribazus.
This document is referred to more than once in Xenophon, always as ‘the
writing’ (a ªæ ÆÆ) of the King: besides the instances seen above, this happens
also in the course of the speeches delivered in Sparta in 372/371 bc by the
Athenian ambassadors Autocles (Hell. 6. 3. 9: ŒÆa a Æ Øºø ªæ ÆÆ) and
Callistratus (Hell. 6. 3. 12): ‘For the King wrote that all the cities in Greece should
be autonomous’, BÆ Øºf b ªaæ  ı ªæÆł  Æ a K B fi  EººØ ºØ
ÆP ı r ÆØ.
However, the notional equivalence betwen grammata and epistole when talking
of a writing by the Persian king addressed to all Greeks is made explicit in the
narrative of the Peace that the Thebans tried to impose, with the agreement of the
King, in 367 bc; the passage (Xen. Hell. 7. 1. 36–9) gives, moreover, a fascinating
glimpse into how these texts were constructed (or into how Xenophon imagines
these texts were constructed). Pelopidas is first asked by the King’s representative
what he would like to stand in the text (›   ºØ ÆıfiH ªæÆçBÆØ); his
conditions (mainly autonomy for Messene) are then written down and read out
to the ambassadors (ªæÆçø b  ø ŒÆd Iƪø Łø E æ  Ø),
presumably for approval. But at this point, an Athenian ambassador, Leon, makes
a disobliging comment in the hearing of the King. As this is translated to the King,
the answer again comes in writing: ‘And if the Athenians are aware of anything
juster than these provisions, let them come to the King and inform him.’ (Kd b
I ªªغ › ªæÆ Æf L r  › #ŁÅÆE, ºØ K ªŒ æ ªªæÆ Æ· N
 Ø ØŒÆØæ  ø ªØª! Œı Ø ƒ #ŁÅÆEØ, NÆ æe Æ ØºÆ
Ø ŒØ, 7. 1. 37.) Once the agreement has been reached, the Thebans invite
all the representatives from the cities to hear the letter of the King (IŒı  ı
B Ææa Æ Øºø KØ ºB), and at the meeting the representative of the King
shows to all the seal and reads the text (ŒÆd › —æ Å › çæø a ªæ ÆÆ  Æ
c Æ Øºø çæƪEÆ Iªø a ªªæÆ Æ). Communication does indeed
happen through writing on the Persians’ part.
The rest of Xenophon’s oeuvre shows features that correspond to the picture
emerging from the analysis of the Hellenica. There are quite a few Persian letters,
some of them treasonous, others reflecting the normal practices of Persian
communication. Among the former is the letter sent by Orontes to the King, an
object lesson in the dangers of communication by letters: with it, Orontes pro-
posed to change camp and join the army of the King, reminding the latter of his
former friendhip and fidelity. Orontes gave it to a man he considered trustworthy;
but this person took the letter and handed it directly to Cyrus (Anab. 1. 6. 3), with
the result that Orontes was condemned to death, utterly disappearing, so that ‘no
grave of his was ever seen’.
As for the letters exemplifying normal Persian communication, the Cyropaedia
offers examples pertaining both to official communication (Cyr. 4. 5. 26–34, cf. 5.
5. 4; 8. 2. 16–18) and to private, personal exchanges (2. 2. 9–10). Accounting
through written lists (ªªæÆ Æ . . . IŒæØH, or a ªæ ÆÆ) is presented as
central to the Persian system, as in Croesus’ gesture of passing to Cyrus a list of
everything that was being transported in each of the wagons leaving Sardis. Cyrus
hands over the list to his subordinates, stating that for himself it is enough to trust
that those in charge deserve those valuables; but the friends and officers are
154 Greek Beginnings

requested to exercise control (Cyr. 4. 7. 12–13). It is also in the Cyropaedia that

Xenophon offers a full picture of the Persian postal system (Cyr. 8. 6. 17–18):
Xenophon is fully aware of the importance of communication for the empire. The
fact that written communication is defined a ÅåÅ Æ (Cyr. 8. 6. 17), a device to
cope with the dimension of the empire explains also, e contrario, why the Greek
poleis never had a postal system, an organized way of transmitting written
communications. Yet Xenophon’s comment stresses the positive aspects of such
a system (‘and it is a fine thing to have immediate intelligence of everything, in
order to attend to it as quickly as possible’, Cyr. 8. 6. 18).
On the Greek side, two letters play a role at the beginning and towards the end
of Xenophon’s personal anabasis. A letter from his friend Proxenus was the reason
Xenophon decided to join the expedition, as the reader learns from a famous
passage of the Anabasis in which, after the generals and commanders of the Greek
army have been seized and put to death following Cyrus’ defeat, Xenophon
introduces himself and his role in the expedition:
There was a man in the army named Xenophon, an Athenian, who was neither general
nor captain nor private, but had accompanied the expedition because Proxenus, an
old friend of his, had sent him at his home an invitation to go with him . . . [5]
After reading Proxenus’ letter (KØ ºc) Xenophon conferred with Socrates . . .
(Anab. 3. 1. 4–5).
Socrates suggested he should consult Delphi before joining the adventure; but
the letter had already done its work, and Xenophon asked to what god he should
sacrifice in order to accomplish successfully the journey he was contemplating—
and not, as Socrates reproached him afterwards, whether he should go or not.
A second letter seems to play a role towards the end of the long journey
homewards, one given by Anaxibius to Xenophon that should have allowed the
army a safe passage from Perinthus (on the European coast of the Propontis) to
Asia (Anab. 7. 2. 8). This letter will prove in the event useless because of a change
in those placed in command.
Finally, there is a unique and very important scene in which the Spartan king
Agesilaus refuses to accept a letter. After having stressed the latter’s openness, ease
of access, and kindness, Xenophon adds that although cheerful and friendly, he
could show great dignity ( ªÆºªø  Å), and gives this refusal as an instance:
KŒE ªaæ ‹’ qºŁ ÆPfiH KØ ºc Ææa Æ Øºø, m › a ˚ƺº Æ F
¸ÆŒÆØ  ı —æ Å XªŒ, æd  Æ  ŒÆd çغ Æ [ÆPfiH], Æ Å b PŒ
KÆ, fiH b çæØ r  IƪªEºÆØ Æ ØºE ‰ N fi Æ b æe ÆPe Pb Ø
KØ ºa  Ø, j b ç º B
fi ¸ÆŒÆ Ø ŒÆd Bfi  EººØ hı J çÆ ÅÆØ, ‹Ø
ŒÆd ÆPe ç º Ia Œæ ÆPfiH  Ø· j Ø, çÅ, KØıº ø ±º ŒÅÆØ, Å
i ı ººa KØ ºa åø ÆØ, ç º (Ø  N Łø. (Ages. 8. 3)
Thus, when the Persian envoy who came with Callias the Lacedaemonian handed him
a letter from the King containing offers of friendship and hospitality, he declined to
accept it, and said to the bearer that he should tell the King that there was no need for
him to send him private letters, but, if he gave proof of friendship for Lacedaemon and
goodwill towards Greece, he too for his part would be his friend with all his heart.
‘But’, he said, ‘if he is found plotting, let him not hope to have a friend in me, however
many letters I may receive.’ (Trans. Marchant, slightly modified)
 When a Letter and Why? 155

This is a fascinating passage, bringing out well the distinction between public
position and personal epistolary communication. That it should be a Spartan who
points out this distinction is paradoxical, but then Agesilaus (and especially
Xenophon’s Agesilaus) is in many ways a paradoxical Spartan.
It would seem thus that the Greek historians of the fifth and early fourth
century (down to and including Xenophon) connect epistolary writing rather
closely with a despotic or oriental milieu by underlining the connotations of
privateness and individuality of the letter. There are in Thucydides and Xenophon
some instances of official letters, but they are still fairly rare. That this is not
something historiographical, but is due to the infrequent use of letters for official
communication in reality is shown in my opinion by the fact that these few letters
do not yet present a fully developed formulary, either for the introduction or for
the conclusion, and even more by the fact that until the end of the fifth century
there is not a term that unequivocally denotes a ‘letter’.150 This does not of course
mean that the Greeks by the end of the fifth century did not exchange letters:
rather, in the context of a more general tendency to avoid or mention with a
certain economy documents and decrees, that is, to avoid the intrusion of ‘other’
writing into one’s work, the infrequent references to letters must be interpreted as
part of a strategy aimed at defining and characterizing in particular ways those
individuals that choose letters as their instrument of communication. This also
explains why, in Herodotus as in Thucydides and Xenophon, the content of most
letters is simply summarized (or plainly rewritten): documentary precision does
not play a role here.

4.3.2. Ctesias

Of Ctesias’ Persica we have only the summary of Photius, and, with some notable
exceptions, fragments (not clearly marked as such) transmitted through indirect
tradition (mainly Diodorus Siculus and Plutarch, but also Nicolaus of Damascus
and Athenaeus). It is thus very difficult to give a just appraisal of his work.151 Even
in such a situation, some aspects emerge clearly. In particular, it is evident that
writing and written documents (whether real or invented is for the moment
irrelevant) played an important role in his work and in how he presented it.
Photius’ summary makes it clear that Ctesias pretended to have been present at
most of the events narrated, or, where this was impossible, to have learnt about
them from Persian sources.152 This is supported by other passages of Ctesias’

150
See above, ch. 1. 4, and esp. 17–18.
151
Among recent work, see Drews 1973: 103–16; Dorati 1995; Lenfant 2004; Tuplin 2004; Stronk
2007.
152
T 8 Lenfant = FGrH 688 T 8 = Phot. Bibl. 72. 35b 35: çÅ d b Æe H ºØø L ƒ æE
ÆPÅ ª  j Ææ’ ÆPH —æ H, ŁÆ e ›æA c Kå!æØ, ÆP Œ ŒÆÆ Æ, oø c
ƒ æ Æ ıªªæłÆØ. On Ctesias’ stress on personal observation or personal hearing, which he
contrasted with Herodotus’ supposed failings, see Marincola 1997: 227–8; detailed discussion of
Ctesias’ sources and of his way of presenting them in Lenfant 2004: xxvii–xxxix. Regarding the degree
of correspondence to reality of Ctesias’ account, I tend to side with the sceptics (Jacoby’s position is
nowadays less well represented than it was, but see e.g. Dorati 1995; Binder 2008: 55–7); for a more
optimistic view, see e.g. Lenfant 2004; Stronk 2007. While making a case for the existence of Persian
archives, even Stronk has to admit that the type of stories narrated by Ctesias would not have been
156 Greek Beginnings

work: the following statement, preserved by Diodorus (2. 32. 4), probably stood in
the proem of Ctesias’ work:153
˚Å Æ ’ › ˚ Ø E b åæØ Bæ ŒÆa c ˚ æı æÆ Æ Kd
#æÆæÅ e Iºç, ª  ’ ÆNå ºø ŒÆd Øa c NÆæØŒc KØ  Å
IƺÅçŁd e F Æ Øºø, ÆŒÆ ŒÆ Šغ  Ø !  ’ ÆPF. y
s çÅ Ø KŒ H Æ ØºØŒH ØçŁæH, K Æx  ƒ —æ ÆØ a ƺÆØa æØ ŒÆ ØÆ
  r å ıƪ Æ, ºıæƪ B ÆØ a ŒÆŁ’ (ŒÆ , ŒÆd ıÆ  c
ƒ æ Æ N f  ‚ººÅÆ KªŒE. (F 5 Lenfant = FGrH 688 F 5)
Ctesias of Cnidus, on the other hand, lived during the time when Cyrus made his
expedition against Artaxerxes his brother, and having been made prisoner and then
retained by Artaxerxes because of his medical knowledge, he enjoyed a position of
honour with him for seventeen years. Now Ctesias says that from the royal records, in
which the Persians in accordance with a certain law of theirs kept an account of their
ancient affairs, he carefully investigated the facts about each king, and when he had
composed his history he published it among the Greeks.
Ctesias thus claimed to have made use of basilikai diphtherai for his work,
which he consulted at the time of his stay at the Persian court. Similarly, when
narrating accounts of the earliest kings of Assyria in his second book, he asserted
that the Memnon sent as an ally to the Trojans was the son of Tithonos, an
eparchos of Persis, and added that ‘The barbarians say that this is what is told of
Memnon in the royal texts’, æd b s   ØÆF’ K ÆE Æ ØºØŒÆE
IƪæÆçÆE ƒ æE ŁÆ çÆ Ø ƒ æÆæØ (F 1b Lenfant = FGrH 688 F1b =
D. S. 2. 22. 5). The meagre fragments we have show that Ctesias quoted inscrip-
tions: his version of the Behistun inscriptions, which he attributed to Semiramis, is
a memorable instance—but ludicrous might be a more appropriate epithet in this
case. According to Ctesias, the queen having smoothed out the cliff wrote on it in
Syrian letters: ‘Semiramis, with the pack-saddles of the beasts of burden in her
army, built up a mound from the plain and thereby climbed this precipice, even to
this very ridge’ (F 1b Lenfant = FGrH F1b = D. S. 2. 13. 2). Again, one cannot but
admire Ctesias’ ingenuity when, in citing a Greek epigram four hexameters in
length to prove his point on Sardanapalus and his inclinations towards pleasure,
he adds that this was originally in Barbarian language, and was only later
translated into Greek (FGrH 688 F1b = D. S. 2. 23. 3). Interestingly, this is in
tune with Xenophon’s use of inscriptions: he cited only two, but of these one, from
the Anabasis, was a dedication; the other one, from the Cyropaedia, concerned
eunuchs (and was written in Syrian letters).154
With Ctesias, thus, we are dealing with an author extremely conscious of the
power of writing to attest the truthfulness or enhance the reliability of any
statement; and not only that, but also an author who can consciously manipulate

recorded there. And when he settles for Ctesias as a writer of historical fiction (see esp. 49–50), he is
basically endorsing the judgements of Wilamowitz and Jacoby (respectively ‘Er hat den ersten
historischen Roman geschrieben’ and ‘Vater des historischen Romans’), as cited by Stronk himself
(2007: 48 n. 94).
153
See Tuplin 2004: 307, who thinks of this as an intermediate preface.
154
See above, n. 143. The Ctesian paternity of the epigram is doubted by Jacoby, who prints this part
of Diodorus in small characters and notes in apparatus: ‘From Cleitarchus?’; Lenfant 2004 does not print
this at all; but see Drews 1970; Walbank 1967: 83–5 (‘perhaps from Ctesias’); Weißbach 1920: 2441.
 When a Letter and Why? 157

his writing.155 This attention to writing played a role beyond the part of his work
concerning the earliest times and the oriental world: the summary of Photius as
well as some scanty fragments show that even when discussing more recent times,
and in situations where he could have relied on personal observation, he used a
good deal of epistolary material, even presenting himself as personally involved in
epistolary dealings. Photius, in summarizing the more recent part of Ctesias’
Persica (books 21, 22, and 23), states that in it were narrated
causes of the quarrel of king Artaxerxes with Evagoras, king of Salamis (ÆN ÆØ Ø’ L
¯PƪæÆ fi Æ ØºE ÆºÆ E Æ Øºf #ææÅ ØÅåŁÅ); and messengers
(¼ªªºØ) sent by Evagoras to Ctesias about the receiving of letters from Abuletes
(bæ F ºÆE Ææa #ıº ı a KØ º); and a letter (KØ º ) of Ctesias to
Evagoras concerning reconciliation with Anaxagoras, king of the Cypriots. Return of
the messengers (Iªªºø) of Evagoras to Cyprus and delivery of the letters
(ªæÆ ø I Ø) from Ctesias to Evagoras. 73. And a speech (ºª) of
Conon to Evagoras about visiting the king; and a letter (KØ º ) of Evagoras on
the honours he had received from him. And a letter (KØ º ) of Conon to Ctesias; an
agreement of Evagoras to pay tribute to the king; and the delivery of the letters (H
KØ ºH I Ø) to Ctesias. A speech (ºª) of Ctesias to the king about Conon;
and a letter (KØ º ) to him. Delivery to Satibarzanes of the presents sent by
Evagoras; arrival of the messengers (H Iªªºø) in Cyprus. Letter (KØ º ) of
Conon to the King and Ctesias. 74. Detention of the Spartan ambassadors (¼ªªºØ) to
the king. A letter (KØ º ) from the king to Conon and the Spartans, delivered to
them by Ctesias himself (L ˚Å Æ ÆPe KŒ Ø ). Conon appointed commander
of the fleet by Pharnabazus. Visit of Ctesias to Cnidus, his native city, and to Sparta.
Proceedings against the Spartan ambassadors (Iªªºı) at Rhodes, and their acquit-
tal. Numbers of stations, days, and parasangs from Ephesus to Bactria and India.
Catalogue of kings from Ninos and Semiramis to Artoxerxes. And here is the end.
(Lenfant F 27 = FGrH 688 F 27 = Phot. Bibl. 72. 44 b20—45 a4)
The loss of Ctesias’ work is regrettable for a number of reasons; but even such a
bare summary makes it clear that the letters were central to the construction of his
very last books: they seem to have been the nucleus around which the events were
organized. That Ctesias did not limit himself to playing the role of innocent
messenger is shown from another passage, which gives another glimpse of how
affairs were conducted in the Aegean at the beginning of the fourth century.
Plutarch in his Life of Artaxerxes (21. 2–4) asserts that for his projects to succeed,
Conon realized that he needed power ( Æ Ø), and that the King needed an able
man (Icæ  çæ). Conon thus sent a letter to the King ( ł KØ º ).
Moreover, he suggested that the carrier transmit it, if at all possible, through Zeno
of Crete or through Polycritus of Mende, the first a dancer, the second a physician;
but if these were not present, through Ctesias the physician. At this point, things
become more complex:
ºªÆØ ’ › ˚Å Æ c KØ ºc ºÆg Æ檪æłÆØ E e F ˚ø
K ƺ Ø, ‹ø ŒÆd ˚Å Æ I  ºfiÅ æe ÆP, ‰ TçºØ  ZÆ ÆE Kd
ŁÆº fi Å æ Ø· › b ˚Å Æ ÆPe Iç’ ÆıF Æ ØºÆ çÅ d æ ŁEÆØ c
ºØıæª Æ ÆPfiH Æ Å.
(Plut. Art. 21. 4 = T 7d and F 32 Lenfant = FGrH 688 T 7d and F 32)

155
Ctesias and writing: Tuplin 2007: 307–10.
158 Greek Beginnings
And it is said that Ctesias, on receiving the letter, added to the suggestions which
Conon made to the king a request to send Ctesias also to him, as likely to be of service
in matters on the sea coast. Ctesias, however, says that the king of his own accord
conferred upon him this new duty.
Ctesias is here presented as not having been content with the simple addition of
an oral message to the letter of Conon; it is claimed that in order to convey the
impression that the request had come from Conon himself, he added a few words
to the letter. Unluckily, we are not in a position to know who exactly it was that
accused Ctesias of having tampered with the text of the letter, and whether this
was external information, or information deduced from Ctesias’ own work; the
hypothesis that here Plutarch depends on Dinon’s account, who would have relied
for his accusation on Ctesian materials, has much in its favour.156 Whether there
may be some reality in the role that Ctesias attributed to himself in these events or
not, the insistence on letters demands our attention: especially if the part he
claimed to play did not really correspond to his actual impact on the diplomatic
negotiations, Ctesias would not have mentioned documents that rendered his
account less plausible—quite to the contrary, in fact. Letters would have been the
perfect document: written; but not kept and not official.
Letters thus were significant in recent history as recounted by Ctesias, and
probably also in reality; they are present also in the narrative of more ancient
periods, as Ctesias’ fragments show. The Indian king Staurobates sends, for
instance, a letter full of insults to Semiramis, who simply laughs when she reads
it (F 1b Lenfant = FGrH 688 F 1b = D. S. 2. 18. 1–2); a Mede, hearing by accident
the treacherous plans being hatched by the sons of Semiramis and a eunuch,
writes them down on a piece of parchment and passes the message to Semiramis
herself (ªæłÆ –Æ N ØçŁæÆ  Øæ Ø  Ø Ø Ø, F1l Lenfant = Nic.
Dam. FGrH 90 F 1); in the story of the love affair between Teritouchmes and
Roxane, letters of the king play a role in deciding Oudiastes to intervene and save
Amestris, the king’s daughter, wife of Teritouchmes (Lenfant F 15 = FGrH 688
F 15 =Phot. Bibl. 72. 43a 17–28).
But the most important instance of a letter, the only one for which we have the
text in extenso, is that mentioned in the context of the love-story between
Stryangaeus and Zarinaea.157 Zarinaea, the queen of the Sacae, falls from her
horse in battle, and the Mede Stryangaeus, although her enemy, spares her life.
Later, Stryangaeus is taken prisoner by the queen’s husband; Zarinaea asks for his
life, since he saved her, but the King refuses; she then kills her husband and frees
the Mede. Meawhile, he has fallen in love with her; but she rejects him, ostensibly
on the grounds that he is already married. Stryangaeus decides then to commit
suicide, but before doing so, writes her a letter. Besides the version of the story by
Nicolaus of Damascus (Lenfant F 8c* = FGrH 90 F 5), a papyrus probably

156
Dorati 1995: 45, with earlier bibliography in n. 56; see also the careful analysis by Lenfant (2004:
xi–xv and xix–xxii) of Ctesias’ diplomatic activity; and Binder 2008: 284–6 and 288–90.
157
Background and full discussion of the various sources for the affair in Gera 1997: 84–100; Gera
sensibly concludes that even if one can vaguely reconstruct the overall historical context, ‘the chief
figures in our story—Cydraeus, Mermerus, Astibaras, Stryangaeus and, of course, Zarinaea—are
unknown outside the Persica (or later sources dependent upon that work) and there is no evidence
that Ctesias’ characters ever existed’ (88).
 When a Letter and Why? 159

preserves Ctesias’ own text of this letter, so famous that Demetrius, in the treatise
De elocutione, cited one sentence of it: ‘I saved you, and you were saved by me, and
yet I perish because of you’ (Kªg b b  ø Æ, ŒÆd f b Ø’ K b K !ŁÅ, Kªg b
Øa b Iøº Å, De elocut. 213).158
The papyrus begins with a conversation between two men: Stryangaeus (as is
clear from the text following) and possibly a eunuch. From Nicolaus’ version, we
learn that having written his letter on a diphthera, a parchment, Stryangaeus has
the eunuch swear that he will give the diphthera to Zarinaea, once Stryangaeus is
dead, without mentioning his suicide; this request is not in the preserved part of
the papyrus, but it might have come either earlier or after the letter.159
| [ ]Æ [ ] º ÆÆł  ’ K[ ]- | Ø ‹Ø ¼ª KºØ. › ’ r- | · çæ e ªF
_ __ _ |_ [ª]æ
æH __ _ _ÆÆ
_ _ _[ª]æłø_ __ ZÆæØ- _ _| Æ Æ. ŒÆd ªæçØ· æıƪ- | ªÆE
æe
ZÆæ[Ø]Æ ÆØ oø ºªØ· _ | Kªg _ b b  ø Æ, ŒÆd _f Ø’ _K-_ | b K _ [!]ŁÅ, _ _ _ _ Kªg b_ Øa
b Iø[º] Å, ŒÆd IŒØÆ | ÆPe K Æı· P ªæ Ø f K- |  ºı åÆæ[ ] Æ ŁÆØ.
Kªg_ b_ ÆF-_ | Æ a ŒÆŒa ŒÆd e æøÆ  |  PŒ ÆPe ƒº Å, Iºº |  æø
I!º . › b_ Łe | y[] K Ø ŒØe ŒÆd d _ ŒÆd | –Æ Ø IŁæ!Ø Ø. ‹øØ b
s ¥ ºø  ºŁÅØ, º - | Æ ª a  ø Ø, ŒÆd ¼º-_ | ºÆ _ _ºE Æ IªÆŁa K_ Å _ _
ÆP- | , ‹_ Æ_ b OæªØÇ _ _ _ 
_ _ |  ºŁÅ<Ø>_ _ _ _ [x ]æ K d F, ºE- | Æ_ Œ[ÆŒa
Kæ]ªÆ _  _ e ºı- | ÆE  æææØ _ Ç
_ I!º  | ŒÆd K[æ]<Ø>ł. Œ Æ æ _ ÆØ
| b HØ K HØ ŁÆøØ. [K]ªg_| ªæ Ø _ŒÆÆæ  ÆØ b P- |_ _, _ K  ÆØ  Ø c |
ØŒÆØ][]Å På · N _b f K b [] _Œ[Æ]ØÆ K Å Æ, º[ _ _ _
_ ___ _ _ _
(POxy 2330; cf. Lenfant F 8b = FGrH 688 F 8b)160
. . . that you were leaving a curse. And he said: ‘First of all, I shall write a letter to
Zareinaea.’ And he writes: ‘Stryangaeus to Zareinaea says thus: I saved you, and you
were saved through me, but I have been ruined by you and I have killed myself; for you
were unwilling to grant me your favours. I did not of myself choose these evils and this
love, but love destroyed me. This god is common to you and all mankind. Whomever
he approaches favourably, he gives the utmost pleasures and contrives for him very
many other benefits, but whomever he comes to in anger, as with me now, having
brought to him the utmost sufferings, he finally destroys him completely and wipes
him out. I come to this conclusion from my own death. I will not curse you in any way,
but will make this most righteous prayer of you: if you acted justly by me, . . . ’
Although a greeting with åÆ æØ might have been very appropriate, in this
letter Ctesias uses the old-fashioned Herodotean letter opening (Nicolaus has the
variant ºªØ , which is just as old-fashioned); the reason will have been that
already at the end of the fifth century this opening was commonly felt to be
oriental, and thus appropriate to the situation. As for the content of the letter, it is
extraordinary: this is the first love-letter in Greek literature, doubling up as a
suicide note; and for a long time it will remain unique. It is all the more regrettable
that the other letters mentioned in Ctesias’ work are lost; comparison of the
prescripts would have been informative.

158
= Lenfant F 8a = FGrH 688 F 8a. The story is also recounted in the anonymous treatise De
mulier. 2 (Lenfant F 7 = FGrH 688 F 7); on its relationship to Ctesias’ original account see Gera 1997.
Diodorus 2. 34. 1–5 describes the reign of Zarinaea, but does not mention the love story.
159
For a detailed comparison of the papyrus and Nicolaus see Biltcliffe 1968.
160
Text based on the revised readings of Giannattasio Andria 2003; cf. Stronk 2007: 52–3. At l. 24,
the papyrus has K[æ]ł, corrected by Lenfant to K[æ]Øł. A form of KŒæ ø makes indeed more
_ _ _ Lenfant is followed by Stronk 2007:
sense in the context; _ _ _ 53–4.
160 Greek Beginnings

4 . 4 . F RO M T H E F O U R T H T O T H E S E C O ND C E N T U R Y

Herodotus, Thucydides, and Xenophon offered a continuous narrative, which

made it possible to evaluate the place and role given to documents, and in
particular letters, within that narrative. The next such extended narrative to
have survived at least in part is the work of Polybius (c.200–118 bc). This is
why, after providing a few indications of ‘trends’ in letter writing in the fragmen-
tary historians of the fourth and third centuries, I shall focus on Polybius’
Histories, whose first five books offer a continuous account, and whose work
offers us a sense of the role played by epistolary communication in the Hellenistic
period (and Hellenistic historiography).
Of the numerous historical works that were written in the fourth and third
century, not one survives. We have fragments, mostly of limited size; and because
these fragments were chosen and preserved by authors with specific interests and
agendas, they may not reflect the overall character of the work from which they
come. In such a situation, it is impossible to speak of ‘epistolary strategies’, or to
discuss the comparative frequency of various types of documents: absence of
quoted letters or of references to letters in the fragments of course need not
mean that letters or references to letters were absent from the original work. We
can thus only note presences that may provide hints of the directions taken by,
and that reflect on, letter writing in this period.
The first remarkable development is that historians now write letters which
enter into public circulation. This is part of a larger trend, a new development
specific to the fourth century, which will be discussed more fully later: the letters
of some orators (Isocrates, Demosthenes) and some philosophers (Plato, Epi-
curus) are better known, simply because they have survived; but writing open
letters is a phenomenon frequent also among those we classify as ‘historians’—a
salutary reminder of the shared roots of rhetoric and historiography.161 Address-
ing one individual person had the advantage that the author ‘did not have to
consider whether his subject and treatment were of sufficient public importance,
or seriousness, or respectability, to interest an ideal fellow citizen. He built on the
fact—or created the fiction—that there was one other person ready to be inter-
ested in his topic, and addressed that one person, writing, as personally as he liked,
about as minute a subject as he chose.’162
Among the historians we find two main types of letters, depending on the
interests of the writer: erudite ones, engaging in intellectual disputes of one sort or
another, and political ones, addressed to kings or personalities active in politics.163

161
The letters of Isocrates are discussed below, 286–92. Shared roots: Nicolai 1992; Nicolai 2004;
also Porciani 1997 for a discussion of the epistolary proem of historical works. On literary letters, cf.
Dalby 2000, who speaks of ‘a coterie of men who wrote letters to one another, letters which circulated
among others and which continued to circulate long enough for Plutarch and Athenaeus to refer to
them, five centuries later’. Dalby discusses specifically the epistolary exchange between Lynceus and
Hippolochus, Poseidippus, Apollodorus, and Diagoras. Randall’s fascinating study (2008) of the
continuities and changes in the use of letters from antiquity to the Renaissance helps put the change
in fourth-century Greece in a wider perspective.
162
Dalby 2000: 375.
163
A third type of letter, common among those Dalby (2000) calls the ‘anecdotists’, seems to have
mainly focused on describing royal dinners, wedding, processions, and similar exceptional events;
 When a Letter and Why? 161

The Attidographer Philochorus (born c.340 bc) offers two good examples of the
first type. To judge from the one fragment of each that has survived, his Letter to
Alypus and Letter to Asclepiades must have been erudite letters, small scholarly
recherché works on literary issues. Starting from a passage in Aristophanes, the
first one probably discussed the use of garlic in ancient Athens during the festival
of the Skira. The other concerned the genealogy of Hecuba.164 Neither Alypus nor
Asclepiades are known, but the erudite character of the letters is borne out by the
fact that the only references to the two men come from a lexicon and an ancient
commentary. Letters of this type are also used for (more or less polemical)
discussions of specific aspects of an event. Thus, Polybius himself states in his
Histories that he wrote a letter to the Rhodian historian Zeno, because of an
important error that the latter had made in the topography of Laconia. Polybius
continues by saying that on receipt of the letter Zeno had to admit his mistake, but
because his work was already published he could not correct it; still, although very
much grieved, he took the criticism in good part.165 This may have been a private
letter, or it may have circulated widely. But had Polybius not seen fit to mention
the exchange (and underline in this way his superior knowledge of Laconian
geography), we would not have known of it.166
Besides this kind of erudite small treatise in the form of a letter, the epistolary
medium is also used for pamphlets, to address political issues, to give advice, and
to attack enemies. Several such letters are attributed to Theopompus (otherwise
mostly known for his—lost—Epitome of Herodotus, Hellenica, and Philippica). An
epigraphical booklist from Rhodes mentions a Letter to Philip and a Letter to
Evagoras among his works;167 a Letter to Alexander, Counsels to Alexander, and
Chian Letters are mentioned in Athenaeus.168 Jacoby’s hypothesis that these
letters, some of them longer, other less so, were put together in a volume with
the overall title of Letters from Chios is very attractive: it would explain the very
similar content of passages attributed to different letters.169 Of the Letter to Philip
one fragment remains, preserved in Didymus’ commentary to Demosthenes: in it,

although such letters certainly offer material for social history, their writers are not known for also
having written ‘traditional’ historical works, and for this reason I shall not take them into account here.
164
See FGrH 328 F 89 (= Photius, s.v. æź ), for the letter to Alypus, and FGrH 328 F 91
(= schol. Eur. Hec. 1) for that to Asclepiades.
165
Plb. 16. 20: › b [Zenon] ºÆg c KØ º , ŒÆd ªf I Æ s Æ c Ł Ø Øa e
æŒøŒÆØ a ıØ, Kºı ŁÅ b ‰ Ø ºØ Æ, ØE  Pb r å,   ª c  æÆ
Æ¥ æ Ø IÆ çغçæø. See McGing 2010: 79–80.
166
Whether it circulated widely or not, the letter (or its mention) is unlikely to have made Polybius
popular with contemporary historians: Walbank 1967: 525.
167
FGrH 115 T 48 = Maiuri, Nuova silloge epigrafica (Florence 1925), no. 11, col. 1, ll. 22–3 (at 24,
the ı ıºı[ØŒe æe] | #ºÆ[æ is also mentioned). Both titles are restored, but in the case of
the Letter to Philip independent evidence (Didymus, see below) makes the restoration almost certain.
168
FGrH 115 F 252, 253 and 254 = Athen. 6. 230 ef, 13. 595 ac, and 13. 586 c. Note also FGrH 115
T 20a = Dion. Hal. ad Pomp. 6. 1, for the title Chian Letters. In his Ad Atticum 12. 40. 2 (= FGrH 115
F 251), Cicero refers in the same sentence to the books To Alexander (æe  `ºÆæ) of Aristotle
and Theopompus, adding that ‘they wrote what was both creditable to themselves and pleasing to
Alexander’. Considering the character of the extracts from the letters of Theopompus that we have, this
is slightly surprising, and makes caution in the use of the fragments even more necessary.
169
Jacoby, FGrH 2B (Kommentar), 390.
162 Greek Beginnings

Theopompus viciously attacked Hermias of Atarneus.170 In the one certain

fragment from the Letter to Alexander Theopompus reprimanded Harpalus for
the grandiose burial he gave to the courtesan Pythionice, ‘not only three times a
slave, but also three times a whore’ ( c  æ ıº Iººa ŒÆd æ æ
ÆP ); attacks on Harpalus, again for his relationship with the courtesans
Pythionice and Glycera, also figured in the so-called Letters from Chios (KØ º
 a .ØÆŒa KتæÆç Æ).171 Finally, the one fragment we have of the
Counsels to Alexander (—æe  `ºÆæ ı ıºÆ ) contains a harsh verbal
assault on another citizen of Chios, Theocritus, who is accused of having become
suddenly so rich that he now drinks only from gold or silver vessels.172 As Jacoby
comments, the comparison (or rather the contrast) with Isocrates’ letters is
certainly instructive.173 The same Theocritus who attracted the barbs of Theo-
pompus is known, apart from his political activity, for having written Anecdotes, a
History of Libya, and Marvellous Letters ( ¯ Ø ºÆ ŁÆı Æ ÆØ). The Suda, the
only ancient text to mention them, does not provide any information as to the
possible content of the Marvellous Letters; we are thus reduced to guesswork.174
These letters may have contained anecdotes, possibly on the contemporary events
at the Macedonian court (the title might have been ironic): after all, Theocritus
was famous for his anecdotes and his witty repartee (one example is mentioned
below). Finally, the trend begun by Ctesias with his love-letter found followers,
and not just in the novel: another instance of the genre is the moving story about
Ariadne narrated by a certain Paeon, a local historian of Amathus in Cyprus
(possibly active in the third century bc). According to Paeon, when Theseus
abandoned Ariadne pregnant on Cyprus, the women of the island composed
and gave her a forged letter which appeared to have been written by Theseus
(ªæ ÆÆ ºÆ a æ çæØ, ‰ F ¨Å ø ªæç ÆPB fi ).175
Letters along the lines established by Herodotus and Thucydides continue to
feature within historical works. Stories of treason revolving around letters still
appear, but now they concern not only the east, but also the west. East first: within
the very little that we have of the Hellenica Oxyrhynchia (a continuation of
Thucydides’ work, composed in the first half of the fourth century), two extremely
fragmentary passages tell of incidents in which communication through letters
seems to play an important role, in the one case involving letters exchanged
between Greeks, in the other between Persians. During the siege of an unknown
city (possibly Byzantion in 409 bc), a guard lets down a rope from the wall, and

170
FGrH 115 F 250 = Didym. in Demosth. V 21.
171
Letter to Alexander: FGrH 115 F 253 = Athen. 13. 595 ac; Letters from Chios: FGrH 115 F 254a
and b = Athen. 13.586 c and 13. 595 de. These various titles might refer to one work only; however, the
title Letters from Chios is confirmed by Dion. Hal. ad Pomp. 6.1 = FGrH 115 T 20a. On the historical
context of these letters, see Flower 1997: 258–62.
172
FGrH 115 F252 = Athen. 6. 230 ef. This work may have been the same as the Letter to Alexander.
On Theocritus, see below.
173
Jacoby, FGrH 2B (Kommentar), 390. This has of course nothing to do with the question of
whether Theopompus (and Ephorus) were students of Isocrates or not; on this issue see Nicolai 1992:
159–63 and Flower 1997: 42–62, who both draw a negative conclusion.
174
See the excellent discussion by Roller 2013, in BNJ 760 T 1 (= Suda s.v. ¨ŒæØ .E), as well
as his ‘bibliographical essay’ in the same entry.
175
FGrH/BNJ 757 F 2 = Plut. Thes. 20.
 When a Letter and Why? 163

exchanges by this means letters with someone (a Myndian?) waiting in the woods
outside; we do not know how the affair ended.176 The other passage of the
Hellenica Oxyrhynchia in which letters figure (KØ º, three times repeated)
concerns the story of how the newly nominated satrap Tithraustes eliminated
Tissaphernes with the help of Ariaeus. What role exactly the letters played in the
affair is, again, difficult to tell; but since a similar story is also told in Polyaenus’
Stratagems, we may deduce that a trick was involved.177 As for the west, treacher-
ous letters are now also one of the main features in the stories concerning the
Sicilian tyrants. A good instance is offered by the tussle between Dionysius II and
Dion over Syracuse, as narrated by the historians Timonides (a supporter of Dion
and contemporary with the events) and Timaeus of Tauromenium (c.356–260
bc), whose narratives of this affair have been preserved by Plutarch. The first
moment in the drama is an intercepted letter. When Dionysius was already getting
suspicious in c.366 bc,
a certain letter was secretly carried to him (KŒ ŁÅ Ø KØ ºc Œæ çÆ) which Dion
had written to the Carthaginian authorities, urging them, whenever they should treat
with Dionysius for peace, not to open discussions without including him, since he
would help them to arrange everything securely. [5] This letter Dionysius read to
Philistus, and after consulting with him, as Timaeus records, fooled Dion by a feigned
reconciliation. [6] That is, after modest protests and a declaration that their quarrel
was over, he led him off alone beneath the acropolis down to the sea, showed him the
letter (Ø c KØ ºc), and accused him of conspiring with the Carthaginians
against him. [7] When Dion tried to defend himself, he would not allow it, but at once
put him just as he was on a small boat, commanding the sailors to deposit him in Italy.
(Plutarch, Dion 14. 4–7 = BNJ 566 F 113)
In this instance, Dionysius’ behaviour is extremely moderate. As a result, Dion
makes a comeback, and in 357 bc he is in Syracuse, occupying all of the city
excepted Ortygia, where the garrison of Dionysius resists. At this point, the friends
of the tyrant send him letters to inform him of the arrival of Dion (KŒ Ø ŒÆa
å ¼ªªº fiH ˜Øı ø fi ªæ ÆÆ Œ ÇÆ: Dionysius is at the moment in
southern Italy)—in vain, because the courier has a bizarre mishap:
But a strange misfortune befell the man who had been sent with the letters
(ªæÆ Æçæø fi ). After he had crossed to Italy and passed through the territory of
Rhegium, and as he was hastening on to Dionysius at Caulonia, he met one of his
acquaintances who was carrying an animal that had been recently sacrificed, and after
accepting from him a portion of the flesh, went on his way with all speed. 8 But after

176
Hell. Oxy., Florence fragment, fr. C, V 2: › | b  <Ø> Kº[Łg K]Œ B | oºÅ æH b Y
Ø | ªæÆ ÆE YÅ Ææ’ KŒ ı ŒÆŁØ  [[]] Kº Æ- | [] ŒÆd Øç ºÆ[], [Ø]Æ b
æ[ ]Bł ÆPe i | (æ [fiH æø fi ª]æÆ [Æ]k [ . . . , with Bruce 1967: 47–8 (who mentions the
possibility that  ‘dumb’ or ‘silent’ may be meant here, only to reject it); McKechnie and Kern
1988: 129–31; Russell 1999: 156. What follows is lost.
177
Hell. Oxy., London fragment, fr. B, XIII, cols. 7–8. In D. S. 14. 80. 7, King Artaxerxes appoints
Tithraustes as commander with orders to arrest Tissaphernes and sends letters to the cities and the
satraps requesting that all perform whatever Tithraustes may command. Polyaen. 7. 16 has a detailed
account, very close to that of the Hellenica Oxyrhynchia, in which Artaxerxes sends Tithraustes to
arrest Tissaphernes, having given him two letters: one gives him command, the other is for Ariaeus,
ordering the latter to help Tithraustes in the arrest. Ariaeus reads the letter, and immediately proceeds.
Cf. Bruce 1967: 80–92; McKechnie and Kern 1988: 148.
164 Greek Beginnings
travelling part of the night, he was compelled by weariness to take a little sleep, and lay
down, just as he was, in a wood by the side of the road. [9] Then a wolf came to the
spot, attracted by the scent, and seizing the flesh which had been fastened to the wallet
in which the man had his letters, went off with it and the wallet too. [10] When the
man awoke and perceived what had happened, he wandered about a long time in
search of what he had lost, but could not find it, and therefore determined not to go to
the tyrant without the letters, but to run away and disappear.
(Plutarch, Dion 26. 7–10, trans. Perrin)
Plutarch does not say who is his source for this specific incident, but this story
fits very well with the role played by letters in the overall narrative—besides giving
a close insight into the difficulties of the life of a messenger. Eventually, Dionysius
hears of Dion’s return, comes back to Syracuse, and tries to get in touch with Dion
by sending envoys; he also mounts a surprise attack against Syracuse, which fails.
At this point, the tyrant moves to letters: he decides to forward to Dion letters
ostensibly written by members of Dion’s family who are in his power, but inserts
among them one of his own, written as if by Dion’s son (here, Plutarch relies on
both Timaeus and Timonides):
Heralds came down from Dionysius bringing letters to Dion from the women of his
family. There was also one addressed outside, ‘To his father, from Hipparinus’ ( Æ  q
øŁ KتªæÆ Å ῾HØ Ææd Ææ ῾$Ææ ı ); for this was the name of Dion’s
son. [2] Timaeus, it is true, says he was called Aretaeus, from his mother Arete; but on this
point at least, in my opinion, Timonides, who was a friend and fellow-soldier of Dion’s, is
rather to be trusted. Well, then, the rest of the letters were read aloud to the Syracusans
(ƃ b s ¼ººÆØ E ıæÆŒ Ø Iª! ŁÅ Æ KØ ºÆ ), and contained many
supplications and entreaties from the women; but that which purported to be from
Dion’s son the people would not allow to be opened in public. Dion, however, insisted
upon it, and opened the letter. The letter was from Dionysius, ostensibly to Dion, but in
fact to the Syracusans (q b Ææa F ˜Øı ı, E b ªæ Æ Ø æe e ˜ øÆ,
E b æª Æ Ø æe f ıæÆŒ ı Øƺª ı). Cast in a tone of entreaty and
self-justification, it was intended to foment ill-will against Dion. [3] There were remind-
ers of his eager services in the interests of the tyranny, and threats against those closest to
him—his sister, his child, and his wife; there were also serious urgings [ . . . ] that he
should seize power himself, protecting the safety of his friends and family. (32) When
these things had been read aloud, the Syracusans were not astonished (though they
should have been) at the steadfastness and great-heartedness of Dion, who was resisting
because of honour and justice such strong family obligations, but they instead began to
regard Dion with suspicion and fear, thinking he was under great compulsion to spare the
tyrant, and they were already looking to other leaders.
(Plutarch, Dion 31. 2–32. 1 = Timonides BNJ 561 F 1 = Timaeus BNJ 556 F 114; trans.
Perrin, modified).
The letter achieves its purpose of casting doubt on Dion: the Syracusans look to
another leader, Heracleides; Dion would die three years later, in 354 bc.178
Of course, stories of this kind were too good to pass over unmentioned. What is
more striking is the large amount of routine letters that are mentioned, more or
less casually. For instance, we owe to Athenaeus’ book on parasites two identical

178
For a discussion of the sources concerning the activity and the end of Dion (which include
Plato’s seventh and eighth letter) see Westlake 1983; Champion 2013a and b.
 When a Letter and Why? 165

stories, attributed to two different authorities, located in two different courts (east
and west, again) but occurring at approximately the same time. Both play on the
arrival of letters. In one version, attributed by Athenaeus to Hegesander of Delphi
(a writer of anecdotal Commentaries, active in the mid-second century bc), Philip
II announces that a letter has been brought to him from Cotys, king of Thrace,
whereupon Cleisophus the parasite exclaims ‘By the gods, excellent!’; when the
king asks back, ‘How do you know?’, the parasite praises him for the well-founded
rebuke. The other version is attributed to Timaeus: here it is the Sicilian tyrant
Dionysius who enters the room announcing that he has received letters from
his officers in Neapolis; the reactions of the parasite (Democles) and of the
king are the same: the parasite simply continues where he had left, ‘Excellent . . .
criticism!’179
Letters to or from Alexander are also often mentioned. For instance, we owe to
Plutarch’s Life of Phocion an important piece of information on Alexander’s
practice in addressing his letters:
 ˇ ªF ˜FæØ YæÅŒ, ‰ ªÆ ª  (#ºÆæ) ŒÆd ˜Ææ ı ŒæÆ Æ
IçEº H KØ ºH e .Æ æØ, ºc K ‹ ÆØ ªæÆç øŒ øØ, F b ,
u æ # Ææ, a F .Æ æØ æ Ūæı. F b ŒÆd .æÅ ƒ æÅŒ.
(Plut. Phoc. 17 = FGrH/BNJ Duris 76 F 51 = FGrH Chares 125 F 10)
For Duris has said that after he had become great and had conquered Darius, he left
out the word ‘greetings’ from his letters, except for those he wrote to Phocion. Him
alone, and Antipater as well, he addressed with the word ‘greetings’. Chares has related
this as well. (Trans. Pownall)180
Another historian, Phylarchus, criticizing the extraordinary luxury in which
Alexander and his friends lived, recounts that the king at some point even wrote to
the Ionian cities, and in particular to Chios, asking for purple, because he wanted
to dress all of his hetairoi in sea-purple clothing. The letter was read out publicly
to the population of Chios (Iƪø Ł Å b B KØ ºB . Ø). We have the
story because on this occasion Theocritus (the contemporary of Theopompus
mentioned above, and a violent opponent of Alexander) aptly quoted the Homeric
verse ººÆ æç æ ŁÆ ŒÆd EæÆ ŒæÆÆØ , ‘and him took purple death
and a harsh fate’.181
The passages stressing the importance of letter writing at the courts of the
Hellenistic kings are legion: accessibility and punctilious attention to correspond-
ence are part of the image of the ideal king presented by the Hellenistic literature

179
First story: Athen. 6.248 e; second story: Athen. 6. 250 ad = FGrH/BNJ 566 F 32. See on both
stories Dalby 2000: 377 n. 24, and 381 for important remarks on the (sometimes very uncomfortable)
closeness between historians and anecdotists.
180
Plutarch gets his information from Duris of Samos (active at the turn of the fourth century, and
an author of numerous now-lost works, among which was a historical work in at least twenty-three
books, variously called Macedonica, Histories, or Hellenica: see Pownall 2013, BNJ 76, ‘Biographical
essay’) and Chares of Mytilene (who acted as the court marshal, N ƪªº , of Alexander, and wrote a
History of Alexander in ten books). The same information is also conveyed, without source references,
in Aelian. VH 1. 25. On the story, besides Jacoby and Pownall 2013 ad loc., see Gerhard 1904.
181
FGrH 81 F 41 = Athen. 12. 539 b–540 a. Phylarchus mentioned letters in other parts of his work,
e.g. twice in connection with Cleomenes (FGrH 81 F 55 and F 59); but as both fragments are preserved
in Polybius and as anyway these are not verbatim quotes, it seems reasonable to discuss them in the
context of Polybius’ own work.
166 Greek Beginnings

on kingship. Hecataeus of Abdera, for instance, in his idealized image of Ptolemy,

emphasizes the importance of dealing promptly with correspondence.182
Finally, long letters are inserted within historical works. Here, the fragmentary
state of the works of the fourth- and third-century historians is particularly
problematic, as often we have to rely on a paraphrase and cannot know whether
a text might have been quoted. But one case stands out, that of Anaximenes.
According to the notice in the Suda (Æ 1989), Anaximenes of Lampsacus was a
didaskalos, a teacher of Alexander, whom he followed in war. The sources say that
he wrote about ( ıªæÆłÆ) ‘the ancient events of Greece and what Philip son
of Amyntas and later Alexander did’ (Paus. 6. 18. 2), that is Hellenica (from the
birth of the gods to the battle of Mantinea) followed by Philippica, and another
work, On Alexander; other kinds of writings were also attributed to him. Of all
these, apart from the Rhetorica ad Alexandrum ascribed to him, only scanty
fragments remain. He is nonetheless worth discussing here for two reasons.
First, just like Ctesias, he seems to have made a rather free use of the written
word. According to Pausanias, Anaximenes retaliated against Theopompus, with
whom he had quarrelled, by writing ‘a treatise abusing Athenians, Lacedaemo-
nians, and Thebans alike. Having imitated the style of Theopompus with perfect
accuracy, he inscribed the latter’s name upon the book and sent it round to the
cities. Though Anaximenes was the author of the treatise, hatred of Theopompus
grew throughout all of Greece.’183 The treatise became known as the Tricaranus
(one wonders whether the title æØŒæÆ may not have been meant as an allusion
to Theopompus’ own description in one of his Chian Letters of the courtesan
Pythionice as æ ıº ŒÆd æ æ), and the story finds a plausible context in
the polemical style that characterizes fourth-century historiography;184 yet no one
had hitherto gone so far as to write an invective against the three most powerful
cities of Greece and then attribute it to a rival.
The second reason why Anaximenes is worth mentioning here has to do
specifically with letters: according to the testimony of Didymus, Anaximenes
inserted a speech of Demosthenes into his historical work, which was put together
from bits and pieces of various real Demosthenic speeches.185 In Anaximenes’
historical work this speech constituted the reaction to a letter of Philip, which
Anaximenes also inserted into his work. This may be the letter preserved in the
Corpus Demosthenicum as [Demosth.] or. 12 (cf. FGrH 72 F 41), which, although

182
FGrH 264 F 25 = D. S. 1. 70. 4 (but the entire passage, D.S. 1. 70. 1–10, is fascinating). On the
ideal monarch see Walbank 1984: 77–8; for the importance of audiences, Savalli 2007; specifically on
the correspondence, Virgilio 2011: 24–30.
183
Paus. 6. 18. 5 = FGrH 72 T 6. The authorship of the Tricaranus remains uncertain; detailed
discussion in Parmeggiani 2012.
184
Pythionice: above, 162. For the new ‘polemical approach’, see Marincola 1997: 228–31.
185
This is the speech transmitted in the Corpus Demosthenicum as [Demosth.] or. 11, cf. FGrH 72
F 11a and b. Our source for this is Didymus On Demosthenes 11.7 = FGrH 72 F 11a:  Ø  ¼
Ø PŒ Ie ŒF ı çæB ŁÆØ e ºª Ø (or. XI) Œ Øø ˜Å  Łı æƪ Æ<Ø>H
KØ ıŁ. ŒÆd N d ¥ çÆ Ø `ÆØ ı r ÆØ F ¸Æ łÆŒÅF c ı ıº , ŒÆd ªaæ c K fiB
 fi Å H[ غØ]ØŒH <‹º>Å Oº ªı E ªæ Æ Ø Æ[PE K][]åŁ[Æ]Ø. See on this Canfora
1991: 70–1; Squillace 2004: 37–8 and 101–7, who suggests that also a letter of Darius (Arr. Anab. 2. 14.
1–3, Curt. 4. 1. 7–14; see also D. S. 17. 39. 1–2 for Alexander’s manipulations) and the answering letter
of Alexander (Arr. Anab. 2. 14. 5–9) may have been the work of Anaximenes; Harding 2006: 216–23;
and below, 276 and n. 40.
 When a Letter and Why? 167

probably not an original letter of Philip, is certainly closely modelled on one; in

this case, both Philip’s letter and Demosthenes’ response in the Corpus Demosthe-
nicum would have come originally from Anaximenes’ work.186 As the story of his
revenge on Theopompus shows, Anaximenes prided himself on having ‘a natural
aptitude for rhetoric and for imitating the style of rhetoricians’, as Pausanias puts
it (6. 18. 5). To find him writing letters or speeches in the style of the Macedonian
king and Demosthenes respectively is thus no surprise. But in his work we have
the earliest attestation of the insertion in a historical narrative of a long Greek
royal letter, supposedly written c.340 bc (Eumenes may already have been active
in Philip’s chancery), and presenting the standard epistolary greeting  ºØ
`
 ŁÅÆ ø fiB ıºfiB ŒÆd fiH  fiø åÆ æØ, ‘Philip to the council and the people of
Athens, greetings’. (Given Anaximenes’ record of precision in staying close to the
language of the originals, we may well believe that this formulation reflects the
authentic prescript.) The two sentences that follow the prescript also present
interesting elements. Right after the prescript, the body of the letter opens with
a causal clause introduced by KØ , as is usually also the case for city decrees:
Philip justifies his sending a letter, and somehow implies that he resorts to a
written document because of the failure of previous oral embassies ( ¯Øc
ººŒØ ı æ Ø I  ºÆ, ¥  K  ø  E ‹æŒØ ŒÆd ÆE
› ºª ÆØ, P Æ KØE Ł KØ æç , þØ Å E  łÆØ æe  A bæ
z IØŒE ŁÆØ  Çø). This positive appreciation of writing is to be expected by a
syngrapheus; whether it was shared by Philip is uncertain. In the following
sentence (the third one of the letter, including the prescript in the count), one
of the elements highlighted as crucial in discussion of epistolary writing is
mentioned: length. Philip (Anaximenes?) excuses himself for the length of his
letter, explaining that it is due to the number of accusations he has to deal with ( c
ŁÆı  Å b e BŒ B KØ ºB· ººH ªaæ Ææåø KªŒºÅ ø
IƪŒÆE K Ø bæ ±ø źH ÆØ ŒÆŁÆæH). It is difficult not to hear behind
this remark an echo of the principle expressed by the later rhetoricians on brevity
as an important characteristic of the letter. As we shall see, from Isocrates onwards
this notion—still absent in Thucydides’ work—had established itself; by now,
epistolary rules exist.

4.4.1. Mediated Long-distance Communication in Polybius

Polybius’ Histories offer in the first five books a continuous narrative of events in
the Mediterranean, with a summary coverage of the years 264 to 220 in the first
two, followed by a detailed narrative of the beginning of the second Punic war
(220–217 bc) in book 3, and a similarly detailed narrative of events in Greece and
the East for the period of the 140th Olympiad (220–217 bc) in books 4 and 5; of
the rest of his work in forty volumes only excerpts survive, some fairly extensive,

186
Non-authentic: Jacoby ad loc.; authentic: Gauger 2000: 73 and 306 n. 20. Two fragments (FGrH 328
F 55a = Dionys. Hal. Ad Amm. 11, and F 55b = Didym. in Demosth. col. 1, 67), show that Philochorus,
whom we have seen as a writer of erudite letters, summarized within his Atthis a letter of Philip to the
Athenians, as well as the speech of Demosthenes in reply, which led to the war: see Jacoby commentary ad
loc.; Harding 2006: 114–15; Canfora 1991: 70–1 assumes the existence of various versions.
168 Greek Beginnings

so that on the whole the Histories offer a reasonably reliable body of evidence for
observing both epistolary practices during the time period covered and their
historiographical representation. Two things leap to one’s attention: first, the
sheer number of references to epistolary communication, which highlights the
existence of a network covering the entire Mediterranean, east to west—even
though very few letters are actually quoted in extenso or even summarized; and
second, the relatively large number of references to forged letters as a strategy
designed to further treasonous plots.187
Let us start with Polybius’ references to epistolary communication. Letters,
messengers (ªæÆ ÆçæØ), and written exchanges between parties in general
are ubiquitous, and occur across all of the ancient Mediterranean.188 This is in
keeping with Polybius’ explicit assessment of what makes the period of time he
has chosen to investigate extraordinary. Before the 140th Olympiad, he says,
events took place in a sporadic way, being specific only to that part of the world
where they were happening; but ‘ever since this date history has been an organic
whole, and the affairs of Italy and Libya have been interlinked with those of
Greece and Asia, all leading up to one end’ (1. 3. 3–4). This notion of a historical
interweaving of events or ı ºŒ is mirrored in the swift movements (at times
over remarkable distances) of generals and troops, of embassies, and of letters
sent and received by the very protagonists of that new interconnectedness. In
most instances, these letters are simple routine exchanges. This fact is in itself
significant since it shows that written communication is by now one of the
staples of diplomacy.189 Significantly, too, the Histories display a new vocabulary
of epistolary communication: ªæÆ Æçæ, ‘letter carrier’, is first found in
Polybius; a letter (or, less frequently, its letter carrier) ‘makes [something] quite
clear’ (ØÆ Æçø).190 It is of course impossible to exclude the possibility that

187
For both text and translation of Polybius I have used Paton’s edition, revised by Walbank and
Habicht (and Olson for the unattributed fragments), 2010–12. I here build on, and develop, observa-
tions made in Pédech 1964: 377–89; Troiani 1979; Zecchini 2003; Prandi 2003; McGing 2010; Virgilio
2011: 30–1. I do not discuss here the issue of the all too real differences between Greek perception and
Roman practice; but see Ando 1999.
188
Numbers in Zecchini 2003 (based on Mauersberger’s Polybios-Lexicon): KØ º appears fifty-
eight times, ªæ Æ or ªæ ÆÆ twenty-three times, ªæÆ Æçæ eleven times. Some of these
references refer to the same exchange; on the other hand, these data do not take into account the
numerous instances in which someone ‘writes to tell’ or ‘writes and sends’ (e.g. 5. 28. 4; 13. 4. 7), which
clearly imply letters.
189
See Walbank 1972: 68–71, and on the transformations in terms of translation, migration, and
communication, which all set in at this moment, Moatti 2006 (although for what concerns communi-
cation, she focuses only on the imperial period). Diplomatic exchanges take place also of course
through the sending of envoys, mainly defined æ Ø; the more or less spontaneous ¼ªªºØ have
however now disappeared (for the very few instances see Mauersberger, Polybios-Lexicon, s.v., and
compare with the very numerous instances of æ Ø, æ  Æ). But contacts can be initiated
through letters: see below, 294–5 and n. 98.
190
On ØÆ ÆçE with the meaning of ‘give instructions per letter’ in Polybius see Walbank 1957:
144. A passage from book 15 presents Agathocles, who thanks to a forged (so Polybius) testament has
become tutor of the young king Ptolemy V, as being kept informed by letters and spies, in a parallelism
expressed by both the choice of the vocabulary and the sentence structure (Plb. 15. 29. 6: › b ªaæ
#ªÆŁŒºB, IåŁ Å æe ÆPe KØ ºB ŒÆd ŒÆÆ Œø KÆÆåŁø, ŒÆd B b KØ ºB
ªªæÆ Å æe a ı Ø Ææa F ºÅº ı ŒÆd ź Å ‹Ø Ææ ÆØ Æåø, H b
ŒÆÆ Œø ØÆ Æç ø ‹Ø æ Ø, oø . . . ). Another such terminological change, which
 When a Letter and Why? 169

such terminology might have been used already in some of the earlier—lost—
historians; nonetheless, its massive presence in Polybius at this juncture in history
is significant, as it reflects not only a specific understanding by Polybius of the
important role played by written communication, but also a change in these very
structures of communication that must have taken place in the wake of the
expansion of the Greek world following Alexander’s conquests. This change
finds a reflection in contemporary official documents: one example will suffice
here, the decree with which Lampsakos in c.195 bc honoured its citizen Hegesias
for having accepted, in a situation of crisis, to lead an embassy to Rome. Hegesias
first travelled to Greece, where he met with the Roman naval praetor (probably
Lucius Quinctius Flamininus) and with the quaestor, and obtained from them a
letter addressed to Lampsakos containing positive reassurances as to the inde-
pendence of the city; the letter was deposited in the city archive. Hegesias then
went on to Massalia, where he persuaded the council of the Six Hundred to give
him delegates that would accompany him to Rome and also a letter for the
Tolostoagian Gauls, beneficial to the city of Lampsakos. He then travelled to
Rome, where he and his fellow ambassadors presented their case to the Senate;
the Senate agreed (in writing) to the main request, and referred him for the rest to
the consul (Titus Quinctius Flamininus) and the commissioners appointed to the
deal with the affairs of Greece. Hegesias thus went to Corinth, and there received
from Titus a decree, and letters addressed to the kings.191 Clearly Hegesias’
mission included soliciting letters and other documents which could be brought
back to the city; equally clearly, these letters were considered ‘useful’ ( ı çæı Æ/
ÆØ referring to KØ º appears four times)—but the very presence and repetition
of such a statement of usefulness is intriguing: How far? In what ways?
At any rate, the decree for Hegesias shows that an envoy might, in the course of
his travels, collect numerous ‘useful’ letters. A survey of Polybius’ work that
catalogues references to epistolary communication further supports the impres-
sion that letters are ubiquitous in Polybius’ world.192 In 240 bc, the Carthaginian
mercenaries in Sardinia can plausibly be thought to have sent letters to the rebels
in Africa (1. 79. 9–10). In 230 bc, the arrival of letters from the Illyrian queen
Teuta forces her troops to return (æ  ø Ææa B  Æ ªæÆ ø, 2.
6. 4). The Hellenistic kings write letters to promise assistance: an example is the
missive Antigonus sent to the Megalopolitans in 226 bc (2. 50. 3).193 Likewise,

implies a change in practice as well (or at any rate in how Polybius views exchanges) is the disappear-
ance of figures such as the—too imprecise?—¼ªªºØ, for which see nn. 9 and 189 above. On Polybius’
(non-)use of ‘technical language’ see Bertrand and Gruenais 1981; Mari forthcoming.
191
I. Lampsakos 4 (= Syll.3 591); translation of the document in Bagnall and Derow 2004, no. 35.
Letters are mentioned at ll. 39–41: [ºÆ ŒÆd ]Ææ ÆPF KØ ºc æe e B [  H], | [m
ªf] ı çæı Æ r ÆØ ŒÆå!æØ  N [a Å ]- | [ ØÆ  H › B ]; 47–9: Œæ Æ b åæ Ø 
r ÆØ IØ![ Æ ºÆ]- | [ Ææa H] ÆŒ ø ı çæı Æ KØ ºc [bæ F  ]- | [ ı æe
e B ] H º ƪ ø ˆÆºÆH; 63–4: [IØ]- | [H ŒÆd KØ ºc º]ÆE ı çæı Æ HØ
 øØ; 75–7: æd z ŒÆd [ºÆ ª]- | [ Æ çغŁæø] ŒÆd KØ ºa æe f Æ ØºE[ - - ] |
[ªf ı çæ] Æ ÆPHØ r ÆØ ØÆ غ[ - - ].
192
Zecchini 2003: 415 singles out as exceptions the Iberian tribes—but that is all.
193
His letter is first discussed in the assembly at Megalopolis, and then the Megalopolitans present
it at the federal assembly of the Achaeans (the wording is interesting: ÆæºŁÆ f ªÆºº Æ
N e ŒØe ıºı æØ   ªæ ÆÆ E #åÆØE KØØŒ ÆØ ŒÆd ØÆ ÆçE c ‹ºÅ hØÆ F
170 Greek Beginnings

after having taken Megalopolis in 223 bc, the Spartan king Cleomenes hastily
dispatches messengers to those Megalopolitans currently in Messene; they carry a
letter, in which he offers to give back the city if they will go over to his side; the
Megalopolitans, however, do not allow the couriers to finish reading the letter (B
KØ ºB IƪØø Œ Å) and come close to stoning them (2. 61. 4–5).194 In
220 bc the Aetolians Scopas and Dorimachus send ªæÆ ÆçæØ to their
general Ariston at Cyllene (4. 9. 9). Later in that same year, Philip sends a letter
to the Aetolians (4. 26. 3), and the Aetolians write back (4. 26. 5–6). Once he has
decided to intervene in the Peloponnese, Philip sends messages (ªæ ÆÆ . . .
KÆ ºº) to the general of the Achaeans and to the cities to inform them of
when and where to meet him (4. 67. 8): the year is 219 bc. Back in Corinth, Philip
again sends his letter carriers to the allied cities (f ªæÆ Æçæı
ØÆ ºº, 5. 17. 9); because of the letter they have received from Philip
(Œ Ø  Ø a ªæ ÆÆ a Ææa F غ ı), the Messenians act rashly in
support and suffer defeat at the hands of the Spartans (5. 20).
Courtiers too write letters. When Leontius, the commander of Philips’ peltasts,
finds himself in difficulty, he repeatedly writes to Apelles, another member of the
Macedonian court, asking him to come over from Chalcis (5. 26. 2). Meanwhile,
when ambassadors from Rhodes and Chios agree to a truce with the Aetolians,
Philip writes to the allies (› b  ºØ,   a Iå, E b ı åØ
ªæÆł, ØÆ ÆçH . . . , 5. 28. 3); more or less at this same moment, other letters
sent secretly to the Aetolians by Philip’s head secretary Megaleas (Kd F
ªæÆ Æ ı, 4. 87. 8), who is actually an accomplice in Apelles’ plot, are
intercepted and found to contain treasonous advice as well as abuse of the
king—this marks the swift end of Megaleas, and of Apelles as well (5. 28. 4).
Ptolemaic Egypt and the Seleucid kings now also enter the epistolary mode,
with the false letter of Nicagoras, used by the Ptolemaic minister Sosibius against
Cleomenes (5. 38), and with the forged letters of the Seleucid minister Hermias
and of the rebel Molon (5. 50. 11–13). Antiochus himself writes official letters to
Achaeus, threatening him (5. 57–8), while Achaeus marks his assumption of the
diadem and of the royal title by writing to the cities in that capacity (ØÅ  
æØŁ ŒÆd Æ Øºf  æH Kº Å  åæÅ Æ ÇØ ŒÆd ªæçØ æe a
ºØ).195 Antiochus first decides to invade Coele-Syria because of the arrival of
letters from Theodotus, ‘an Aetolian by birth’ and a former general of Ptolemy,
who now wants to change sides and invites him (5. 61. 3: æ  ø Ææa
¨ı ªæÆ ø). Back in Greece, while Philip is at the Nemean games of
217 bc, ‘a letter carrier arrived from Macedon with the news (ÆæB KŒ
ÆŒ Æ ªæÆ Æçæ ØÆ ÆçH) that the Romans had lost a major battle

Æ Øºø, 2. 50. 10; the Megalopolitans have, as it were, imbibed the letter, exactly as when it is the
grammatophoros who ‘makes clear’).
194
A nice instance of how much royal letters and their carriers could be disliked. In this passage,
Polybius is explicitly polemicizing against Phylarchus’ version of the events, so the sending of
messengers and the attempt at stoning whoever was reading that letter must have been already in
Phylarchus; but the terminology (in particular, the use of ªæÆ Æçæı for the messengers) is most
likely Polybius’ rather than Phylarchus’.
195
There is a significant connection between the diadem, the royal title, and writing to the cities: see
Virgilio 2011: 25.
 When a Letter and Why? 171

and that Hannibal was in control of the Italian countryside’. Philip initially shares
the letter only with Demetrius of Pharos (c KØ ºc KØ, 5. 101. 6–7,
repeated also at 5. 102. 2). This is a momentous letter: it induces Demetrius to
counsel Philip to subdue Illyria and then attack Italy. As part of this strategic
reorientation, Philip looks into arranging a peace with the Aetolians and sends
letter carriers to the allied cities (æe b a ı Æå Æ ºØ ªæÆ Æçæı
KÆ غ) to invite representatives to participate in the discussion of the terms
(5. 102. 8).
The sixth book contains Polybius’ account of the Roman constitution. All the
remaining books have only survived in fragments. Still, these fragments contain
enough reports of letters and letter carriers to make it clear that the density of
written exchanges did not diminish, and that everyone is involved in such
exchanges. A fragment of book 7 (14b) shows that forged letters (KØ ºc
Ø ŒıÆ Å) can be used for dirty work in Syracuse.196 In book 8, epistolary
intrigue on the part of the Cretans Bolis and Cambylus links the Ptolemaic and the
Seleucid courts. At the instigation of the Ptolemaic minister Sosibius, they get in
touch with friends of Achaeus, from whom they receive a letter. But they then
decide to betray Achaeus to Antiochus, and, with the approval of Antiochus,
receive letters in cipher ( ıŁÅ ÆØŒa ªæ ÆÆ, 8. 15. 9, 16. 9) from Achaeus’
friends; Achaeus’ trust in the letters of his friends Nicomachus and Melancomas,
written in the cipher they used to employ, will be his undoing (8. 17).197
Instances of normal, straightforward letters continue to appear in the fragmen-
tary books as well. Back in the West, Hannibal uses a letter to inform the Capuans
of his sudden decision to appear in front of Rome and thus force the Romans to
abandon the siege of their city; Polybius adds that because of the importance of
this message, great care was taken to ensure its safe delivery, B H ªæÆ ø
I çƺ Æ (9. 5. 1). The letter succeeds in its purpose, since on receiving it the
Capuans are reassured (9. 5. 6). The Roman senate keeps in touch with generals
through letters, as is shown by numerous examples, such as the letter received by
Scipio in 203/202 while he is outside Carthage in which he is informed that the
Senate has ratified the treaty (15. 1. 4: ¼æØ ªaæ wŒ fiH —º ø fi ªæ ÆÆ
ØÆ ÆçFÆ æd H æØæÅ ø).198
Similarly, Hellenistic kings send letters to the Greek cities: after meeting with
the Romans in 172 bc, Perseus will send identical letters to various Greek states,
detailing the arguments on both sides, trying in this way both to assert his superior
position and to sound out the others’ intentions (27. 4. 1).199 When the ambas-
sadors sent by the Ptolemies in 169/168 to ask for military help from the Achaeans
realize that their request will fail, they give to the magistrates a letter from the

196
In 213 bc, after Hieronymus’ death, Syracuse is divided between a pro-Roman and an anti-
Roman party; Hippocrates’ forged letter, handed over to some Cretans, but meant to be found,
represented an attempt to discredit the Syracusan authorities by presenting them as pro-Roman.
Livy 24. 31. 6 supplements the Polybian fragments: Hippocrates is presented as giving a public reading
of the forged letter, of which Livy quotes the address: ‘The praetors of Syracuse to the consul Marcellus’
(praetores Syracusani consuli Marcello); the rest is summarized in indirect speech. The letter achieves
the desired effect: the men rush to the arms.
197
On the letters in cipher (also KØ ºa ªªæÆ Æ ıŁÅ ÆØŒH) see Walbank 1967: 94.
198
Other examples in Livy, e.g. 23. 27. 10 for the letter of the Scipiones to Rome.
199
Cf. Livy 42. 46, listing the cities.
172 Greek Beginnings

kings ‘which they had ready ( Æ å)’, asking the Achaeans to send at
least Lycortas and Polybius (29. 25. 7). And even the Gerrhaei on the Persic gulf
send a letter to Antiochus during his expedition in the Upper Satrapies in 205 bc,
asking the king ‘not to abolish the gifts the gods had bestowed on them, perpetual
peace and freedom’; the letter is translated for the king (æ ÅıŁ Å ƒ B
KØ ºB), who grants the request (13. 9. 4).
As Zecchini points out, ‘this meticulous and almost maniacal attention to the
instrument of communication, when the use of the epistolary form must have
seemed already obvious to readers of Polybius, has a significance: it reflects a
world that fully adhered to the culture of writing’.200 Polybius actually appears as a
direct participant in the epistolary communication displayed in the Histories: he
writes, as we saw, to the Rhodian historian Zeno, to inform him of his topograph-
ical mistakes; and he pens a note (whose text he quotes) to remind Demetrius of
Syria of the plan for his escape.201
Even though everyone uses letters in Polybius’ world, not everyone uses them in
the same way. This brings me to my second point, the frequent use of letters in
political intrigue. The reports of generals to their cities, such as the Rhodian one
discussed below (176), are straightforward documents; so too are most of the
letters from kings (although Philip V is not above making use of epistolary
stratagems; and Polybius sternly comments, ‘Philip became addicted to that kind
of treacherous dealing which no one indeed would say in any way became a king,
but which some maintain to be necessary in practical politics, owing to the present
prevalence of treachery’, 13. 3. 1).202 But numerous letters, instead of facilitating
communication, distort it. As we have seen, writers of forged letters can be found
east and west. The first instance of the misuse of epistolary communication
occurs early in the Histories, in the camp of the rebellious Carthaginian mercenar-
ies (1. 79. 9–10). The leaders, Matho, Spendius, and Autharitus the Gaul, want to
excite the hatred of the troops against the Carthaginians. They arrange for a

200
Zecchini 2003: 415. Part and parcel of this is the fact that the opposition logos/ergon has now
become an opposition grammata/erga. So in the fragment at 9. 20. 3, an appeal for help, to be dated
c.210 bc: Yæ s  ºÆØ c E ªæ Æ Ø , Iººa ŒÆd E æªØ ÅæE c æe ÆPf
ı Æå Æ . . . ; so also at 11. 5. 2–3, in the speech of Thrasycrates of Rhodes, where the antithesis is
twice repeated. At 4. 24. 5, remonstrances in words or in writing are opposed by Philip V to enforcing
redress.
201
The note to Demetrius inscribed on a ØŒØ or writing tablet, on which Polybius puts his
seal, is mentioned at 41. 13. 8–14. 1. (It led to Demetrius’ escape in 163/162.) But Polybius is anyway an
author very much present within the Histories; and of course his own personal trajectory epitomizes the
idea of mobility.
202
Polybius’ text here is very fragmentary, but Polyaen. 5. 17 preserves a very Herodotean stratagem
(strongly reminiscent of the story of Zopyrus, Hdt. 3. 150–60), which must have taken place in this
context: Heracleides, one of the friends of the king, mutilates himself and then runs away from the
palace of the king to Rhodes, where he asks for protection from the king, whom he prevented
(allegedly) from making an unjust war against the Rhodians. As proof Heracleides shows a letter
from Philip, addressed to the Cretans, declaring his intention of launching war against the Rhodians.
The Rhodians thus welcome him. Heracleides proceeds to set fire to their dockyards, destroy their navy,
and go back to Macedon, where he becomes one of the first friends of Philip. Heracleides is mentioned
in Polybius as having received the order to destroy the Rhodian fleet (Plb. 13. 4); the anecdote thus
probably goes back to him. If so, Prandi 2003: 380 simplifies excessively when she writes that the Greek
and Macedonian letters are official ones, whereas those concerning Syria and Egypt, meant for a limited
circulation, are almost always forgeries: the Antigonids could also make use of forged official letters.
 When a Letter and Why? 173

first false letter to come, as if from Sardinia, and they produce the courier and his
letter at the assembly; then, a second courier comes, this time from Tunis,
reinforcing the message, with the result that the mercenaries cut Gisgo and the
other Carthaginian prisoners to pieces. The content of the letters is very briefly
summarized, but their effect is deadly. The next cluster of forged letters concerns
the Hellenistic courts of Macedon, Egypt, and Syria. At the beginning of book 4,
Polybius emphasized the fact that the almost simultaneous accession to the
throne, between 223 and 221 bc, of Philip V in Macedon, Ptolemy IV Philopator
in Egypt, and Antiochus III in Syria marked a watershed (4. 2); books 4 and 5 are
dedicated to introducing the three monarchs and their courts.203 In all three
courts, real or forged letters circulate among courtiers, again with deadly results.
The narrative begins with Philip V and the Antigonid court. As we already had
occasion to note, Megaleas’ treasonous letters to the Aetolians are intercepted—
this results in Megaleas’ death, and in the demise of Apelles, who had been behind
these intrigues from the start (5. 28).204 Then, the Lagid court comes into focus.
Sosibius, the powerful minister of Ptolemy, manages to eliminate Cleomenes of
Sparta by bribing a Messenian, Nicagoras, and convincing him to write a letter, to
be sent to Sosibius himself, accusing Cleomenes of intending to revolt against the
king (5. 37–8). Right after this affair, the focus shifts to the Seleucids. Their court is
the one most beset by forged letter writing, a situation brought about by the fact
that the power is fragmented, with a young king, Antiochus III, and two usurpers,
Molon in the upper satrapies and Achaeus in Asia Minor. Each one of these power
centres is involved, actively or passively, in treasonous epistolary exchanges. In
order to reinforce his hold over the king, the powerful Seleucid minister Hermias
prepares a forged letter, as if from Achaeus (in 223 bc, when this is happening,
Achaeus is still loyal), showing that Ptolemy was ready to support Achaeus in the
attempt to assume the royal title. He then presents the letter to the king to prove
his point that Achaeus is going to secede (Plb. 5. 42. 7: KØ ºc º Æ ‰ Ææ’
#åÆØF ØÆ ƺ Å æ ªŒ fiH Æ ØºE, ØÆ ÆçF Æ ‹Ø —º ÆE
ÆPe ÆæƌƺE Œº.). The king accepts that what is written is true (› b
Æ Øºf Ø  Æ E ªæÆç Ø . . . , 5. 42. 8) and makes his decisions
accordingly.205 Meanwhile, Molon gains the full support of his officers by holding
out the prospect of booty, and by showing them false letters full of menaces
and threats, as if from Antiochus, so as to thwart any thoughts of reconciliation
(5. 43. 1). Finally, when Hermias realizes that he must rid himself of the successful
general Epigenes, he devises a Palamedian plot. He forges a letter, as if from the
usurper Molon, and persuades one of Epigenes’ slaves to put it among the other
papers of Epigenes; he then sends a certain Alexis, the commander of the fortress
of Apamea, who is privy to the plot, to ask Epigenes whether he has received any
letters from Molon. On Epigenes’ denial, Alexis enters the room and looks among

203
On the ‘subtlety of presentation’ and careful design of books 4 and 5 see McGing 2010: 97–127.
204
Polybius does not express any opinion as to the authenticity of the letters (and anyway he
dislikes Megaleas and the courtiers around Apelles), but I find fascinating Thornton’s idea that the
letters might have been forged in order to ruin Megaleas: Thornton 2002: 439. On the background (the
struggle within Philip V’s court) see the comments of Thornton 2002: 419, with further bibliography.
205
Note that according to Walbank 1957: 573 and Schmitt 1964: 161–4 the letter might have been
authentic. This makes Polybius’ epistolary choices all the more interesting.
174 Greek Beginnings

the papers—and with the letter in hand, he immediately kills Epigenes, depriving
the latter of the chance to defend himself.206
These books focus on intrigue within the Hellenistic courts. As McGing
comments: ‘There seems to have been a rash of forging. It was presumably easy
to do. I am not suggesting that it did not happen in any particular case, but
I wonder if it is not also something of a commonplace: it is what evil courtiers
do.’207 Indeed: these two books of Polybius offer a very specific angle on the life of
the Hellenistic court. It cannot be coincidence that they are followed by a book
devoted to the Roman constitution, which is devoid of such intrigues. This is in
part due to the topic, of course, but also highlights the contrast between the
Hellenistic courts and Rome.208 Likewise, there are relatively few explicit reports
of letters in the first three books, although these do contain references to long-
distance communication (expressed through forms of  ø, ‘to send’, people or
messages or news, which, when they arrive, ‘show clearly’, ØÆ Æçø, the same
term used for the letters). Narrative construction is clearly at work here: Polybius
(as it were) obliquely operates with an ‘ethics of epistolary communication’, which
he uses to demarcate cultural differences. The courtiers of the Antigonid, Lagid,
and Seleucid courts write forged or treasonous letters; such letters also circulate in
‘post-Hiero’ Syracuse; they are used by mercenaries, and in general by adven-
turers. Conversely, besides the Carthaginians and marginal populations such as
the Gerrhaei, the groups that are not directly involved in internal power struggles
by means of forged letters (at least in what survives of Polybius) are the Romans
and the Greek cities, even though they employ letters as a means of communi-
cation. Some caution is, however, necessary: a modern reader encounters two
difficulties here. First, the participation of Rome in epistolary exchanges becomes
more intense in the later books of Polybius, but these are precisely the ones which
are very fragmentary. And then one must allow for the possibility that the way in
which Polybius looked at the protagonists may have changed over time.209

206
Plb. 5. 50. 11–13: (Epigenes) ªæłÆ ‰ Ææa ºø I ƺ Å KØ ºc æe e
 ¯ تÅ  ŁØ Øa H KŒ ı Æ ø Kº Ø ªºÆØ łıåƪøª Æ N ªŒÆÆ æe e
 ¯ تÅ ŒÆÆ EÆØ c KØ ºc E KŒ ı ªæ Æ Ø. y ª ı ÆæB PŁø @ºØ, ŒÆd
ØÅæ!Æ e  ¯ تÅ ØÆ KØ ºa ŒŒ Ø ÆØ Ææa F ºø. F ’ IØ ı ØŒæH
KæıAfi XØ. Æåf b ÆæØ ºŁg yæ c KØ º ,fi w åæÅ   Içæ fi B ÆæÆåæB Æ e  ¯ تÅ
Iί.
207
McGing 2010: 118. Forged documents appear also in the later books: a notable instance is the
forged (so Polybius) testament with which Sosibius and Agathocles proclaim themselves tutors of the
young Ptolemy V, Plb. 15. 25. 4; other examples in Prandi 2003; Zecchini 2003.
208
Troiani 1979; Virgilio 2011: 30.
209
Cf. Walbank 1985b [1977]. Other Roman letters (all reasonably straightforward): e.g. the letter
from Lucius and Publius Scipiones sent to Lucius Aemilius Regulus and Eumenes, then in Samos, 21. 8.
2; the letter from Rome ordering the liberation of the Aetolian envoys held by Charadrus, 21. 26. 77; the
letter sent by Gnaeus in the winter of 189/188 to Quintus Fabius Labeo after the treaty between
Antiochus and the Romans, 21. 44. 3; the letter by Marcus Lepidus to the Achaeans, 22. 3. 1; the arrival
of Lacedaemonian exiles with a letter from the senate to the Achaeans, 24. 2. 1 (this same letter is
declared to have been written by the senate only because of the importuning of the exiles by Bippus and
other ambassadors returning from Rome, with the result that the Achaeans take no account of it, 24. 2.
4—a decision which will have grave consequences); the letter sent by Appius Cento from Epirus to the
Achaeans asking for troops, 28. 13. 6–8. And these are fragmentary books.
 When a Letter and Why? 175

The Greek perception of Roman practice in the use of letters for interstate
relations comes out clearly from an incident narrated in book 27. We are in 171
bc: the Roman praetor Lucretius Gallus, at anchor in Cephallenia, sends a letter to
the Rhodians through a certain Socrates, a gymnastic trainer, asking for ships. As
the letter arrives and is presented for discussion, the majority initially tends
towards a positive answer. But some Rhodians who are dissatisfied with the favour
shown to Rome try to shift the mood of the assembly, by—among other things—
disparaging the letter (c KØ ºc Ø ıæ). They pretend that it comes not
from Rome but from Eumenes, who is trying to force them into an unnecessary
war. As proof they adduce ‘the low station of the man who had arrived bearing the
letter, the Romans not being in the habit of proceeding thus, but, as regards their
communications on such matters, employing excessive care and ceremony. They
said this, well knowing Lucretius to be the author of the letter’ (Plb. 27. 7. 1–10).
This is interesting on two counts: first, the politicians (and their public) evince no
qualms in ascribing to a Hellenistic king, Eumenes in this instance, the idea of
making use of a forged a letter for his own purposes; second, there is a general
sense of the seriousness with which the Romans employ written communica-
tion.210 At the same time, the Romans too are not beyond ‘tweaking’ reality: the
letter sent not much later by Marcius Philippus to the Achaeans in 169/168 bc
(Plb. 29. 25) contains a statement which Polybius says that he knew presented a
distorted view of the situation even as the letter was being read, since indeed the
Romans had sent Numisius to reconcile the Ptolemies and Antiochus IV, but
Numisius had failed in his attempt.211
As for the Greek cities, they are certainly aware of letters, and not just as
recipients of them. Cities make use of letters for internal communications: for
example, Polybius twice cites dispatches of Rhodian naval admirals; cities, or at
least the leagues, can also utilize letters to organize a complex network, as is shown
in the story narrated in detail in book 16 of Philopoemen’s expedition against
Tegea (16. 36–7). The year is 200 bc. Philopoemen, the strategos of the Achaean
league, is organizing a march of all the allies towards Tegea. But because the spies
of the tyrant Nabis of Sparta are everywhere, he makes use of the following
stratagem. He writes letters to all the cities in the alliance, and sends them first
to the cities farthest from Tegea, in such a way that each city would also receive the
(sealed) letters addressed to the other cities located between them and Tegea. In
the first group of letters he had written: ‘On receiving this letter you will make all
of military age assemble at once in the market-place for five days, with their
equipment, provisions for five days, and money for as many. As soon as all those
present are collected, take them with you to the next city, and on arrival there
you will hand the letter addressed to it to their commanding officer and obey

210
This emerges also from another, earlier episode, the refusal of the Achaeans to acquiesce in
Caecilius’ request to summon the assembly, because ‘they were prevented from doing so by the laws, as
Caecilius was neither the bearer of letters from the senate nor would he show to their magistrates his
written instructions’ (  ªæ ÆÆ çæØ ÆPe Ææa (B) ıªŒº ı  a Kºa Kªªæı
KŁºØ FÆØ E ¼æåı Ø, 22. 12. 7, cf. 22. 10. 10–12): this is a subterfuge, of course, but written
documents are important. See Prandi 2003: 387–9, who emphasizes the importance of ‘writtenness’ in
Polybius.
211
But as Zecchini 2003: 417 stresses, Polybius here is more interested in his own personal reaction
to the letter when it was read than in the letter itself.
176 Greek Beginnings

the instructions contained in it’ (Plb. 16. 36. 3–4). Each letter contains the name of
the next city to be reached; in this way, no one can know the final destination
of the march. Moreover, Polybius adds, the letters had been sent out timed in such
a way that everyone would converge on Tegea at the same moment. This is one of
the rare instances in which Polybius cites the exact content of a letter; he must
have been deeply impressed with Philopoemen’s stratagem.212 Another instance
of a letter sent by a Greek community is when Andriscus (the false Philip) makes
himself master of Macedonia. The Thessalians then send letters and ambassadors
to the Achaeans, asking for help (36. 10. 5, in c.149 bc). However, apart from such
instances, letters from Greek cities do not seem to play a significant role in the
Histories.
Thornton is thus probably right when he states that the bleak representation of
life at the Hellenistic court also reflects Polybius’ loyalty to the republican and
democratic traditions of the Greek poleis.213 At the same time, Greek individuals
who are in all senses still part of the world of the Greek poleis may be deeply
involved with the kings: after the defeat of Perseus, the letters that the king had
exchanged with Deinon and Polyaratus of Rhodes—both the letters of Perseus to
them and theirs to him, Polybius says—are intercepted and made public (H
ªæÆ ø ƺøŒø ŒÆd çøØ ø).214
Some further observations may be added, to complete the picture of how
Polybius employs epistolary communication in his work. Two Roman letters,
both linked to the Scipiones, belong to a category in themselves. In book 10,
Polybius mentions an autobiographical letter of Scipio Africanus addressed to
Philip V (10. 9. 3); his purpose in referring to this document is to support his
interpretation of the true character and the forethought of Scipio against those that
regard Scipio’s successes as dependent on the gods and on fortune.215 Clearly,
Polybius is here using the letter as a document to support a specific point of view
or interpretation. Similarly in the context of a discussion of the naval battle of Lade
(c.200 bc), Polybius refers to the dispatch sent home by the generals, which he says
was kept in the prytaneum, the official archive, at Rhodes (B KØ ºB Ø
 Å K fiH æıÆ ø fi ). He cites this as evidence against Zeno and Anti-
sthenes’ interpretation of the events (16. 15. 8). In this second case, it is possible,
but not necessary to suppose that he saw the letter;216 but in both cases, the letter
is used as a document to provide evidence for an argument, something that is

212
As Manuela Mari points out to me (per litteras), in Polybius’ world a good statesman must be
also a military commander; but a good military commander should be able to make skilful use of
communication, including written communication. Much earlier, Themistocles had also been able to
play with both written and oral communication (see above).
213
Thornton 2002: 439; see also 434.
214
Plb. 30. 8. Interestingly, they were written in cipher; but Polybius does not trouble himself with
details of the cipher, or with the exact content.
215
Walbank 1967: 204 reasonably suggests that Polybius had access to it through Aemilianus: ‘for
had published versions circulated in Rome, Cicero would surely have known of them, whereas he writes
that nothing remains of such an ingenium. (off. 3. 4)’. This was therefore a ‘private’, personal letter,
which never circulated. It was probably written by Scipio after he and Philip V became friends in
190 bc. Cf. the commentary of Beck 2013, ad BNJ 232 F 1.
216
See Pédech 1964: 379–80, Walbank 1967: 520, and the ample discussion in Zecchini 2003: 418–
19. The same applies to another dispatch sent by a Rhodian general, Theophiliscus, right after the battle
of Chios, and mentioned slightly earlier at 16. 9. 1.
 When a Letter and Why? 177

rarely the case elsewhere.217 Finally, Polybius gives a long and very detailed
summary of the letter addressed by the two Scipiones to Prusias, the king of
Bithynia, in 190 bc (21. 11). This is an important document. It gives an outline of
the Roman policy towards the Hellenistic kings, mentioning very specific cases,
both in the west (Andobales and Colichas in Spain, Massanissa in Africa, and
Pleuratus in Illyria, all examples of kings whom Rome made greater than they
were) and in Greece (Philip V and Nabis, both of whom were not crushed by the
Romans after their defeat). If Polybius decided to dedicate so much space to this
letter and make it the turning point in Prusias’ political strategy, it may have been
done as a gesture towards the two Scipiones aimed at highlighting their diplomatic
ability, and because the text was available to him in the archive of Scipio Aemi-
lianus. We are, however, dealing with a unique instance.218
Polybius’ Histories thus reflect a world in which letters circulate; but what
primarily matters is that people are part of a larger network of communication.
While in the first part of the Histories the majority of letters concern the Hellenis-
tic kings and their courts, as the Histories advance more and more letters from
Rome appear. Most are official ones, although there are also some from individ-
uals (e.g. the letter of Scipio Africanus to Philip V). The Histories reveal that
private individuals in Greek cities joined in the power game, and might write
letters in cipher (to which kings replied in kind); on the other hand, official letters
from Greek cities (as opposed to internal reports) remained extremely rare, and
appear to have been straightforward in their purposes.
The content of most letters is not deemed important enough to be quoted but is
usually simply summarized.219 They are thus not significant qua letters: an oral
message would have accomplished the same purpose. What matters is that the
message is conveyed. It is then all the more interesting that Polybius insists on
stressing that these are letters: writing has by now a specific value of its own
(ØŁÆæåE E KªªªæÆ Ø, as Philopoemen had stated). And yet there is a
tension here, since in quite a few instances the writing proves mendacious, or
simply not adjusted to reality.220
At the same time certain epistolary themes, such as the use of forged letters or
treasonous arrangements conducted through letters, are used by Polybius to
highlight patterns of behaviour that he disapproves of. The clusters of treasonous
letters at specific points in the narrative carry implications that go beyond their
contents: such treasonous letters help construct a specific image of the Hellenistic
courts, as well as of the difficulties into which cities might run when divided
between parties. In Polybius, good leaders, good generals, good kings should not

217
Pédech 1964: 387–9; Zecchini 2003: 418–20.
218
Detailed discussion in Zecchini 2003: 419–22. References to other letters of the Scipiones and
wider discussion of their archive in Pédech 1964: 380–2.
219
Possibly in agreement with Polybius’ expressly stated policy as concerns speeches, to give their
main point: Walbank 1985a [1965]: 248–9, 252; Usher 2009.
220
Mendacious: e.g. the forged letter in which Antiochus trusts (› b Æ Øºf Ø  Æ E
ªæÆç Ø), 5. 42. 8; the truthful but misused letters of Achaeus’ friends, 8. 17; the false will of
Ptolemy IV Philopator, 15. 25. 4. See Prandi 2003: 386–9 for a larger contextualization, involving other
types of—official—documents.
178 Greek Beginnings

attract plots;221 but when the three main Hellenistic kingdoms are introduced for
the first time in the fourth and fifth books, they are marked by plotting, and a
plotting that furthermore utilizes epistolary means. Some Herodotean themes
thus resonate strongly in Polybius’ treatment of epistolary communication.222
How far the picture presented by the fourth and fifth book would have changed in
the rest of the Histories is difficult to tell because of the fragmentary state of books
6–40. It is, however, notable that those who deal in forgeries and treason are
mainly the courtiers; the kings may be aware of what is happening, but usually (in
Polybius’ presentation) they do not have a direct hand in the affair. On the
contrary, Polybius characterizes as truly royal the decision of Philip V, on the
night following his defeat at Cynoscephalae, to send one of his aides-de-camp to
Larisa to destroy and burn the royal correspondence (18. 33. 1–3). Philip was
‘acting like a true king in not forgetting his duty even in the hour of disaster’
(ØH æª Æ Æ ØºØŒ e Å K E ØE º ŁÅ ØE ŁÆØ F ŒÆŁ Œ),
for he was well aware of the fact that these documents would give the Romans
much material against himself and his friends if they fell into their hands.

4 . 5 . C ON C L U S I O N

I began with a contrastive analysis of oral and written communication in Herod-

otus. In his Histories, oral and written messages are couched in exactly the same
terms. Near Eastern kings and tyrants show a marked tendency towards the use of
epistolary communication, in contrast to the Greek habit of sending messengers
(¼ªªºØ or Œ æıŒ). Most of the letters mentioned by Herodotus are recorded
not because of their import, but because of external elements such as a peculiar
support (the tattooed head of a slave, a wooden tablet covered with wax), their
unusual means of conveyance (the letter sent to Cyrus within a hare, a note
attached to an arrow), or their falsity.
In Thucydides, the number of letters is still limited. A good half of them
concern communication with Persia. The others belong to the Greek world;
more specifically, they originate with the Athenians and the Spartans (e.g. the
long letter of Nicias and the flurry of written exchanges in book 8). I have
suggested that these letters are either private letters by individuals seeking
power or advantage (which applies also to some of those concerning Persia), or
internal communications, such as dispatches from generals to their cities—a
practice for which we have evidence also from other sources. There is a specific
rationale in Nicias’ choice of sending a letter; otherwise, the most frequent reason
for an epistolary choice seems to be plotting. Thus, the relatively high number of
letters in Book 8 could be interpreted as a symptom of the unfinished state of that
book, but also as a sign of a change: while in the first part of the Peloponnesian

221
McGing 2010: 96–7. Significantly, the rift between the two sons of Philip V, announced early on
in the Histories, will cause the death of Demetrius (in c.180 bc), also through a forged letter sealed with
a counterfeit of T. Quinctius’ seal (Livy 40. 23–4, with Walbank 1979: 254 and 257).
222
See McGing 2010: 52–8 (with further reference) for the presence of a Herodotean influence.
McGing does not, however, discuss communication.
 When a Letter and Why? 179

War communication is mainly face-to-face open speech or formalized communi-

cation through heralds, the second part is marked by the emergence of a dysfunc-
tional pattern of written communication that favours secrecy and trouble-making
over direct interaction.
The work of Xenophon does not mark any significant change in the treatment
of written communication. Relatively few letters are mentioned in the Hellenica,
and as in Thucydides, these letters concern only Sparta, Athens, and Persia. The
Agesilaus offers a significant statement of the difference between private and
official letter writing. As for Xenophon’s works on Persia, letters and written
communication figure more prominently, in line with the traditional Greek
understanding of Persia; again, epistolary deviousness figures side by side with
examples of letters that are straightforward. With Ctesias, a change is noticeable,
and not just because he presents us with the first instance of a suicide-note-cum-
love-letter. Ctesias mentions in his work epistolary exchanges between Persians
and Greeks, Athenians and Spartans, in the context of the power struggle in the
Aegean; Photius’ summary seems to imply a remarkable increase in the volume of
written communication.
What can be gathered from the fragmentary remains of historians active
between the fourth and the second century confirms the impression of a change,
which can be appreciated in full in what remains of Polybius and which, with the
appearance of ªæÆ ÆçæØ, extends even to the choice of vocabulary. In
Polybius letters are ubiquitous: everyone writes them. Conversely, and strikingly,
the time-honoured Herodotean oral messengers, ¼ªªºØ, have all but disap-
peared. What has also disappeared is the Herodotean interest in the material
support on which letters were written as one of the possibilities for tricking
unintended addressees, an interest that speaks for a growing awareness of the
possibilities of a means of communication that was still relatively new and
untested. Polybius’ letters (at least, those we have) are all on papyrus rather
than on, say, the wood of a tablet covered with wax. It is now the text itself, rather
than the support, that gets encrypted: in at least two instances, the narrative
concerning Achaeus and that concerning Perseus, we see issues of life and death
discussed through the employment of letters written in cipher. The long excursus
of book 10. 43–7, where Polybius exposes a system he himself has perfected, attests
to his interest in encrypted long-distance communication: it is thus all the more
strange that although he mentions encrypted letters (and in one case, states that
they were ‘brought to light’), he makes no attempt at describing the specific type of
cipher used. Possibly such encrypted exchanges were banal—or Polybius may
only have heard of the letters, but not seen them. What remains is a sense that a
devious use of writing marks the Hellenistic courts; but this is a pragmatic
assessment, not an ideological one. To put it differently, writing devious letters
is one of the symptoms of the plotting that afflicts the Hellenistic courtiers, as is
forging documents; but other parties also make use of written communication,
and although there is still a slight tension, the written document has by now
imposed itself as the source to be trusted. Polybius showcases the treacherous
purposes to which writing can be put; but he also emphasizes the potential of the
written document to function as a guarantee of authenticity and power.
 Part II
Letter Writing and the Polis
 5
Writing and Letter Writing
on the Athenian Dramatic Stage

The evidence discussed in the preceding chapters concerned the Greek world as a
whole; it is time to introduce geographical distinctions, and look at the behaviour
of specific poleis. As with so many aspects of ancient Greek culture, the polis that
provides us with the most evidence about letter writing is Athens. It is in fact the
only place where it is possible to study perceptions about letter writing and its
representations between the fifth and fourth century bc, grounded in documents,
literary or otherwise, that mainly address or concern its politai. In the following
two chapters, therefore, the focus will be resolutely Athenocentric. We shall first
look at drama, and then at the way the orators deal with letter writing. A third
chapter, based this time mainly on epigraphical evidence, will allow us to context-
ualize the findings within the wider Greek world.

5 . 1 . T H E A T RI C A L L E T T E RS : TR A G E D Y A N D S A T YR - P L A Y

From the very beginning, images of, and references to, writing are part of drama.1
Both the plays themselves and the archaeological record attest to the impact of
writing on drama. A good instance of the latter is an intriguing Attic vase, a hydria
dated to c.470 bc. On the vase Hermes, the messenger of the gods, appears holding
in one hand a folded writing tablet and his kerykeion, while with the other hand he
is dragging two reluctant old chorusmen, their mouths open in song, towards
Dionysus.2 The chorusmen wear corselets (but no other armour) and wreaths;
Dionysus makes a welcoming gesture towards them. The writing-tablet figures in
a rather prominent position, between the two gods: clearly, the image is making a
special point of the association between the chorusmen’s performance and
writing, an association that will be taken for granted in later documents (for

1
Segal (1985; 1986: 75–83) has suggested that alphabetic writing played an important role in the
move from a predominantly sound-directed, oral mode of orientation (and from epic recitation) to a
visually, spatially organized mode (and thus to drama); Wise 1998 offers a more radical version of the
argument. References to writing in drama also reflect the new self-awareness of the dramatists.
2
Hydria attributed to the Pan painter, now in St Petersburg, Hermitage B 201, ARV2 555. 95; see
Siebert 1990, no. 892; Gasparri 1986, no. 846 and pl. 401; Nieddu 2004: 28–9; Hall 2006: 46–7; and
especially Green 1994: 81–4 and 1995.
184 Letter Writing and the Polis

instance, on the ‘Pronomos vase’ of c.400 bc, a volute-crater celebrating a dra-

matic victory, the poet Demetrios holds nonchalantly a roll).3 What is the
meaning of the tablet? And why a tablet rather than a papyrus roll? The latter
would have been more appropriate for a text of the length of a tragedy. Among the
various possibilities canvassed by Green two stand out. The tablet might have
symbolized writing more readily than a roll in the eyes of the audience: the image
of writing on the tablets of the mind is frequently used as a metaphor for firm
remembrance in drama. Alternatively, the painter may be reflecting contemporary
practice: the tablet may represent a ‘temporary’, partial script. But as Green
acknowledges, Hermes and his tablet might also refer to something quite different,
as for instance the votes of the judges.4 Thus, this vase shows that writing had
already in the 470s a place of sorts in the theatre; but its exact role in this image
and its status in contemporary dramatic performances are still very much unclear.
We are on safer ground with references to writing within the plays. Already in
Persians, the earliest surviving play (brought on the stage in 472 bc), the messen-
ger may have used an image borrowed from writing to deliver his news: he unrolls
(or opens up) the sufferings of the Persians, as if they had been inscribed on a
papyrus roll or on folded tablets.5 However, the only pointers towards such an
interpretation are the general context (a report listing casualties), and the verb
IÆø (‘open out, unfold’), which is in itself non-specific. The passage in
Persians well illustrates one of the difficulties facing a study of the connotations of
writing in tragedy: the allusive character of references to writing. Because of the
setting of the plays in the heroic age (Persians is actually one of the very few
exceptions), references to writing in tragedy tend to be vague, or confined to
aspects that would not be felt as anachronistic: thus, while certain types of written
documents are tendentially avoided on the tragic scene, letters are admitted,
simply because they fit the atmosphere better.6
Granted this general limitation, there is a further difficulty: what can we expect,
as historians, from a study of the references to (and connotations of) letter writing
in tragedy? A literary study will focus on how the letters function, on how skilfully
they are inserted in the plot, on the way they advance the action—these are
legitimate lines of enquiry arising out of narratological concerns, to which answers
can be given. But what of their meaning? Where does writing fit in the relationship
between drama and the society that produced it, what do the presence and

3
Volute-crater Naples, MN 81673, H3240, ARV2 1336.1, 1704; for detailed discussion of this
extraordinary vase, see the essays in Taplin and Wyles 2010.
4
Green 1995: 80–5.
5
Aesch. Pers. 254: ‹ø I ªŒÅ A IÆ ÆØ  Ł , cf. 294–5 (Atossa orders: A IÆ Æ
 Ł  | º  ŒÆÆ ), and 333 (IÆæłÆ  ºØ), with Hall 1996: 130, Barrett 2002: 49 n. 42,
Garvie 2009: 148. Æø means in general ‘to open’, but here the context (messenger report) may
inflect the meaning. In Aesch. fr. 281a 22 Radt, IÆø is used for ‘opening the pinax’; compare
Soph. El. 639 (FŁ  . . . IÆ ÆØ), as well as ØÆø (to examine) in Eur. Hipp. 935, and the
frequent references in tragedy to ‘folds’ of a tablet, º ı ıå  or ØÆıå ; for prose, Hdt. 1. 125,
IÆ Æ e ıº .
6
Easterling 1985: 3–6. Hall 2006: 47 overstates the case, when she speaks of ‘the remarkably
uninhibited way in which the tragedians import examples of writing and metaphors connected with
it into their heroic world’. Cf. the thoughtful reflections of Ford 2002: 152–7.
 The Athenian Dramatic Stage 185

connotations of writing in drama tell us about that society?7 An intial answer is

that the enhanced presence of certain types of writing on the scene reflects a wider
familiarity with them. Of course there is no need to go through a detailed analysis
of Attic tragedy to realize that by the end of the fifth century literacy was fairly
widespread in the Greek world, and that letters in particular circulated; but
tragedy allows us to see, as in a (distorting) mirror, how literacy ‘filtered through’,
how it entered into, and eventually occupied, both scenes, the dramatic one and
that of the polis.
An answer that better addresses the problem of the distorting mirror might take
as its point of departure the problem of anachronism. If the playwrights intro-
duced references to writing, notwithstanding the fact that these could have been
felt as anachronisms, then writing must have mattered to them and to their
audience—or at any rate, it must have been an important part of the conceptual
landscape.8 Tragedy brings on stage the heroic age, but it is concerned with the
polis, inasmuch as it offers a sustained reflection on the discourse of the city as
expressed in myth, ritual, festivals, and political institutions (including explicit
statements of civic ideals or ideologies). We may thus hope that a scrutiny of the
types of writing referred to in drama, of the specific contexts in which these
writings appear (e.g. as parallel to spoken discourse, or in opposition to it), and of
the connotations that their appearances suggest, will tell us something about the
connotations of writing in contemporary society.
As a general rule, references to writing may appear in tragedy for at least four
reasons. First, writing may be mentioned as one element of the surrounding
conceptual landscape, shared in an unproblematic way by playwright and audi-
ence, although not really having a functional role within the plot. Second, writing
may feature in a tragic play because of the aesthetic effect which results from
bringing a written document of one type or another. Such a prop may render a
given scene more effective, serve to advance the plot, or trigger metatheatrical
associations. These two possibilities make sense on their own, and they are enough

7
There is debate as to the forms the relationship between drama and society took (see e.g. Goldhill
1987; Griffin 1998; Seaford 2000; Goldhill 2000; Rhodes 2003; P. Wilson 2009; Ceccarelli 2010), but the
connection is not in doubt: ‘poetry, like all forms of art, . . . expresses and involves the interweaving,
explicit or implicit, of all the codes of the society, linguistic, ritual, familial, sexual, alimentary, and so
on. This network of interwoven codes determines how we perceive and experience what we call reality.
The poet cannot but utilize and to some degree conform to that pre-coded structuring of experience; to
some degree he also goes beyond it, examines it, reflects upon it’ (Segal 1985: 199). A concise discussion
of the various models, metaphors, and theories pressed into service to describe the relationship between
theatre and society can be found in Hall 2006: 1–6.
8
Wise: 1998: 17 n. 6 gives a ‘raw’ list of references to ‘reading, writing, books, inscriptions, letters
and waxed tablets’ in tragedy: her thirty-nine instances (and the list is not complete: Sophocles’
Poimenes is missing, for instance) are enough to show the importance of writing’s presence. Writing
mattered for the playwrights (and for drama) in a very basic way: it is commonly agreed, notwith-
standing Havelock 1982: 261–313 and 1985, that tragedy was composed in a written form, and that
actors would probably memorize written texts (see Wise 1998: 65; Hall 2006: 44–7, esp. 47 for the
notion that the writing out of actors’ lines may have affected the way in which drama represents this
craft). It is rather natural that a degree of self-reflection on the medium that made its performance
possible would enter the genre of tragedy (Segal 1985: 209, 213–17). But again, this is a feeble
explanation: drama was not an egocentric exercise on the part of the playwrights. The audience
mattered. On the nature of the latter, see Sommerstein 1997; Torrance 2010: 249–50.
186 Letter Writing and the Polis

to explain why there is writing in drama. The first of these possibilities suffices to
justify interest in tracing its workings, beyond the narratological.
But, third, in bringing writing on the scene the playwright may also have had in
mind a reflection on his own discourse, on the status of tragedy as compared to
other types of commemorative monuments (statues, paintings, temples), practices
(rituals), and poetry or speeches, oral and written. Finally, this same reflection may
have addressed issues widely shared: a dramatist in bringing writing on the scene
may have had specific issues in mind that would have interrogated the audience, or
presented issues to which the audience could relate. Truth and falsehood, imitation,
the relationship between physis and nomos, persuasion and communication, that is,
language and its power, the meaning of words, are central to drama; writing, as the
‘double’ of speech, is bound to have been involved too.9
In what follows, references to writing will first be contextualized within the play;
their socio-cultural connotations will be examined only in a second step. The
importance of interpreting as a whole multiple references to various types of writing
within a single play (metaphorical writing, real writing, writing as painting, official
or public writing, books, letters) is the reason why I have preferred a rather
traditional division by authors to a thematic subdivision: the latter was attractive
indeed, and would have been unproblematic for those plays containing only a single
reference to writing, or for the fragments, but it would not have allowed for the
treatment as a whole of, say, the Oresteia, the Trachiniae, or the Hippolytus, to cite
only one example from each section below. I want, however, to stress that no
‘evolutionist’ vision is implied in the articulation.
I have made one exception to my examination by author, namely, for those
plays that brought the letters of the alphabet on the stage qua individual letters:
these plays are discussed as a group, because the preserved fragments seem to
present very strong intertextual links between themselves, because the fragmen-
tary state of the specific passages anyway prevents further contextualization
within the play, and because they are all preserved by one source, Athenaeus,
which at least provides an ancient context of sorts. The discussion also includes
those instances in which words formed on the ªæÆç- stem are used to indicate a
painting, because the fact that the same words were used for both ‘to paint’ and ‘to
write’ tells us something central about the way writing is looked at. The remark
that ªæÆç means both ‘writing’ and ‘painting’ has often been the starting-point
for inquiries into the ‘decorative’ use of writing on Attic vases (for instance), or for
inquiries into the status and interplay at the conceptual level of the various types
of signs on these same vases (images, words, and the further level of signification
evoked by the scene in its entirety).10 But just as writing is an ‘image’, images are a

9
Once a decision had been taken in a public debate, it was usually inscribed on stone, and put in a
public place, ‘for whoever wants to observe’; records of temple administrators and lists of all sorts could
be seen on display, just as dedicatory or funerary epigrams. In all these instances, writing functions as
an independent power, but it is also the residuum of words once pronounced and meant to be repeated.
See e.g. Faraguna 2006: 54–5, and below.
10
See the seminal papers by Lissarrague, 1985 and 1992 (‘Graphein: écrire et dessiner’); Neer 1995
(esp. the analysis of the Hermitage pelike with the first swallow, 132–3); Neer 1998. Both Lissarrague
and Neer see a change around the mid-fifth century: in the archaic period all signs function in the same
way: ‘the painters treat letters and pictures as different aspects of a single phenomenon’ (Neer 1998: 18).
With the development of new techniques (foreshortening, landscape, depth) and in short of a more
 The Athenian Dramatic Stage 187

means of preserving and transmitting information; and references to images

through the specific word ªæÆç may legitimately be put in a series with other
references to ‘writing’.11 Only those images referred to with words of the ªæÆç-
stem will be considered: the focus is on writing, not visualization.12
Against sweeping statements to the effect that in drama writing is very often
negatively marked, feminine, undemocratic, or conversely, democratic, I suggest
that the evidence defies such generalizations. Rather, the ideological connotations
of writing as a cultural practice differ drastically in our surviving scripts, depen-
ding on the context and the type of writing at issue; a monolithic interpretation
is not adequate. Of course, what we have is only a minimal percentage of all that was
performed on the Athenian stage, and this has to be borne in mind when trying to
delineate changes or innovations. However, if we are allowed to take what survives
as a representative sample, the way different types of writing are referred to in
tragedy is in itself revealing; not least, it gives us clues as to how writing, and
specifically letter writing, are positioned with respect to the spoken word.

5.1.1. Aeschylus

Writing is present to a greater or lesser degree in all of Aeschylus’ plays (Persians

being the one potential exception), often as a metaphor for memory, often to refer
to a painting, but also as literal writing.13 We cannot know whether it was listed
among Palamedes’ inventions in the homonymous play; but it certainly occupies a
special place in the Prometheus transmitted among Aeschylus’ plays. Some fluidity
between the series of inventions attributed to each character had been noticed
already by a scholiast to the Prometheus Vinctus, who in discussing vv. 457ff.
(i.e. exactly that part of the list of inventions dealing with the rising and setting of
stars, the science of numbers, and writing) remarked: ‘He [Aeschylus] has also
ascribed the discovery of these things to Palamedes’ ( ø c oæØ ŒÆd
—ƺÆ fi Å æ Bł, schol. Aesch. PV 457 = Aesch fr. 182a Radt).14 Indeed, the
discovery of numbers is mentioned in almost the same words in the Prometheus
vinctus (vv. 459–60: ŒÆd c IæØŁ,  å   çØ ø, | K ÅFæ  ÆP E) and in
fr. 181a 3–4 Radt, from the Palamedes (æHÆ b e   ç  | IæØŁe ÅoæÅŒ ,

mimetic art, inscriptions and images part company: naturalistic painting emphasizes issues of ambi-
guity, while inscriptions, reduced to their linguistic function, tend to offer interpretive stability to the
uncertain pictures.
11
Segal 1985: 223 (‘In some texts of tragedy . . . the notion of painting has associations with the
ambiguous “truth” of writing’) is one of the rare explicit acknowledgements of this; Wright 2005: 318–19
another. Tellingly, in quite a few instances interpreters disagree as to whether the ªæÆçÆ being discussed
are paintings or writings; see below.
12
The issue of visualization has been addressed by Froma Zeitlin in numerous essays; Zeitlin 1994 is
particularly pertinent. See also Steiner 2001: 44–56; Hall 2006: 99–141; Stieber 2011, passim.
13
Metaphor: besides Persians, mentioned above, see Aesch. Suppl. 179; 708–9, and 991–2; Choeph.
450 and 699; Eum. 275; PV 789. Painting: Aesch. Agam. 242, 801, and 1329. Literal writing: Aesch. Sept.
434, 468, 646, 660; the Dike fragment, fr. 281a Radt.
14
More precisions as to the three slightly different versions of this scholion, and their exact location
in the various manuscripts, in Radt’s apparatus; cf. the remarks of Sommerstein 2008, fr. 181a.
188 Letter Writing and the Polis

 å   çØ ø).15 Clearly there is some relationship between the two plays;
and Sommerstein’s suggestion that the list of inventions attributed to Prometheus
derives from a similar list in a Palamedes play is attractive.16 But we do not know for
certain whether Aeschylus attributed the invention of writing to Palamedes; and we
have even less idea of what the purpose of the invention might have been.
The situation is different for the Prometheus Vinctus: writing, characterized as
‘memory of everything, occupation mother of the Muses’, is one of the inventions
highlighted by Prometheus when listing his gifts to humanity (PV 460). The new
techne finds a metaphorical application in the play itself, with the order addressed
by the Titan to Io, to inscribe his words in the tablets of her mind (PV 789). In a
fascinating, if slightly one-sided, analysis of the Prometheus Viuctus, Segal has
described the meeting between Prometheus and Io as ‘the confrontation of oral
and written mentalities’: a Prometheus in control, although fixed to his rock, is
opposed to an Io who enters the scene running, without any sense of direction or
time, exactly as men were before the Titan gave them writing, creatures who saw
without seeing and heard without hearing (442–506).17 According to Segal, the
suggestion of the Titan that Io ‘write down on the remembering tablets of her
mind’ her travels (789) underlines the capacity of writing to give an orientation in
time and space, to visualize in the inner mind what is being said. For Segal, the
metaphor of writing into one’s mind is closely linked to the visual: the move from
the Homeric formula f Kd çæd  ºº B fi Ø to expressions having exactly the
same meaning, but involving writing, signals the move from a memory attached to
hearing and speech to a memory based on visualization.18
This applies, however, only up to a point. When asked by Io to provide her with
details of her future travels, Prometheus obliges, dividing the information in two
long rheseis (707–41 and 790–815). At the beginning of the first rhesis, the Titan
exhorts the daughter of Inachus to ‘put my words in your soul, so as to learn the
end of your wanderings’ ( f K f ºª ı ŁıfiH  º , ‰ i æÆ KŒ Łfi Å
› F, 705–6). After a pause, asked to complete the picture of Io’s wandering, the
Titan accepts, and requires again of Io that she memorizes what he will say. But
this time he also suggests that this memory is to be assimilated to a written one:
‘First, Io, I shall tell you of your much-driven wandering, and may you inscribe it
on the memory-tablets of your mind’ ( º   º Å çæ ø, m Kªªæ ç ı f

15
Similarly, the opening of Prometheus’ rhesis, with its accent on the confused state of mankind
(Aesch. PV 449–50), finds a parallel in fr. 181a Radt (PV 449–50: e ÆŒæe   | çıæ  NŒB
fi  Æ ~
fr 181a Radt:   . . . ZÆ æd çıæ ). Stobaeus, who quoted fr. 181a Radt directly after the
verses 454–9 of Aeschylus’ Prometheus, had already made the connection (see Radt’s apparatus).
Further Sommerstein 2000: 122; Gera 2003: 120–7, and esp. 123.
16
Sommerstein 2000; already Kleingünther 1933: 83 and n. 38, convincingly arguing that the move
will have been from a Palamedes, already known for numerous, but circumscribed inventions, to the
majestic Aeschylean Prometheus, bringer of all arts (AÆØ åÆØ, v. 506) to humanity. See also above,
66–7, 79.
17
Segal (1986): 84–6. For my purposes whether the Prometheus Vinctus is by Aeschylus is not
important; see above, 67 n. 29; Saïd 1985; and West 1990: 51–72 (who denies Aeschylean authorship).
18
Segal 1986: 81. The invention of a mnemonic technique involving visual representation was
attributed to Simonides, to whom was also ascribed the invention of some letters of the alphabet: see
above, 66, and below, Appendix 2 nos. 1 and 6. Recent studies, however, have emphasized that
visualization (in the form of a ‘theatre of memory’) is already at the basis of epic composition: see
e.g. Clay 2011.
 The Athenian Dramatic Stage 189

 Ø º Ø çæH, 788–9). The two expressions function in exactly the
same way, to introduce a long description of places and names. The choice of a
metaphorical writing to represent memorization in the introduction to the second
rhesis may have had to do with the length and complexity of the details the girl is
by now asked to retain, or with the fact that the information takes on an almost
oracular colouring (especially this part, with her arrival to Egypt)—and oracles
were typically written down;19 but there does not seem to be within the play any
qualitative difference between ‘putting the words in one’s soul’ and ‘engraving
them in the tablets of the mind’. However, the second request serves the purpose
of showing one of the Titan’s inventions as it is being put into action; and the very
existence of the play—from the point of view of the audience, its performance—is
a metatextual instance of the other aspect of Prometheus’ gift, the production of
poetry.
In the Oresteia, inscription in the mind and visual image, painting, appear side
by side; notions expressed by ªæ çø and connected forms are used to link
together characters and themes. The thread begins at the opening of the trilogy,
when the chorus remembers Iphigenia and her sacrifice, the girl with her mouth
kept closed by the silencing strength of the bit, and standing out, ‘as in a painting’,
desiring to speak (æ ı Ł ‰ K ªæÆçÆE, æ Ø | Łº ı , Ag. 242–3).
ˆæ ÆÆ speak, of course, although silently and in a mediated way, and can even
scream, as we shall see with the shield scene of the Seven or the tablet of Phaedra.
Iphigenia, standing out as if in a painting, speaks in two ways: although silenced
by the bit in her mouth, she silently speaks through her glances; in the context of
the trilogy, the image speaks through the network of relations built around it.20
This happens next when the chorus, in greeting Agamemnon later in the play,
admits that at the time he was assembling the army for Helen, the king had been
described (or rather, depicted) to him most unfavourably, nor steering the helm of
his mind rightly (Œ æ I  ø qŁÆ ªªæÆ , | P s æÆ ø YÆŒÆ
ø, Ag. 801–2). I  ø (‘inartistically’) is not exactly the term one would
expect: it may have been chosen here to further inflect the meaning of
ªªæÆ  towards a painting; but it is very difficult to choose here between
the two meanings (the uncertainty may actually only be ours: this is one of those
passages that show how identical writing and painting are).21 The following two
verses, if they can be taken to refer to the sacrifice of Iphigenia, may strengthen the
impression of an allusion back to the ‘speaking painting’ (but they might also refer

19
For the oracular colouring cf. Jenkins 2006: 87. To consider the Titan’s rhesis an epistle, as Jenkins
does, is going too far: Io and Prometheus are in a face-to-face situation.
20
For the connection between the ªæÆç as painting here and the silently speaking ªæ ÆÆ so
often mentioned in connection with tablets, see Segal 1985: 223–5, with the further important point
that the graphe here occurs when violence is done to speech. So also Stieber 2011: 148: the comparison
with a painting emphasizes the silence, rather than the beauty, of Iphigenia.
21
Hall 2006: 134–5 discusses interpretations of ªªæÆ  as ‘written’ (‘inscribed upon the
memory’), or as a metadramatic reference to the judges’ writing down their decision on the competi-
tion; she chooses the ‘painting’ interpretation, but her reason—that an interlinear gloss explains the
text with KÇøªæÆçÅ , showing that the ancients understood the verb as referring to a painting—is
not compelling. It is the intertextual relationship which exists between all these references to images in
the Agamemnon that orients the interpretation.
190 Letter Writing and the Polis

to Helen’s forwardness).22 At any rate a painting (ªæÆç), quickly wiped out,

again figures in Cassandra’s final words, pronounced before entering the palace
where she will be killed: sad is human fortune, for a shadow may arise even when
the situation is prosperous, while in misfortune, ‘One wet sponge-sweep wipes all
our trace away’ (Ag. 1329, Morshead’s translation of  ºÆE ªæø ªª 
þº ªæÆç). The three passages are closely interconnected: these humans are
but traces, images fixed in their description; by the end of the play, they will have
been wiped away.
The thread continues through the two other plays of the trilogy. In the famous
recognition scene of Choephoroi, Electra twice uses words compounded from
ªæ çø to state that the outlines (æتæÆç , 207) of the feet and the imprints
( ªæÆçÆ, 209) of the tendons, left by visitors on the tomb, are similar to her
own; Orestes confirms her in her intuition by showing her a tissue she herself had
woven, and pointing out to her the strokes of the wooden blade and the design of
wild animals (ŁæØ  ªæÆç, 232). In the recognition of the siblings, forms of
literal writing thus play an important role. The other two instances of ‘writing’ in
Choephoroi concern metaphorical writing. At v. 450, Electra enjoins Orestes to
write her sufferings in his heart ( ØÆF’ IŒ ø K çæd ªæ ç ı; although the
text is corrupt, this part is certain; the request was probably reiterated, again with
the imperative ªæ ç ı, by the chorus). Later in the play, when given news of
Orestes’ supposed death, Clytaemestra laments the disappearance of the last hope
for the house of Atreus, and orders that this betrayal be inscribed (æ FÆ
ªªæÆç, 699, a metaphorical inscription, as in a list of traitors—an odd expres-
sion, that would have been arresting for the public).
These two types of writing (the image and the metaphorical inscription)
reappear in the last play of the trilogy, placed this time squarely on the side of
death, horror, and punishment. The first one, the painting, is used at the opening
of the Eumenides by the Pythia: in her tentative attempt to describe the scene she
has just seen in the temple, the priestess turns first to comparison with mute,
petrifying statues (ˆæªØ Ø  Ø, v. 49), and then to a painting (r   ’
X Å . . . ªªæÆÆ, ‘I saw once depicted’, 50). The priestess may here be alluding
to a notional painting, such as those decorating Attic vases; perhaps to some
famous painting; or, as has recently been suggested, to the masks of Harpies that
the chorus had worn in Aeschylus’ play Phineus fourteen years earlier, in 472 bc.23
In turn, the Erinyes refer to metaphorical writing to stress the impossibility for
men of avoiding punishment: for Hades ‘supervises all things with his recording
mind’ ( º ªæ çø fi b   Køfi A çæ, 275). From this point onwards words
linked to the root ªæÆç- disappear: Orestes is now in Athens, the images of writing
cease, and the play proceeds towards a public trial and the resolution.

22
The transmitted text is Ł æ  Œ Ø  (willing boldness of Helen). The connection between Ag.
801–2 and Ag. 242–3 is tighter if Ahrens’s suggestion Łæ   KŒ ŁıØH (with reference to the sacrifice
of Iphigenia) at Ag. 803 is accepted, as e.g. by Denniston and Page 1957: 139–40. Palaeographically the
change is minimal; for some arguments against it see West 1990: 202–3. The reference to images of
Helen in Ag. 416–17 is not pertinent here: the language and the overall meaning are somewhat
different.
23
Hall 2006: 116–18, who goes on to stress that ‘allusions to works of visual art may well “mask” a
specific inter-performative reference’.
 The Athenian Dramatic Stage 191

In the Oresteia, then, words from the field of writing function as a way of
interlinking images and themes one with the other. References to a metaphorical
writing serve to convey the image of an enduring memory, typically of griefs; what
is emphasized is the notion of permanency, rather than secrecy. This use is not
negatively marked, although one may wonder whether inflexibility and inability to
forget are necessarily to be viewed as positive values.24 The painting image has
different, rather ominous connotations; it shares with metaphorical writing an
accent on rigidity (the person is fixed in an image, the accent is on the power of the
image to last beyond its human counterpart); but as soon as the writing is thought
of as not metaphorical but real, it becomes possible to wipe it out altogether.25
Painted images also play a significant role in a fragment of Aeschylus’ satyr play
Theoroi, or Isthmiastai. At the beginning of fr. 78a Radt, the satyrs appear proudly
carrying images extraordinarily resemblant of themselves (probably satyr masks,
in a reduplication of their own masks closely resembling a mise en abyme); these
are images closer to life than Daedalus’ models, as they proudly say—only lacking
in voice (Y øº  r ÆØ  F’ KBfi  æçBfi º  | e ˜ÆØ º ı []ÅÆ· çøB E
 , 6–7).26 Their _ stated
_ _ _ purpose is to adorn the_ _temple
_ of _ Poseidon
_ _ with
these votives, defined as Œ  . . . ŒÆººªæÆ , beautifully painted ornament
__
(11–12), a term that while again stressing the aesthetic aspect of the images, points
to their artificial, crafted nature. (May the reference to the lack of voice be taken to
imply that voice is located on the side of the natural?) The aspect most stressed is
the life-like quality of these images: they are so close to reality that the chorus of
satyrs next imagines the horror and confusion of their mother if she were to see
them.27 The imagined horrified response of the mother might reflect a mistaken
assumption on her part that her children have been decapitated; but it may also be
linked to the fact that images elicit particularly strong responses, sometimes

24
For Segal 1986: 81, the association of writing and emotional interiority found in the image of the
tablets of the mind is aided by ‘the tendency in an oral culture to connect writing with private, secret or
deceitful communication (particularly of an erotic nature) and the importance that writing gives to
vision, for the Greek the most powerful stimulus of eros’. This is an important perspective, but needs
qualifications, since the part about eros is true only in some instances, and the same applies to
deceitfulness or secrecy.
25
Agamemnon might seem to have changed, in the appreciation of the chorus (Aesch. Ag. 799–
809); but his words show that he does not understand (note the emphasis on Helen in his words,
ªıÆØŒe oŒÆ, 823, just after the chorus has sung of how destructive Helen has been); and although
he himself takes up the image of mirrors and doubles (›غÆ Œ  æ , Y øº  ŒØA, 839), he
misunderstands the chorus’s warnings and continues to steer his course unwisely—e.g. stepping on the
purple carpets in the next scene. When discussing graphein as painted image, other imagery-related
terms should be taken into account; but this would exceed the boundaries of this study. Segal 1986
offers a good entry; see also Zeitlin 1994; Stieber 2011 (mainly focused on Euripides).
26
Discussion of the play in Wessels and Krumeich 1999; O’Sullivan 2000; Voelke 2001: 283–9.
Wessels and Krumeich 1999, 142–4 propose that the images carried by the satyrs might be wooden or
ceramic pinakes with portraits, rather than masks; but the argument (that for a mask the painting
would not be stressed, 143 n. 54) is not really convincing; O’Sullivan 2000 and Voelke 2001: 286–7
present persuasive arguments for masks.
27
Zeitlin 1994: 138–9; Steiner 2001: 45–9, who stresses that the accent on the images’ extraordinary
techne, together with the mention of Daedalus (Daedalus’ sculptures were supposed to be living,
moving, and speaking ones), is meant to underline not so much their naturalistic quality as the fact
that they, at least in the perception of the satyrs, ‘straddle the boundaries between the manufactured
and the living’; O’Sullivan 2000.
192 Letter Writing and the Polis

actually stronger than the ‘reality’ they are depicting.28 Next, we see the satyrs
nailing the images to the temple, exactly as votive dedications and bronze tablets
with public decisions and decrees were nailed on (or directly inscribed upon) the
walls of Greek temples:
ŒIØÆ ºı’ ŒÆ  B Œ[Æ]ºB  æçB. [- | ¼ªªº , ŒæıŒ’ [¼]Æı , Kæø
_ _ª KØåØ ŒºŁ ı  f  [ı] ç [ -
Œøº æ[Æ, | ‹ [] _
_ _ _ _ _ (Aesch. Isthm. fr. 78a, 19–21).

Each of you nail up there a messenger [ . . . ] of his handsome self, a voiceless herald, a
restrainer of travellers, which will make visitors halt in their path [by the] fear[some
look in their eye].29
How exactly the paintings were supposed to function in the play is unclear; but
the image of a voiceless herald is strikingly appropriate for a public display of
writing (or as here, of painted, speaking although voiceless, images—even more so if
they are indeed masks).30 The effects and power of visual artifices are here presented
through the mirror of (and enhanced by) the satyrs’ curiosity and inexperience.
Literal writing has an important place in the shield scene of the Seven (375–
767); among the various codes through which, in the report of the herald, the
message of the aggressors is expressed (gestural, sonorous, verbal, iconic, and
graphic, variously combined), writing appears three times, to mark the messages
of Capaneus, Eteoclus, and Polynices.31 The ªæ ÆÆ are thus only one of the
expressive codes mobilized, a specifically human one (writing has no part on those
shields whose imagery is cosmic or monstrous, and it accompanies gods only
when these interfere in the affairs of mortals, as is the case of Dike).32 The letters
function in the same way as the images, transposing the living aggressivity of the
warriors (gestural, vocal, and sonorous) into a permanent graphic sign; but they
also serve an ecphrastic function, in making clear the intentions of the figures
depicted on the shields (in their turn coinciding with the intentions of the bearers
of the shields). While it is important to keep in mind the overall scene, and while
the various communicative codes bring out their full meaning only in their
interaction, an analysis of the entire scene would go beyond what is required
here. We shall thus concentrate only on how the grammata function in the three
cases in which the shield bears an inscription.

28
For the various interpretive possibilities, see O’Sullivan 2000: 361–3. The interpretation suggested
above takes its cue from Zeitlin 1994 on ‘hyperviewing’ and the power of images (the mother here
would be cast in the role of the ‘spectator’); see below on Hecuba in the Hecuba.
29
The Loeb text of Sommerstein 2008 (fr. 78c = 78a + 78c); his translation, modified.
30
The overlap of visual and aural here is noted by O’Sullivan 2000: 363–4, who compares with
Aesch. Sept. 518 (æe ºª   F Æ , ‘in accordance with the symbol’, referring to the image on
the shield of Hyperbios); in both Aesch. Sept. 82 and Suppl. 180, the dust raised by the enemy is
described as ¼Æı  ¼ªªº , ‘voiceless messenger’. An intriguing comparison is offered by a small
inscribed bronze disk from Olympia dated 500–475 bc (Minon 2007, no. 16, SEG 51, 532), with holes,
presumably to put the disk up against a wall: the Eleans give the citizenship to two individuals and their
descendants, who among other things will also receive the thearoi (sacred ambassadors); the text closes
with ‘the tablet is a dedication to Zeus’, › b Æ ¼ªÆºÆ  Ð ˜Ø.
31
The bibliography is immense; the fundamental essay is Zeitlin 1982; see also the shorter, excellent
discussion of the interaction between the various communicative codes in Magini 1998: 25–35.
32
Magini 1998: 28.
 The Athenian Dramatic Stage 193

In the case of Capaneus, the messenger reports his words (‘he says that he will
sack the city (KŒæØ ºØ | . . . çÅ), whether the god is willing or not . . . he
compares Zeus’ lightning and thunderbolts to the heat of midday’, 427–31),
describes the device on the shield (a naked man carrying a torch), and adds that
‘in golden letters he declares: “I will burn the city” ’ (åæı E b çøE ªæ ÆØ
‘æø ºØ’, 434). From the way in which this is presented, it is unclear—
purposely so—whether it is the image, or Capaneus himself, who speaks through
golden letters: the various codes are here superposed; at the same time, the
relations established between them emphasize their differences.33 The next war-
rior to use letters to enhance the message of his shield is Eteoclus. Eteoclus is
silent, but circles with his horses, who emit hissing, barbarian sounds; the image
on his shield, unlike that of Capaneus, is a true mirror of himself, an armed hoplite
climbing on a ladder over the top of the wall of a city which he wants to sack
(KŒæÆØ Łºø); ‘he too shouts in syllables of written letters that not even Ares
could throw him off the battlements’ ( AØ b å s  ªæÆ ø K ıººÆÆE | ‰
P ’ i @æÅ ç’ KŒ º Ø ıæªø ø, 468–9). Here, the letters do not replicate
an oral message, as in Capaneus’ case; the image on the shield—and thus Eteoclus
himself—speaks through the letters only.34 As for the seventh warrior, Polynices,
he makes use of all the communicative codes mentioned, body language, sound,
words, icon, and writing. The messenger first reports Polynices’ words (that he
may climb the walls, be proclaimed king, and sing the victory paean, and either kill
or send Eteocles into exile, Sept. 634–8). He then describes his shield, on which a
double device ( غ F BÆ) of a woman leading a warrior is cunningly engraved
(æ ÅåÆÅ ); the woman affirms she is Justice, ‘Dike, as the letters say:
“I will bring back this man from exile, and he shall possess his father’s city and
the right to dwell in his home” ’ (˜ŒÅ ’ ¼æ’ r Æ çÅØ, ‰ a ªæ ÆÆ |
ºªØ· ‘ŒÆ ø ’ ¼ æÆ  , ŒÆd ºØ |  Ø ÆæfiÆ ø ø ’ KØæ ç ’,
646–8).35 Here, the identification between the second image (the warrior) and
Polynices is total; just as total seems to be the identification between the image of
the modest woman and Dike, the goddess daughter of Zeus, the one goddess who
does make a limited use of writing, because she has to deal with mortals. And yet,
the repetition of verbs of saying insinuates doubt, as does the fact that this is the
first time that the letters, rather than the image/referent, speak; it is as if the
message (its vocalization) was detached from the image (in fact, unless we assume
that there is another identifying inscription stating that this is Dike, what the
woman says in the report of the messenger and what the letters say are two

33
Capaneus is the subject of the first verb of saying, at l. 428, and his words are reported in indirect
speech by the messenger; he is the subject of åØ at 432 (åØ b BÆ ªıe ¼ æÆ ıæçæ ); but
both Capaneus and the naked man may be the subjects of vv. 433–4: çºªØ b ºÆa Øa åæH
‰ºØÅ, | åæı E b çøE ªæ ÆØ ‘æø ºØ.’ The parallelism of thunderbolts and torch,
çÅ and çøE ªæ ÆØ, enhances the contrast between Capaneus, a hoplite, and the naked man
depicted on the shield. See Zeitlin 1982: 63–5; on letters as ‘double’ of the oral message see Magini 1998:
23–4, and on this passage, 29.
34
Magini 1998: 30. For the cultural progression, see Zeitlin 1982: 74–5; for the name Eteoclus, itself
a message, Zeitlin 1982: 77–8 (Eteoclus is a morphological variant of Eteocles; the anti-Theban hero
Eteoclus replaces here the epic Adrastus, so as to create a thematic variant of the Theban Eteocles).
35
On this last gate, see Zeitlin 1982: 135–45 (in part. 138, on غ F, and on the fusion of man,
iconic sign, and verbal sign; and 139 on the relation between Eteocles and Polynices).
194 Letter Writing and the Polis

different things). Eventually, the words of Dike are shown to be a projection of

Polynices: Eteocles reads them as babbling letters (‘we shall soon know whether
these letters worked in gold on the shield, bubbling with his mind’s wanderings,
will bring him home’, Y Ø ŒÆ Ø åæııŒÆ ªæ ÆÆ | K’ I  çº Æ
f ç ø fi çæH, 660–1).
What are then the connotations of writing here? In the first case, that of
Capaneus, writing doubles the oral message; in the case of Eteoclus, it substitutes
for it; in Polynices’ case, it adds a further aspect. Throughout, the connotations of
the messages transmitted are negative—but the negativity would seem to touch
the message rather than the means. Writing is here used exactly as the voice is
used; the various communicative codes interact with an extraordinary degree of
sophistication. The entire scene can be taken as the result of a reflection on the use
of all available communicative codes, in synaesthetic combination and separated,
to produce the most vivid description possible.
Supplices presents four explicit references to writing, the highest number in any
Aeschylean play. This number impresses, especially considering that other pas-
sages may imply writing too; one of the explicit references to writing, it has been
claimed, concerns letters. A closer look shows that two of the four mentions of
writing concern metaphorical writing: early in the play, the Danaids are asked by
their father to preserve his words as if putting them on tablets (179: ÆNH çıº ÆØ
¼’ Å º ıÆ); later, once the decision of the Argives has been announced,
Danaus tells his daughters to add his present counsels to those they have already
written down, in a sentence that clearly harks back to his first request (ŒÆd ÆFŁ –
Kªªæ łÆŁ æe ªªæÆ Ø |  ºº EØ ¼ºº Ø øçæ ÆØ Ææ, 991–2).
At 708–9, in praying that the Argives may always worship the gods who possess
their land, the chorus (i.e. the Danaids) mentions a divine, and probably again only
metaphorical, writing: respect of the parent is inscribed in Dike’s thesmia.36 These
references all concern the Danaids’ relationship to writing: they have an idea of what
this skill is, but they practise it only metaphorically.
The remaining mention of writing is very different: it is the statement made by
the king of Argos, Pelasgus, that a vote passed by the people of the city is stronger
than a tablet:
 Ø  ÅæÆŒ  KŒ ºø Æ | łBç  ŒŒæÆÆØ,   KŒ FÆØ Æ
fi | º 
ªıÆØŒH. H KçºøÆØ  æH | ªç  ØÆ , ‰ Ø IæÆæø. | ÆF P
Æ  KØ KªªªæÆÆ | P K ıåÆE ºø ŒÆçæƪØÆ, | ÆçB
IŒ Ø K KºıŁæ  ı | ªºÅ. (Suppl. 942–9)
That is the unanimous vote passed by the people of the city, never to surrender under
compulsion this band of women. This decision has been nailed down with a nail that
has pierced right through, so that it stays fixed. Not on tablets are these decisions
inscribed, nor in folded papyrus sheets have they been sealed: you hear them plainly
from free-speaking lips.
This statement Pelasgus makes in answer to the request of the Egyptian herald
to know the name of the person who stops him from bringing back the fifty
women. Pelasgus refuses to give his name; he actually begins his answer with this

36
On the meaning (literal or metaphorical) of ªªæÆÆØ here see Friis-Johansen and Whittle 1980:
iii, 69–70. Note also the writing by Dike in the Dike fragment, fr. 281a Radt, discussed below, 197–8.
 The Athenian Dramatic Stage 195

refusal (  Ø ºªØ åæc  h  ; 938), which underlines the fact that the decision
is not a personal one, and then goes on to relate the decision of the people.37 The
final sentence of Pelasgus’ answer, where his own free words are opposed to tablets
or sealed papyrus, implies that for the Argive king the oral promise has a stronger
value than anything written on tablets, or sealed in the folds of papyrus sheets;38
yet, since at issue here is the stability and permanency of the Argives’ decision, as
shown by the image of the nail in the wall, writing would prima facie seem to fit
the bill better. Faced with this difficulty, interpreters have argued that an inscribed
decree might have been modified, or might have been unclear; that an oral
messenger is generally felt to be more reliable; or that this passage should be
understood in light of the (later) Platonic critique of writing, as expressed in
the Phaedrus.39 As Grethlein has convincingly shown, the first two solutions are
desperate remedies, and the third approach does not work here: what the passage
highlights is the immutability of the spoken word as opposed to written state-
ments, something very different from what Plato had in mind. Moreover, the
metaphor of writing for secure remembrance is very present in Greek tragedy, and
particularly in this play: we thus need to explain the apparent sudden shift in the
perception of writing.
One solution to all these problems is to see in this passage a contrast between
oral communication, marked as free and democratic, and written communication,
marked as despotic. ‘The oral message is more befitting than a letter for the
representative of free citizens, presumably because there is no suggestion of
secrecy about it’; the democratic, open way of conducting affairs is here opposed
to the oriental, despotic way, a thesis that finds support in what precedes and
follows this passage.40 More specifically, oral communication as the most demo-
cratic form would be opposed here to the Greek letter (Æ ) and to the Egyptian
writing (º ). But Æ may be used for other types of writing besides letters
(including decrees); and º , a (here sealed) papyrus sheet or roll, cannot be
interpreted in fifth-century Athens as typical of Egyptian writing.41 An alternative
solution, the one proposed recently by Grethlein, lies in understanding these two
forms of written documents not as opposed to the oral communication, but rather

37
For the use of a specific type of political language in this play, see Easterling 1985: 3–4.
38
So e.g. Segal 1986: 88; Musti 1997: 49–53; a very different interpretation in Grethlein 2001. The
interpretation of Pelasgus’ words here is independent from the larger issue of his role in the play
(Sommerstein 1997 makes of him a manipulator of the assembly—words could be used to bad purposes
too).
39
See Grethlein 2001: 176–8 for references and a good discussion; I fully agree with Grethlein’s
point, that a critique of writing in the Platonic sense cannot be what is meant here.
40
Friis-Johansen and Whittle 1980: iii, 250–2; cf. Easterling 1985: 5. While ultimately unconvinced
by this interpretation, Grethlein himself (2001: 178–9) brings forward arguments for it.
41
For the interpretation that sees in Æ a Greek format (and specifically a letter) and in byblos the
oriental format, see e.g. Friis-Johansen and Whittle 1980: iii, 250–2; cf. Easterling 1985: 5. Turner 1975:
10 interprets everything as directly related to the workings of Athenian democracy: the nail would be
the one with which decrees are publicly nailed; the Æ would correspond to the minutes of the
secretary; the folded and sealed º  would be the sheet of papyrus kept in the archive (he compares
the ‘document of the decree’, mentioned in Rhodes and Osborne 2003, no. 2, ll. 61–2 (e b غ  |
[ F łÅçÆ  ÆæÆ Ð ÆØ ÆP]HØ ªæÆÆÆ B  ıºB, 403 bc). Contra, Grethlein (2001: 178–9).
196 Letter Writing and the Polis

as (positive) instances of permanency, mentioned in order to further stress the

permanency of Pelasgus’ words.42
But attractive as such an interpretation may be, it does not really account for the
flow of the passage. It seems difficult to deny that there is here a contrast between
ÆŒ and º Ø on the one hand, and the free voice of the representative of the
Argives, a contrast that implies an opposition between a despotic and a demo-
cratic way of conducting affairs. The way out of the interpretive difficulty may be
to look at the play first, rather than at the Athenian context, in order to under-
stand what is meant by ÆŒ and º Ø here. For both terms appear earlier in
the play. At vv. 438ff., after the pressing supplications of the Danaids, when
Pelasgus realizes the inevitability of what will happen, he expresses his realization
through the image of a rivet or a nail fixing solidly a ship (A  I ªŒÅ, ŒÆd
ªªçøÆØ . . . , 440: compare ªç  at 945).43 The Danaids seize the occasion
to make their strongest and final appeal (æÆ’, 455): they first show him their
belts (or breast-bands, æ ç ) and girdles, which they qualify as ‘gatherers of
their robes’ (it is worth noting that the Greek here has ‘syllables’, ıººÆa
ºø, 457); Pelasgus answers with mild irony, noting that these are appropriate
items for a woman (ªıÆØŒH ÆFÆ ıæB º Ø, 458). The girls then proceed
to inform the king that if their appeal is not heard, with the help of these they will
‘adorn these images with tablets of a strange sort’ ( Ø Æ Ø æÆ Œ BÆØ
 , 463). When Pelasgus does not understand, they explain that they plan to
hang themselves from the statues of the gods: they will themselves be ƌ,
messages of a new type, with the help of their ‘syllables’.44 Significantly, the entire
conversation between Pelasgus and the Danaids (from v. 234, his entrance on the
stage) is meant to foreground their diversity: the king remarks throughout on the
girls’ alterity, while they do not understand Pelasgus’ desire to defer the decision to
the people, nor do they comprehend the workings of the Argive system.
I suggest then that if Pelasgus stresses in his answer to the herald that his words
are firm and not written on ƌ, he is going back in his mind to the first time
he has used the image of a nail, but at the same time he is keeping his distance
from the novel kind of writing represented by the Danaids. As for the sealed
º , while it is true that in a fifth-century Athenian context it cannot be taken
to mean Egyptian writing, yet in the context of this specific play it may very well
have had this specific meaning. Only a few verses earlier Danaus, talking to his
daughters, had opposed Argives and Egyptians in affirming: ‘Yet rumour has it
that wolves are stronger than dogs; the papyrus-fruit does not conquer ears of
corn’ (Iºº Ø çÅ  f ºŒ ı Œæ ı ŒıH | r ÆØ: º ı b ŒÆæe P

42
Grethlein 2001: 181, who proposes the following translation: ‘Die ist <zwar> nicht auf Holztafeln
geschrieben und auch nicht in Papyrusrollen “eingesiegelt”, aber—<was viel mehr ist>—deutlich hörst
du es von einer freisprechenden Zunge.’
43
The passage is problematic (see Friis-Johansen and Whittle 1980: ii, 345–7), but the ‘nailing’
image is certain, and appropriate.
44
Suppl. 464–5. Friis-Johansen and Whittle 1980: ii, 110 remark: ‘The tablets will, as it were, contain
a very unusual inscription, to the effect that because of Pelasgus’ attitude the gods have not been able to
help their suppliants.’ Cf. also vv. 282–3, where even if the Danaids are being compared with real
Cypriot women, the overall imagery is that of artistic representation (as seen by Kranz 1933: 74), here
specifically sculpture. Mitchell 2006 offers a fascinating discussion of the way in which the Danaids are
both kin (Greek) and foreigners.
 The Athenian Dramatic Stage 197

ŒæÆE  åı, 760–1). Wolves and corn here represent the Argives, while the
dogs and the papyrus stand for the Egyptians.45 In that same passage, Danaus had
been careful to locate himself on the side of orality: in concluding his dialogue
with his daughters, he had stated that ‘The city will find no fault with a messenger
old in years, but young in his well-speaking mind’ (¼ªªº  P łÆØ | ºØ
ªæ Ł , HÆ Pªºø fi çæ, 774–5). Thus, when Pelasgus affirms that his
answer is not written on a sealed papyrus sheet, he is keeping his distance from the
male counterpart of the Danaids, from another type of writing: personal, private,
sealed. As a result, I should not look for any specific reference to letters in the
ÆŒ or the sealed º  of Suppl. 946–7. Of course, the ÆÆ ºªæÆ
(‘baneful signs’) inscribed on Bellerophontes’ folded Æ (Hom. Il. 6. 169) will
have come to the mind of some spectators; the reference may have been mobilized
already in the earlier passage where ÆŒ and death are linked—after all,
Bellerophontes’ letter, even though written by a man, was instigated by a
woman, and was meant to cause death; but foremost here, in Pelasgus’ words, is
a contrast between the free mouth of the representative of the Greek demos, the
silent and deadly feminine writing proposed by the Danaids, and the sealed and
just as deadly writing of the violent Egyptians. Even as the king defends the
Danaids, he lumps them together with their cousins, in a gesture that may have
pointed to the events narrated in the rest of the trilogy. Such an interpretation
makes of Supplices a play with no references to ‘real’, or at any rate Greek, writing:
writing is here either metaphorical, and on the whole positive, or external to the
Greek world, and negative, because located on the side of the feminine, the
personal, the oriental, the secret.46
And yet interestingly Dike is mentioned more than once in connection with
very literal writing. As we have seen, the chorus of the Suppliants states that
respect for one’s parents is written in Dike’s thesmia (Aesch. Suppl. 708–9); this
may be literal, or metaphorical writing. Dike appears also in the shield scene of the
Seven against Thebes, as one of the emblems which speak through writing; of
course, this is a problematic Justice, as the events will show, and the غ F,
‘double’, that characterizes the BÆ there implies certainly more than the simple
fact that two figures form it. Finally, a fragment of an unknown play of Aeschylus
(F 281a Radt, possibly from the Aetnaean Women) describes how Dike accom-
plishes her divine role of rewarding the just and chastening the reckless through
writing the mortals’ offences on the tablet of Zeus: at the right moment, the list of
crimes will be unrolled (ÆŒ’ IÆØ[], v. 22). In the dialogue between the
chorus and Dike, writing on a tablet (and _this seems indeed a very literal writing)
is presented as stronger than either magic or force (Cho. K]ø fi ÆE  j ŒÆ’ Nå 
æ []; Di. ªæ ç ıÆ] I<>ºÆŒÆ’ K ºø _
fi ˜Ø[., vv. 20–21). This is the
_ __ _ _ _ _

45
A legend associated Danaus with the wolf (Paus. 2. 19. 3–5) and with Apollo ¸ŒØ , invoked by
the chorus of Danaids at v. 686; the wolf appears on fifth century Argive coins, and is perceived a
symbolic of the city (schol. Soph. El. 6: fiH  ÆØ H æªø KªåÆæ ŁÆØ <e> ºŒ  ‰ ŒÆd
a ªºÆFŒÆ ŁÆÇ). Cf. further Friis-Johansen and Whittle 1980: iii, 49–50 and 108–10; Sommerstein
2008, ad loc.
46
Aeschylus would certainly not have let himself be constrained by mythical chronologies; yet if we
think that Danaus was credited with bringing writing to Greece, it makes sense that only the Danaids
and the Egyptians are credited with (deviant) forms of writing, while the Greeks speak.
198 Letter Writing and the Polis

first appearance in Greek literature of divine tablets of destinies, a notion well-

known in the Near East, where it plays an important role.47 But the image,
whatever its origins, is employed in fifth-century Attic drama in a very specific
way: it mediates between the human and the divine world, in a context that
concerns the administration of divine justice.48 Why the notion of administering
(divine) justice should have attracted writing so forcefully is disputed: besides the
possibility of an Oriental influence, whose channels are not so easy to trace, the
actual writing of the laws as a display at the centre of the polis appears an
interesting possibility. At any rate, the idea of a heavenly register, kept by Dike,
Zeus, or also the stars, will have some future in Greek literature: but it will never
become central.49 What is indubitable is that whenever the gods write (and this
happens a few times in drama), it is because of men.
A typology of the forms of writing (and of written documents) present in
Aeschylean drama thus includes writing as a metaphor for firm recollection,
often implying lists (so Dike’ thesmia in the Suppliants, so also Hades’ ‘delto-
graphic mind’ surveying men’s actions in Eumenides, or the ‘memory-tablets’ of
Io’s mind registering her future wanderings, in Prometheus Bound ); ‘writing’ as a
painted image, rigid and fixed, yet surprisingly labile, and often feminine; writing
on shields, that is, writing on a private and personal support, but for public
display; and writing of non-identified documents on normal, everyday supports,
such as pinakes and papyrus, but a writing attributed to ‘others’. In all this, there is
evident a reflection on writing and on the possibilities it involves. The lists are

47
For instance in the Gilgamesh, Bēlet-seri, the scribe of the netherworld, reads from tablets to Queen
Ereshkigal (Gilgamesh VII 194–5); the prophet Malachi mentions a book of remembrance, written in
front of the Lord, of those who revered him (Malachi 3:16); further references in West 1997: 562 n. 36. In
the Near East tablets held by gods extend beyond the sphere of retribution, to include planning and
organization of the world in general: so, for instance, Nisaba consults a tablet when organizing the
building of the temple of Gudea (ETCSL 2.1.7): writing underwrites human and divine society.
48
On the ‘Tablets of Zeus’, see Solmsen 1944; on the Dike fragment, Wessels 1999, who gives a
competent status quaestionis, and considers that the fragment comes from a satyr play. Patrito 2001: 83
suggests some external parallels to Dike’s action of writing on a tablet: the Athenian organization of the
Dionysia, in which the judges would write their decision on a ªæÆÆE  (compare the vase with
Hermes and a tablet discussed above); the political project of Hippodamus of Miletus (Arist. Pol. 2. 5.
8); even more interesting, if this fragment was indeed part of Aeschylus’ Women of Aetna, the tradition
according to which in Sicily oaths were inscribed on a tablet, then immersed in the water, to test the
truthfulness of the taker of the oath ([Arist.] Mir. ausc. 57, 834b; Polem. fr. 83 Müller; Steph. Byz., s.v.
Palike). Quite a few texts connect writing, water, and women: besides Aesch. Ag. 1329 (discussed
above) and Soph. fr. 811 Radt (below), note also the comic references to oaths written in water or wine,
below, n. 57.
49
Patrito 2001: 82 n. 17 and Magini 1997/98: 242 n. 84 list references. For the future of the idea, cf.
Plaut. Rudens 13–15 (the stars report in writing to Zeus the names of those who swear false testimony
or false oath) and 21 (Zeus keeps written on other tablets the names of the good men, bonos in aliis
tabulis exscriptos habet, an idea derived from Diphilus), with Solmsen 1944; also pertinent, even if
tablets are not explicitly mentioned, is Callimachus Hymn 6 to Demeter, 56 (Nemesis ‘put down in
writing his evil speech’, ŒÆŒa Kªæ łÆ çø ), a passage taken up by Nonnus Dion. 1. 481; see also
Babrius 127 (= Perry 1952, no. 313: Zeus orders that Hermes write on ostraka men’s wicked deeds);
Babrius 75 (= Perry 1952, no. 317: Kore and Pluto threaten to write down the names of those doctors
who do not let people die); Luc. De merc. cond. 12 KŒ H ˜Øe ºø ›  æı (in an ironical context);
schol. B Hom. Il. 1. 175; and Hesych.  1190 Œı ºÆØ· ÆŒÅæÆØ. ŒÆd ƃ ƒØŒÆd r ºÆØ. ŒÆd Zç[æ]ø
r . çæƪººØÆ, ºHæ Ø. ÆŒ, Kç’ x   ˜ŒÅ ªæ çØ a H IŁæø ±ÆæÆÆ. j ŁºÆŒ
æ Ø Ø. Behind Hesychius is most likely the Dike fragment (Aesch. fr. 281a Radt).
 The Athenian Dramatic Stage 199

particularly interesting, as these are also, as we have seen, one of the formats
chosen for defixiones; clearly, here the influence of official lists is felt. The absence
of writing from Persians, in a play that contains much that is catalogic, might also
be significant: if there is a play in which oriental, tyrannical writing could have
been signposted, that is Persians (one has only to compare with the picture offered
in Herodotus’ Histories). Persians, produced in 472 bc, is our earliest preserved
play: explicit allusions to writing may have been purposely avoided, or better, may
have been felt not to be pertinent. The messenger’s report in Persians is particu-
larly close to the voice of the epic narrator;50 avoiding writing makes the narrative
of the defeat even more something sanctioned by the gods, rather than a ‘text’,
with all of the human connotations this implies.
None of the plays refers explicitly, much less brings onto the stage, any concrete
‘Greek’ document; this applies also to letters. ¯Ø ºÆ (in the plural), that is
what will become the specific term for letters, appears four times in Aeschylus, in
contexts where the meaning is clearly that of ‘counsels, injunctions’ given in a
face-to-face encounter by one person to another person or to a group, with no
implication of writing.51 It would be interesting to know more about the context
of the ‘old writing’ mentioned in fr. 331 inc. fab. Radt (‰ ºªØ ªæ  ªæ Æ); as
it is, we can only see in this fragment one more proof of the interest borne by
Aeschylus to writing. Finally, while there is much that is unclear about fr. 358 inc.
fab. Radt, ‘Face (?) him at a distance, because at close range you won’t [hit?] him;
be like a man who reads clearly in his old age’ ({ P b{ ¼ø[Ł] ÆP· P ªaæ
KªªŁ | [x –]· ªæø b ªæÆÆf ª F Æç),52 the allusion to changes in
vision depending on age, put forward here (in an extraordinarily anachronistic
way!) in connection with the office of grammateus, betrays a remarkable familiar-
ity, on the part of the dramatist but also of the public, with writing. If the context
was indeed a combat, then the Homeric use of Kتæ çø to mean ‘to graze, to
inflict a light wound’, typically with an arrow from afar, may have been the subtext
of this allusion to writing.

5.1.2. Sophocles

In Sophocles, as in Aeschylus, KØ ºÆ was probably used only for oral mes-
sages. Out of the four instances of the term, two are unproblematic: the contexts
show that the KØ ºÆ referred to in the Ajax (748) and in the Oedipus at
Colonus (1601) are indeed oral requests. The lack of context of fr. 128 Radt, from

50
Barrett 2002: 40–6.
51
Probably all (certainly three out of four) are instances of ‘instructions’ given by a father to his son
or daughter: by an unknown character to another, in Aesch fr. 293 Radt (
— ¼Œ ı a Ka
KØ º , possibly words of Helios to Phaethon); by Darius to Xerxes, who does not remember
them, in Pers. 783; by Zeus to Hephaistus, in P.V. 3–4; by Danaus to his daughters, in Suppl. 1012. In
this last example, it might be just remotely possible to ‘hear’ writing behind the term: the exhortation to
follow the paternal counsel (çºÆ ÆØ   ’ KØ ºa Ææ) closes a speech that Danaus had opened
with the suggestion that the girls write down his counsels (991–2, and cf. the use of çºÆ ÆØ for Danaus’
suggestions in 179). But the situation is that of a face-to-face encounter.
52
I prefer here to Radt’s text the cruces and unrestored lacunae of Sommerstein 2008, whose
translation I follow.
200 Letter Writing and the Polis

the Andromeda, ‘do not be afraid of new orders/letters’ (Å b ç EŁ

æ ç  ı KØ º ), renders discussion impossible.53 The fourth, and most
interesting, instance occurs in Trachiniae: Deianira, after having learnt the truth
about Iole, suggests that they all enter the house so that Lichas ‘may take away the
message of words’ (‰ ºªø  KØ º  çæfi Å, 493) as well as gifts in exchange
of gifts. It could not be clearer that the KØ ºÆ are here an oral message—and
yet: why the precision, why ‘of words’? The iunctura sounds odd.54 While a formal
explanation for it can be found in the fact that ºªø  KØ º  here parallels
the æø HæÆ of the following verse, Deianira’s words also find their sense in the
overall context of the play which, as we shall see, gives a large place to written
transmission of information. Finally, another tantalizing fragment must be men-
tioned in this context: fr. 784 inc. fab. Radt mentions a ‘writing of the herald’
(ªæ Æ ŒÅæŒØ ), whatever that may mean exactly. Once again, the lack of
context renders any definite interpretation impossible; but whether a symbol or a
long piece of writing (a message) was meant, the passage clearly had to do with the
transmission of official information in a context in which some form of writing
figured.55
At any rate: letters stricto sensu do not seem to appear in what remains of
Sophocles.56 Writing, however, figures, and writing of a type rather different from
what we have seen in Aeschylus. Granted, some instances of metaphorical writing
are also present in Sophocles. Thus, at the end of the Philoctetes (produced in 409
bc), Neoptolemus, before making his final plea to Philoctetes, exhorts the hero to
write his words in his mind (ŒÆd ÆF’ Kø, ŒÆd ªæ ç ı çæH ø, 1325); he
will not succeed, and the authoritative words of Heracles will instead make their
traditional, unwritten way into Philoctetes’ mind. Writing in the tablets of the
mind as a metaphor for recollection also appears in fr. 597 Radt, from the
Triptolemos (one of the earliest plays by Sophocles, to be dated in the 460s).

53
Lloyd-Jones in the Loeb edition translates ‘letters’—but ‘orders’ would be equally possible. Segal
1986: 95 is probably thinking of this fragment when he mentions as instances of letters by women who
can bring salvation ‘that of Andromeda in Euripides’ lost play of that name’ (but I cannot understand
where the notion of salvation through letters fits in any of the known Andromeda plays).
54
Davies 1991: 135 simply comments: ‘a defining genitive’, adding that ºª ı çæØ is tragic idiom
for ‘bring a message’. Rosenmeyer 2001: 62 suggests as the reason the fact that Deianira will send as a
present a robe in a casket, sealed with the stamp of her ring (614–15): ‘the robe sealed in its casket
functions almost as a letter would: it attempts to bridge the distance between the separated lovers,
defends its authenticity by way of a signet ring, and metaphorically transfers Deianira’s love with it to
its recipient, Heracles.’ One could add that, just as some letters, it kills. See already Segal 1986 for an
interpretation along these lines.
55
The fragment is preserved in Etym. Gen. A (= Et. Magnum 511.53) = Et. Gud. 320.23 Sturz: Ø
b ŒÆd ŒÅæŒØ  ‹æ ŒÅØŒ KØ ŒÆd ÅÆØ e  F ŒæıŒ , ‰ Ææa  ç ŒºE ªæ Æ ŒÅæŒØ .
Radt mentions Wagner’s suggestion that the gloss ŒÅæŒØ  ªæÆç, transmitted without further
interpretation in Suda Π1547 Adler, may refer to this passage. It might have been a text, or an
image announcing further developments.
56
A letter may have played a role in the Iobates: but only three fragments of the play are preserved,
and the plot is very sketchily known (it revolved around the story of Bellerophontes). A number of
Italiote vases represent the arrival of Bellerophontes at Iobates’s court, holding the letter, or also Iobates
reading the letter (see below, 223). Because of the centrality of Iobates to these paintings, it has been
suggested that they refer not to Euripides’ Stheneboea, but to Sophocles’ Iobates. Somehow, Sophocles
must have managed to get the information from Proetus to Iobates, and a letter is a reasonable guess—
but it is simply impossible to say anything more with the material we have.
 The Athenian Dramatic Stage 201

This happened most likely in the words of Demeter when advising Triptolemos
(Ł F ’ K çæe º ØØ  f K f ºª ı). Again a metaphor, but a very
different one, is used in fr. 811 inc. fab. Radt, ‘The oaths of women I write in
water’ (‹æŒ ı Kªg ªıÆØŒe N o øæ ªæ çø). The lack of context renders
interpretation difficult, but clearly in this passage the stability inherent in writing
(and in oaths) is played off against the impermanency of water, to the disadvan-
tage of women.57 Another instance of a metaphorical reference to writing,
pointing, however, to a very normal, everyday use of it, appears in the Oedipus
King: Teiresias refuses Oedipus’ accusation, and states that he will not accept
being considered (but the Greek has ‘to be inscribed’) as a client of Creon ( P
˚æ   æ   ı ªªæ ł ÆØ, 411).
Writing of a different kind is briefly alluded to in the Antigone. In a famous
passage, the non-written and stable laws of the gods are pitted against the kerygma
ordered by Creon:
It was not Zeus who made this proclamation (› ŒÅæ Æ  ), nor was it Justice who
lives with the gods below that established such laws among men ( Ø  . . . uæØ
 ı), nor did I think your proclamations (a a ŒÅæªÆŁ’) strong enough to have
power to overrule, mortal as they were, the unwritten and unfailing ordinances of the
gods (u ¼ªæÆÆ ŒIçƺB ŁH | ØÆ ÆŁÆØ ŁÅ ª ZŁ æ æÆE).
(Soph. Ant. 450–5, text and trans. Lloyd-Jones)
This passage illustrates perfectly Easterling’s point about anachronisms: if non-
written laws are here mentioned, then written laws must exist as well—but they
are nowhere mentioned. References to nomoi in the play are frequent (there are
eighteen instances, not including our passage, where Antigone refers to ØÆ);
and they will have brought to the minds of the public the familiar texts of the city
decrees. (In this context it is worth noting that at v. 448 Antigone admits that the
ŒÅæıåŁÆ were KçÆB, ‘manifest’, using a word from the register of vision.) Yet
it is nevertheless the case that in our passage the non-written laws of the gods are
not pitted against a written law, but against an oral proclamation by a mortal.58
This is, however, not just a matter of avoiding anachronisms: written documents
may be authoritative, but they become so only after a public proclamation; in a
number of cases the written text itself states that the proclamation has to be
repeated for the decision to be ‘reactivated’. The number of references to terms
with the root ŒÅæıª- in the play (twelve times) makes it clear that public
proclamation is important (that public, authoritative proclamations may still be
wrong is another matter). Thus writing is alluded to, as part of the debate; but the

57
The sentence quickly became proverbial and found fortune among the comic poets, who further
played on the image: cf. Philonides (a contemporary of Aristophanes), fr. 7 K.–A.: ‹æŒ ı b  ØåH N
çæÆ Kªg ªæ çø, ‘I write the oaths of the adulterers in ashes’, with an allusion to the punishment
(depilation with hot ashes) inflicted to those caught in the act, Ar. Nub. 1083; and Xenarchos (a fourth-
century comic playwright), fr. 6 K.–A.: ‹æŒ  ’ Kªg ªıÆØŒe N r  ªæ çø, ‘I write a woman’s oath
in wine’.
58
Arist. Rhet. 1368b5–25 distinguishes between Y Ø   , partly written, partly unwritten,
because it consists of customs, and Œ Øe  , the universal, unwritten law of nature. For the latter,
cf. Soph. OT 865 ff.; Thuc. 2. 37. ØÆ, observances sanctioned by usage, is the more correct word: so
Plat. Leg. 793A. I am not entering here on the legal issue of ‘who is right’; on this, see Harris 2006: 41–80,
esp. 53–7, for the relationship of written and unwritten laws, and 61–6 for the Antigone.
202 Letter Writing and the Polis

issue is one of levels of authority and of correct decision-making, and not of

writtenness against unwrittenness.59
Among the plays preserved in their entirety, the only one in which writing has
an important role is Trachiniae: two written documents figure in the plot, one real
(an ancient tablet written by Heracles for his own use), the other one metaphor-
ical. Throughout the play, information about these two ‘documents’ is delivered in
steps: piecemeal information concerning the first tablet spans the entire play, so
that complete clarity and understanding comes only at the end; as for the second,
its action is exerted mainly in the central part of the play—but here too, the details
are learnt a posteriori.
In the prologue, the audience is alerted to the existence of a tablet ( º ), left
by Heracles to Deianira; now that the hero has not been heard from since fifteen
months, this tablet makes her fear for the worst (46–8).60 Slightly later, when
talking to her son Hyllus, Deianira affirms that on leaving, Heracles entrusted
prophecies to her (ÆEÆ Ø , 77), and gives some hints as to their content;
that these prophecies and the tablet are one and the same thing becomes clear
later, during the first episode, in the course of a dialogue between Deianira and the
chorus. Deianira recounts that Heracles, who had never before left any instruc-
tions, this time left in the house an ancient tablet, inscribed with signs (’ K
 Ø | ºØ ƺÆØa º  KªªªæÆÅ | ıŁÆŁ’, 156–8), while also
giving oral instructions. But instead of setting Deianira’s mind at peace, instruc-
tions and tablet (especially the latter) keep worrying her: for before, Heracles had
always simply left, and the assumption had been that he would come back at some
point; this time, he acted as if he was going to die, giving testamentary instruc-
tions, and leaving the tablet as a kind of substitute—a substitute that does not help
Deianira at all.61
The oddity of the word ıŁÆÆ, used to describe the writing on the tablet,
has not escaped interpreters: a ŁÅÆ is an agreed sign, a token; the use of
ıŁÆÆ here marks writing as something mysterious, while also expressing the
uneasiness of Deianira in front of these signs (it has been suggested that here, too,
an echo of the ÆÆ ºıªæ inscribed by Proetus on the tablet given to Beller-
ophontes may be lurking).62 However, it is not just the shape of the signs, it is also
their content, their meaning, that is difficult to understand: the tablet is the
transcription made by Heracles for himself (cf.  ºÆØ . . . NªæÆł Å, 1165–7)
of an oracle rendered at Dodona, predicting that after an absence of one year and
three months Heracles would either die or live a life free from pain. Only too late
do the characters understand the real meaning of the oracle, only too late do they
realize that the second alternative is not an alternative: freedom from pain means
death (not coincidentally at the end of the play the tree that rendered the oracle is

59
There might be a specific point in the juxtaposition of ‘unwritten’ and ‘stable’ (IçƺB): Antigone
might be implicitly contesting here the commonly held notion that writing is more stable than words.
60
On vv. 44–8, often suspected of being an actor’s interpolation, see the excellent defence of Davies
1991: 66, pointing out how ‘the crucial information is allowed to leak out slowly and at dramatically
effective stages of the play . . . as . . . a progressive, climactic revelation of truth’.
61
Magini 1998: 41–2, with bibliography.
62
Davies 1991: 93, with further references.
 The Athenian Dramatic Stage 203

characterized as an ‘oak of many tongues’,  ºıªº ı æı, 1168).63 The

mysteriousness and obscurity of the written document is thus brought back to
the ambiguities of oracular speech; these will be clarified by Heracles in person at
the end of the play. But for much of the play, the tablet with its mysterious writing
functions as a substitute for the absent Heracles.
Besides Heracles’ tablet, another tablet haunts Deianira, one of her own
making; but the audience will learn of its existence only after it has exerted its
power. Having learnt about Iole, Deianira proposes to give to the herald Lichas a
‘message of words’ (ºªø  KØ º , 493) and a gift, and to this end enters the
house; meanwhile, the chorus sings of the mighty power of the Cyprian goddess,
narrating how Achelous and Heracles fought for Deianira. When Deianira, soon
after, re-enters the scene, it is to explain to the chorus what she has done: she is
sending Heracles a tunic, on which she has smeared the blood of the Centaur
Nessus. This blood (an ancient gift, ƺÆØe Hæ , hidden in a vase of bronze kept
in the house, 555–6) was the result of Heracles’ shooting an arrow at the Centaur,
as the latter, taken by desire for Deianira, had laid hands on her (again, a story of
desire, similar to the desire of Heracles for Iole, and of Achelous for Deianira,
remembered both in the prologue and in the song of the chorus). While exhaling
his last breath, the Centaur gave Deianira instructions, which she now repeats
literally, in direct spech, as if the Centaur were still speaking: the blood, if she
would collect it and keep it, would be a charm controlling the mind of Heracles
(ÆØ çæ  Ø  F ŒÅºÅæØ  | B  HæÆŒºÆ, 575–6), so that he would
never love another woman. Nessus’ words too are duplicitous, just like the oracle
concerning Heracles, true and false at the same time: after his death, Heracles will
not be able to love any other woman. The efficacy of the blood will prove terrible;
but the words of the Centaur can be shown to have already exercised a charm over
the mind of Deianira. She is still hesitating (that is why she consults the chorus),
but her precise remembrance of the words of the Centaur shows that she too has
been seized by the fatal force of desire, and not just now, but since that time.64
Later, after having sent the robe, Deianira will be overcome by fear, realizing
that the wool with which she has dyed the robe is vanishing, crumbling; she will
then go back to the words of the Centaur and to her own attitude towards them
with a new understanding. This time, she promises to tell the entire story (678):
for
ÆæBŒÆ ŁH P , Iºº’ Kø
fi ÇÅ, | åƺŒB ‹ø Ø  KŒ º ı ªæÆç
(Trach. 682–3).
‘I neglected none of the instructions [of the Centaur], but observed them, like writing
hard to wash away from a bronze tablet.’ (Trans. Lloyd-Jones)
There are a number of striking features in these two lines. First, the usual
metaphor of writing on the tablet of memory is here replaced by a comparison.

63
See Magini 1998: 38–40 for an excellent discussion of the passage. Davies 1991: 253 defends
Elmsley’s and Dobree’s conjecture K ªæÆł  at 1167 (instead of the transmitted NªæÆł Å),
adducing a parallel passage from Ar. Av. 982: (åæÅe) n Kªg Ææa Iººø  K ªæÆł Å. In both
Aristophanes and Sophocles (independently of whether one changes the transmitted text of the latter),
the middle voice indicates that the consulter of the oracle had the answer written out for him.
64
See Segal 1986: 94–5.
204 Letter Writing and the Polis

Deianira remembers, preserves the exact words, as a writing hard to wash off does:
she becomes herself a writing. Then, there is the qualification ‘brazen’ for the
tablet: this is the only time in drama that such a precision is given.65 Here this
epithet serves to emphasize the hardness, the stability with which the writing has
implanted itself in Deianira’s mind, as opposed to the ease with which a message
written on a waxed tablet could be erased; and to show how, in Deianira’s mind,
the Centaur’s words have risen to the level of an ordinance, a law (Ł , 682).66
But the reference to bronze also helps correlate the tablet, the poison (hidden in a
bronze vase, ºÅØ åƺŒø fi ŒŒæı , 556), and the robe (folded, ı Æ ,
and sent in a container, Çıª æø fi , 691–2, whose material is, however, not
specified). As for the hapax Ø , it emphasizes the permanency of the
writing; and yet it is not really appropriate, since bronze would be engraved, not
(only) painted. It is, however, exactly this lack of appropriateness that brings
about a deeper layer of understanding through an intertextual gesture towards
Aeschylus’ Agamemnon: the writing difficult to wash away, a writing that is a
person, Deianira, will be wiped out all too easily, exactly as Cassandra had likened
the humans to a ªæÆç that one stroke of a sponge may wipe away when
misfortune hits (Aesch. Ag. 1329).
The dyed robe is thus the transposition into artefact of the words written in
Deianira’s mind as on a tablet. This robe she sends to Heracles, both as a gift67 and
to accomplish her vow for his safety. (It is the first time such a vow is mentioned—
from the point of view of the audience, Deianira is advancing fast down the path
of dissimulation; as will be understood later, she had all along kept a secret in her
house—the poison, and in herself—the ‘writing’: the second is but the internal-
ization of the first). The herald, Lichas, pledges to behave like a good messenger
(this time: for he had delivered a false message earlier), and to bring directly the
sealed vessel with the robe to Heracles, adding to it the assurance of her words
(Ø ºªø, 623), an expression that closely recalls the ºªø KØ º 
promised by Deianira before entering the house.68

65
A º , ‘tablet’, is simply a º , although some words may imply a certain type of material,
such as ŒÅ, literally ‘fir’, and hence something made of wood, in Eur. Hipp. 1253–4 (see below, 221).
66
Magini 1998: 42, who compares Plut. Rect. rat. aud. 47 e: Cleanthes and Xenocrates, being slow in
learning, compared themselves to bronze tablets (ØÆŒØ åƺŒÆE), on which words are incised slowly,
but which retain them faithfully.
67
A gift from Deianira to Heracles, in exchange of his own gift (Iole?), is first mentioned at 494, Id
æø HæÆ; at 555 the term refers to the ƺÆØe Hæ  of the Centaur, old as Heracles’ tablet is old; at
663 æÅ’ is used for the peplos, as also at 668, in a question of the chorus; at 692 Deianira recounts
how she has put her gift, Hæ , in a container—but she begins to fear; at 758 the æÅÆ is qualified, in
Hyllus’ words, as ŁÆ Ø  º ; at 776 Heracles receives the æÅÆ; at 871–2 the Nurse connects
the gift to the death of Deianira: t ÆE , ‰ ¼æ’ d P ØŒæH ŒÆŒH | qæ  e Hæ   HæÆŒºE e
Ø .
68
Ø is another important theme, running through the play: at 77 Ø defines the prophecies
(ÆEÆ) left by Heracles to Deianira, i.e. the tablet: reliable, certainly, but definitely obscure; at 286
Lichas defines himself as Ø, while lying to Deianira; at 398 Deianira asks of Lichas whether he will
give her the truth, e Øe B IºÅŁÆ; at 541 she marvels at Heracles’ being known as Ø, he
who has sent home this girl; at 588 and 590 the term refers to the chorus and Deianira’s belief in the
power of the ‘love philtre’; at 623 it appears in Lichas’ promise to deliver Deianira’s faithful words; at
the end of the play it marks Hyllus’ attitude towards his father (1182, 1251). But for Hyllus, not many of
the characters mentioned in the play are ‘reliable’; the writing in a sense is, but it perpetuates unreliable
information.
 The Athenian Dramatic Stage 205

So what are the connotations, what is the function of writing in Trachiniae? In

the play, writing is, in both cases, the faithful transcription of an oral message: if it
turns out to be deceptive, the reason lies not in any intrinsic ‘deceitfulness’ of
writing. Communication in a larger sense is at issue in the play, and writing is only
part of it; with Lichas, for instance, the play offers an instance of untruthful
speech, as it does with the words Deianira sends to Heracles. However, the two
tablets allow for the inclusion of a deep chronological dimension, whereby the
past is brought tangibly (physically) to bear on the present: the words of Nessus
(or of the oracle at Dodona) might have been forgotten but for their inscription;
moreover, in the case of Deianira’s metaphorical writing what had once been a
loud pronouncement is silenced and made for ever private through an inner,
secret writing, a writing that mediates between the civilized world of humans and
the bestial one of the Centaur, and that renders possible a second performance of
the very words of the Centaur.69 But most of all, the use of writing for both
pronouncements, the oracular one and the words of the Centaur, marks (a
posteriori) Heracles’ death as something inescapable, fixed in writing years before
the beginning of the action.70
If no other Sophoclean play refers explicitly to writing, a number of fragments
present tantalizing references to it. Thus, the fr. 144 Radt, from the Achaion
syllogos (‘Gathering of the Achaeans’), has someone, probably Odysseus, suggest-
ing to another character (Agamemnon) that ‘you, on your chair, and holding the
folds of the writing, mark off any who has sworn the oath but is not present’ (f ’
K Łæ ØØ ªæÆ ø ıåa åø | ’ Y Ø P  æØ n ı ). The
writing here mentioned is a list, a long list since it has folds, and an authoritative
one, deriving its authority from that of the person who reads it, sitting on a throne.
Power and written lists, seen as instruments of control, in this passage go hand in
hand. Writing was also certainly at least alluded to in the plays connected with
Palamedes; and a few more references to writing, of uncertain interpretation, are
present in plays linked to the Trojan war.71 In particular, writing may have played
a role in the plot of Sophocles’ Poimenes, a play that narrated the arrival of the
Greeks at Troy and, among other things, the deaths of Protesilaus and Cycnus (the
fragments do not allow any further steps in reconstructing the play—it cannot
actually be determined whether this is a tragedy or a satyr play).72 The exact
meaning, in the context of the play, of the expression ‘Phoenician letters’ (fr. 514

69
A point stressed in Segal 1985: 223. Again the comparison with defixiones is striking: in quite a
few cases, re-actualization of the writing through reading it aloud changes the magic power of the
tablet, activating it or rendering it inactive. See Davies 1991: 156 on ƺÆØe Hæ  at 555: the gift is old,
the tablet is old, Nessus persuaded Deianira long before (1141–2).
70
See on these issues Magini 1998: 44–7, although I would disagree with her final view of writing in
this play as entirely positive; and Steiner 1994: 84–6, who emphasizes the talismatic status of Heracles’
oracular tablet.
71
Fr. 429 Radt, from the Nauplius navigans, mentions the invention of a ªæÆÆ, the
draughts-board with five lines; fr. 177 Radt, from the Helenes apaitesis, is irrecoverably corrupted; it is a
very negative portrayal of Helen, depicted as she ‘tortures her cheek, lately faded, with pencils that she
digs in’ (Lloyd-Jones’ translation).
72
See Rosen 2003, who does not, however, discuss fr. 514 Radt. The hypothesis of a satyr play is
attractive: it makes sense of some aspects difficult to explain otherwise. References to writing may have
been relatively frequent in satyr plays: see discussion below, 236–9.
206 Letter Writing and the Polis

Radt = Hesych. ç 688 ØØŒ Ø ªæ ÆØ·  ç ŒºB — ØØ. Kd ŒE

˚   ÆPa KŒ ،ŠŒŒ ØŒÆØ), is unclear. Is this an allusion meant to
stress, for whatever reason, the Phoenician origin of the alphabet? A way of
referring in general to the alphabet? Or might a letter (or letters) written in
Phoenician script (i.e. in a non-Greek alphabet) have played a role in the plot?
Why writing should have been mentioned at all in a play narrating the arrival of
the Greeks at Troy, a play in which the chorus was formed by shepherds, is
anyone’s guess. However, an informative element is offered by fr. 520 Radt,
ÆæÆ ªªÆØ, a Persian measure of distance, but here, in this specific context, a
Persian word for messengers.73 Possibly the reference to Phoenician letters was
not just a passing, incidental one, and (foreign) writing played some role in the
plot.74
Finally, Philomela’s weaving of her story on a ‘patterned gown’ (fr. 586 Radt,
from the Tereus) will also have been a type of writing: fixed signs are, however, in
this play needed to overcome not distance, spatial or temporal, but enforced
silence. Already in the Homeric poems, weaving had been presented as the female
way of narrating a story (or of suspending it: think of Helen and Penelope). In the
Tereus, weaving substituted for writing in a situation in which communication
through direct speech had been made impossible by the brutal mutilation inflicted
on Philomela. The play is lost, so it is impossible to know what exactly was woven
on the cloth; writing seems on the whole more likely than a pictorial representa-
tion.75 At any rate, a verse quoted by Aristotle in his Poetics (1454b30), ‘And there
is “the voice of the loom” in Sophocles’ Tereus’ (ŒÆd K fiH  ç Œº ı !ÅæE  B
ŒæŒ  çø, fr. 595 Radt), shows that the shuttle was endowed with a voice, as
is more usually the case with tablets.76
In his use of writing as a metaphor for a permanent remembering, Sophocles
thus follows in Aeschylus’ footsteps; but he also innovates, when he reformulates
the image as a comparison in Trachiniae, or when he uses the image of writing
oaths in the water to highlight the untrustworthiness of female discourse. Besides

73
Fr. 520 Radt = Claudius Casilo, —Ææa  E ØŒ E Þ æØ ÇÅ Æ ed. Miller, Mél. 397 ≊
Lex. Rhet. Cantabr. 32 (= Lex. Gr. Min. 83), 6 Houtsma: ƪª  ÆØ b ƒ I ºº Ø ŒÆº FÆØ·
 ç ŒºB b K — ØØ ŒÆd ¯PæØ Å K ŒıæÆØ ÆæÆ ªªÆ ÆP f ŒŒºŒÆØ· KåæB b NE
ƪª  Æ· › ªaæ ÆæÆ ªªÅ æ  Kd › F. Cf. Hesych.  658 ÆæÆ ªªfiÅ· <Iª>ªºø fi . ƒ —æÆØ
 f Øƪªºº Æ oø ºª ıØ. The play may have been performed in the sixties (Heubeck 1979:
108; but see Radt’s sceptical comment).
74
Corcella 1986: 60–1 discusses the passage, and concludes that here phoinikia is simply an
‘epithetum ornans’, and that phoinikia grammata simply means ‘the alphabet’, comparing with two
verses of Timon of Phlius (active in the fourth century bc): ªæÆÆØŒ, B h Ø IÆŒ c P
I ŁæÅØ | I æd Ø ÆŒ ø fi Ø،،a ÆÆ ˚  ı, Suppl. Hell. 835.
75
Dobrov 1993: 204–6 and nn. 38 and 47, 222–3 (writing—and glossectomy—as Sophoclean
innovations); Fitzpatrick 2001: 97–8, who advances the suggestion that literacy may have been an
issue in a play featuring the confrontation of Greek and barbarian. For Steiner 1994: 36–8, woven
garment and writing collapse into one token.
76
One would expect the woven tissue to speak, rather than the instrument (the loom or the shuttle);
but the notion that the ξΠsings is well-engrained in Greek tradition (see Restani 1995, and cf. Soph.
fr. 890: ŒæŒ  o Ø, with P.Oxy. 4807 and the further remarks and references of Mülke (ad loc.);
Eur. fr. 528a K.: ŒæŒ  I Ø F; Eur. IT 222–3; and metaphorically,  Æ I ÆÆÆØ çÆÆØ
ŒæŒØ Ær Æ Lyr. Adesp. ap. Stob. 1. 5. 11= fr. 1018 Campbell). Ample discussion of weaving in tragedy
in Stieber 2011: 315–34 (see in particular 315–6 for Sophocles’ Tereus, and 331–4 for weaving
metaphors).
 The Athenian Dramatic Stage 207

metaphorical inscription in lists, actual reading off a list was probably presented
on the stage. This might have been a further innovation. To judge from what
tragedies survive, the hint of written laws in the Antigone was likewise new. Two
written documents played a central role in one of Sophocles’ surviving plays, their
content, revealed bit by bit, accompanying and steering the course of the action.
Finally, it is uncertain whether letters were brought on stage (the Andromeda is an
unlikely candidate, but the Iobates is a plausible one, as well as fr. 784 Radt); issues
of communication, involving both oral and written transmission of messages,
were certainly touched upon. Comparison with the Aeschylean use of words from
the stem ªæÆç- to denote paintings shows another difference: there is not one
instance in the extant verses of Sophocles of the use of ªæÆç to indicate a painting
(the closest instance might be the already mentioned writing difficult to wash out
of a bronze tablet in Trachiniae 683, which may resonate with the image used by
Cassandra in the Agamemnon). Possibly Sophocles was not interested in a con-
nection between language and the visual arts; but this impression may well depend
on the unrepresentative nature of the surviving evidence.77 Thus while references
to writing and written documents are not evenly distributed in all extant plays,
written documents could be physically brought on the stage as props; in at least
one play they have an important part in the plot.

5.1.3. Euripides

A marked change is visible in Euripides. Metaphorical uses of writing to indicate

intellectual activity are absent; there are in fact very few instances of metaphorical
writing altogether.78 However, the range covered by forms of writing is much
wider, and includes letters, laws, references to books, funerary epigrams, victory
inscriptions, paintings, in other words, the whole gamut of ‘real-life’ writings that
surrounded the audience. Some of the Euripidean references to writing simply
continue the earlier tradition: for instance, the image of the painting easily wiped
out is taken up in a fragment of the Peleus, but modified into a comparison.79
Euripides’ description of images (especially paintings) through words employing
the ªæÆç- stem, while corresponding to one of the normal uses of the term in Greek,
is often very specific in its pairing of image and sound. In the Troades (produced
415 bc), for instance, Hecuba claims never to have set foot on a ship, yet to know
about them from seeing them in pictures and by hearsay: the two channels of

77
Hall 2006: 115 n. 60, on the basis of the lost Tereus; but see above for the likelihood that a textual
message, rather than a pictorial one, may have been woven in the peplos handed to Procne.
78
ˆæÆ, the ‘finishing line’, is used twice metaphorically, in fr. 169 from the Antigone (of evils),
and in the Electra, v. 956 (of life). An instance of metaphorical ‘writing off ’ a list appears in the Electra,
v. 1073: the heroine affirms of a wife who takes care of her beauty when her husband is away, ‘strike her
off the list as a whore’, Ø ªæÆç’ ‰ sÆ ŒÆŒ, possibly echoing Clytaemestra’s words in Aeschylus’
Choephoroi 699 (see above, 190).
79
Fr. 618 Kannicht: e Zº  P b P Æ F Œæø æ  E, | ‹ ª’ K ƺçØ Þfi A  j ªæÆçc Ł.
The ‘wiping out’ motif reappears in fr. 1041. On ‘wiping out’ see now Stieber 2011: 172–8, with an in-
depth discussion of the uses of K ƺçø.
208 Letter Writing and the Polis

information are put side by side.80 Similarly, in a striking fragment of an unidenti-

fied play, painting and speech are paired in their inability to describe what an evil a
woman is: ‘there could be no such picture drawn nor could speech describe it’ ( P ’
i ª Ø ªæ Æ  Ø F  ªæÆçB fi | P ’ i ºª   Ø).81 Another instance of
this pairing occurs in the Hippolytus (produced 428 bc): the eponymous character
defends himself from knowing anything about ‘such a practice’ (love) except for
what he has heard or seen in paintings ( PŒ r Æ æA Ø   ºc ºªø fi Œºø |
ªæÆçBfi  ºø, 1004–5). In fifth-century thought seeing (ZłØ) is considered
the best among the tools of knowledge, while hearing (IΠ), being mediated by
some other narrative, usually receives only second place. Here, however, they are on
the same level, because the seeing is ‘distanced’ by the fact that what is seen is not the
thing itself, but a ªæÆç, an image: this kind of seeing, the seeing of letters or images,
is on the same cognitive level as hearing, since letters or paintings ‘speak’. References
to paintings or woven images are also made in the Ion (produced between 418 and
413 bc) through the use of ªæÆç and ªæ ÆÆ, respectively at v. 271 (and here
again, the visual/written source of information is paired with an auditory one,
ŁıÆØ, 265), and at v. 1146.82 And there is an imaged use of  ªæ çø in
the Heracles: the hero, having recovered sanity but not knowing yet what he has
done in his attack of madness, asks his father ‘tell me, if you are sketching in outline
some new change to my life’ (Y’ Y Ø ŒÆØe  ªæ çfi Å TfiH ø fi , 1118). This is a
particularly interesting instance, because of its strong hint of metatheatricality: as
Foley has pointed out, the madness of Heracles is here ‘written over’ the story of
Heracles: the hero feels that he is being ‘redrawn’ or rewritten; but the use of
 ªæ çø may imply that the rewriting was already there from the beginning,
underwriting Heracles’ life, and that it is only now becoming evident.83
Literal writing is conspicuously absent from the Phoenician Women’s version of
the Aeschylean shield scene. The only remnants of what had figured so import-
antly in Aeschylus are to be found on the two occasions when ªæÆç is employed
for images. The likening of Hippomedon’s dazzling face to a picture, possibly
silhouetted (Iæøe <‰æ> K ªæÆçÆEØ, 129–30), is one of the relatively
few passages where a male is compared to an image; it elicits terror rather than
pity. The word is also used in the picture of a hundred snakes filling up the shield
of Adrastus (ŒÆe Kå ÆØ I KŒºÅæH ªæÆçB fi , Phoen. 1135, the second
shield scene).84 Thus, where Aeschylus in his Seven, even while using images,

80
Eur. Tro. 686–7: ÆPc b hø Æe NÅ Œ ç , | ªæÆçB fi ’ N FÆ ŒÆd Œº ı’ KÆÆØ,
where the graphe appears as the source of both visual and sonorous information. Cf. Stieber 2011:
221–2 (‘Paintings as Instructors’).
81
Fr. 1059, 5–6; I follow the translation of Collard and Cropp 2008; see also their note on the
strained Greek of l. 5, for which a number of conjectures have been advanced.
82
Ecphrasis is a prominent aspect of the play; Zeitlin 1994: 147–56 shows how Ion is so deeply
immersed in images that he is not convinced by Creusa’s words concerning his father, and needs a
visual sign: the final epiphany of Athena.
83
Foley 1985: 94 n. 54. For a detailed discussion of the meaning of  ªæ çø (to sketch in outline,
as in the ‘underdrawing’ preparatory for a painting; hence here to provide a partial vision) see Stieber
2011: 237–40.
84
See Mastronarde 1994: 187–9 for discussion of the difficult passage concerning Hippomedon,
and Stieber 2011: 262–7; Zeitlin 1994: 174–8 for the first viewing (Antigone’s), 178–85 for the second,
and 171–96 for a general discussion of how ‘hyperviewing’ (moments in which words and image join in
pictorial language, Zeitlin 1994: 145, thus forcing the spectators to reflect on their own viewing)
 The Athenian Dramatic Stage 209

had avoided the term ªæÆç, instead focusing with this term on the ‘literal’ writing
on the shield, Euripides here shifts the term onto the images and dispenses with
‘literal’ writing. While it is absolutely true that references in tragedy to something
seen only in art generally imply ‘the speaker’s lack of first-hand experience or the
monstrosity or foreignness of the thing referred to’,85 it is all the more fascinating
to realize that a form of ‘writing’ is the means to apprehend and understand the
unknown.
Of course, explicitly creating ‘images’ implies creating viewers, internal ones
but also—implicitly—external ones, the audience. This is apparent in a striking
passage of the Hecuba. On finding out about the death of Polydorus, the queen
addresses Agamemnon, asking him to punish the assassin of her son and show
respect for divine justice. To reinforce her plea, she adds:
YŒØæ  A, ‰ ªæÆç  I ÆŁd | N F  ŒI ŁæÅ  x åø ŒÆŒ .
(Hec. 807–8)
‘Have pity on us, and standing back like a painter look on me and scrutinize the evils
I bear.’
Hecuba objectifies herself and her sufferings on the stage (the actor may have
taken a striking pose, possibly modelled on some famous contemporary art-work,
or on a well-known schema); she presents herself as a painting, for Agamemnon
(and for herself) to look at—but for the audience as well: she thus presents a
spectacle of second degree.86 What Hecuba does to Agamemnon is no less
interesting than what she does to herself: she puts him in the position of the
artist—painter, but also writer: ªæÆç, and, in this play, in this moment, the
‘author’, the person in control of her destiny. This is definitely a case of ‘hyper-
viewing’—the audience is invited to look at Hecuba as much as Agamemnon is—
and there may be here more than a hint of metatheatrical reflection. As we shall
see, Agamemnon will refuse to be put in the role of the artist, he will refuse to help.
What plays itself out in this passage is the uncertainty, the lack of firm distinction,
and ultimately the struggle over the roles of subject and object, of painter/writer
and image.87
Just as interesting are other new elements. For the first time in extant tragedy,
there are explicit references to written laws. One appears fleetingly in the Ion:
having learnt of Apollo’s violence towards a friend of the Athenian visitor, Ion
wonders, ‘How is it right that you who prescribe laws for men incur a charge of

functions in this play; Hall 2006: 134 for the gendered distinction between eliciting pity and eliciting
terror.
85
Mastronarde 1994: 187, who lists, besides Phoen. 129, Aesch. Suppl. 282–3; Eum. 49–51; Eur.
Hipp. 1005; Tro. 687; Ion 271.
86
Zeitlin 1994: 142; Steiner 2001: 51; Stieber 2011: 223–5. This outward move of Hecuba’s own self
has been prepared by vv. 726–51: when Agamemnon enters and addresses her directly (726, ‘Hecuba’,
at beginning of verse), she does not answer, but addresses herself ( Å —KÆıc ªaæ ºªø ºª ıÆ
, |  EŒ Å— æ ø, ‘Unhappy one—in saying you, Hecuba, I mean myself—what shall I do?’, 736–
7), in a long interior monologue (736–8, 741–2, 745–6, and 749–51, until at 752 she addresses
Agamemnon, again at the beginning of the verse), punctuated by Agamemnon’s uncomprehending
remarks.
87
Hall 2006: 133–4.
210 Letter Writing and the Polis

lawlessness?’ (Ion 442–3: H s ŒÆØ   f  ı A æ  E | ªæ łÆÆ

ÆP f I Æ OçºØŒ Ø;). The significance of this passage is not so much in
what the text states, for the use of writing here is hardly likely to be pregnant, but
rather in the fact that by now (the Ion was produced c.413 bc) the notion of
written laws is so self-evident that it is used also for the higher laws of the gods.88
The mild questioning of the Ion appeared also, in a much more radical form,
in one of Euripides’ Melanippe. The play is lost; but in one of the preserved
fragments, a female character (Melanippe, or possibly Hippo) mocks the foolish
belief in a distant divine justice, based on written records, and asserts the existence
of a closer, immanent justice:
‘You think crimes leap up to the gods on wings, and someone writes them on Zeus’
folded tablet (Œ¼Ø K ˜Øe º ı ıåÆE ªæ çØ Ø ÆP ), and Zeus looks at
them and delivers justice to men? Even the whole sky would not suffice for Zeus to
write (˜Øe ªæ ç  ) men’s sins on it, nor could he study them and send punish-
ment for each of them. In fact, Justice is somewhere close . . . ’
(Eur. fr. 506 Kannicht, trans. Collard and Cropp)
Euripides is here making fun of the notion of a ‘book-keeping Zeus’; as we have
seen, the idea of the tablets of destiny can be traced back to Mesopotamia, but it is
not deeply rooted in Greek thought. There are no other instances of it apart from
the ‘tablet of Zeus’ mentioned in Aeschylus’ Dike fragment (fr. 281a Radt),
discussed above. It can actually be argued that Euripides is here alluding directly
to the Aeschylean passage, pushing to its paradoxical limits the earlier formula-
tion, felt as sufficiently new to attract attention.89
More interesting are the two other Euripidean passages in which written laws
are mentioned as a positive feature of the polis because they ensure equality for the
poor and for the wealthy, or on the contrary as one of the many constraints that
bind men. The latter, negative aspect is highlighted by Hecuba in the homonym-
ous play.90 When she discovers Polydorus’ fate, the queen asks for revenge,
appealing to the strong nomos that makes men believe in the gods and live
according to a distinction between justice and injustice; if this nomos were
destroyed, there would be no more justice among men (798–805). She then offers
herself and her evils for Agamemnon to look at and take pity on. Because
Agamemnon hesitates, she adds a personal reason, Cassandra, and entreats
him again, once more using an image from art, but this time adding a voice to
the art-work. This second supplication elicits from the chorus the comment that it
is terrible how nomoi determine our necessities, turning former enemies into
friends and former friends into enemies (846–9): the plural nomoi that establish

88
Lee (1997: 207). This passage does not say anything as to the Greek gods’ literacy: the writing is
metaphorical, the gods ‘write’ laws for man, in the sense that they dictate them.
89
So West 1997: 561–2 and n. 35, who accepts that the metaphorical use of writing on the tablet to
indicate memory might have come to the Greeks together with the tablet itself (the word, for it, deltos,
has a Semitic origin), but stresses the alienness of the notion of a recording Zeus; so already Solmsen
1944. Collard and Cropp 2008: 609 refer to West, but also state: ‘An old belief reflecting the centralized
administrative record-keeping of the ancient Near East, and recommended in Aesch. fr. 281a.’
90
See Collard 1991: 23–32 for the main issues of the play, 34–5 for the date, and 174 for vv. 864–7.
The treatment of nomos and nomoi in the Hecuba is addressed in Segal 1993: 191–213, esp. 192–205,
where he stresses that the gap between free and slave can be bridged from above (the strong nomos) or
from below (all men are slaves)—the in-between is open to negotiation.
 The Athenian Dramatic Stage 211

necessities, in other words, that are limiting, are implicitly contrasted with the
singular, strong nomos. Although at the personal level admitting to a feeling of
pity for Hecuba, Agamemnon refuses, however, to do anything officially, out of fear
of the army (a larger contrast between public and private runs through this passage);
in answer, Hecuba, by now herself a slave, remarks on the lack of freedom of man:
PŒ Ø ŁÅH ‹Ø ’ KºŁæ · | j åæÅ ø ªaæ Fº KØ j åÅ | j ºBŁ 
ÆPe º  j ø ªæÆçÆd | Yæª ıØ åæBŁÆØ c ŒÆa ªÅ æ Ø.
(Eur. Hec. 864–7)
‘There is no mortal who is free: either he is a slave to money or to chance, or the
masses of the city or written laws prevent him from following his bent according to his
own judgement.’
This remark takes its meaning from the context of the play. Hecuba had already
(vainly) invoked an isos nomos (egalitarian law) when asking Odysseus to save
Polyxena; but Odysseus had pointed at the necessity of individual, specific poleis,
to honour the most valiant men. The free acceptance of her fate by Polyxena had
been a source of pride for Hecuba: notwithstanding her ruin, the queen had taken
comfort in this inner freedom. Now, Hecuba has to accept that there is no
freedom. But why writing (written laws) here? The other categories mentioned
in vv. 864–7 all find a respondent in the play (Polymestor has acted from greed;
Agamemnon does not act for fear of the army; Hecuba, her children, and the
Trojans are victims of chance); thus, the written laws must here coincide with the
‘laws’ in the plural, the principle that had been invoked by Odysseus as regulating
the various poleis. The higher kind of law, equal for all (the ison stressed at vv.
291–2,   K E  E  KºıŁæ Ø Y  | ŒÆd  EØ º Ø Æ¥ Æ  ŒEÆØ
æØ, ‘there is a law in your country, the same for free men and for slaves,
concerning blood’, and at vv. 798–805) is contrasted in the course of the play
with plural nomoi; the writtenness of the laws in v. 866 may be meant to underline
the man-made and limited character of the latter, as opposed to the higher law. In
a sense, the written laws mentioned by Hecuba are also egalitarian, since they
apply to everyone; their being paired with greed, fear of the masses, and chance,
however, makes it clear that it is a negative equality Hecuba is thinking of here.
As for the positive side of written law, this is stressed in two famous verses
of the Suppliant Women, pronounced by Theseus in his answer to the Theban
herald: against the latter, who affirms proudly that he comes from a city ruled
by one man and not by the rabble, a city where no man can fool the others
with slick, flattering words, Theseus stresses that in a city held by a tyrant there
are no common laws ( Ø Œ Ø , 430–1), for one person only has power,
holding the law by himself—and this is unjust (ŒÆd  PŒ  Y , 432).
However,
ªªæÆø b H ø ‹ ’ IŁc | › º Ø  c ŒÅ YÅ åØ.
‘When the laws are written, both the weak and the rich have justice equally.’
(Eur. Suppl. 433–4)
As pointed out by Stinton, this is the only passage in all of fifth-century
literature stating explicitly that written laws are a guarantee of equality, not in
itself a self-evident notion: ‘statutory laws in states where they were not written
212 Letter Writing and the Polis

down, e.g. Sparta, were not thought any the less binding, or less nomoi, on that
account.’91 However, the historical context (e.g. the move of written laws by
Ephialtes from the acropolis to agora; the appearance of formulae of disclosure,
such as Œ E fiH  ıº ø fi , ῾so that anyone who desires may see’, attested in
five Athenian inscriptions all dated to the 430s/420s; and more generally the
emphasis in inscriptions on publishing their text) shows that the idea that written
laws help enforce equality (and democracy) must have been widely diffused.92 The
notion that writing down the laws was a central element of their polis’ identity
may thus have been widely shared in Athens; even so, it is remarkable that it
should have surfaced, and so forcefully, in this one tragedy only. It is probably not
an accident that the play is Suppliant Women, a play in which persuasion and
writing are presented in an optimistic perspective; and it is certainly not an
accident that these words are put in the mouth of the ‘democratic’ Theseus.93
This same play presents at its very conclusion another instance of real writing,
functioning exactly as in real-life interstate alliances. After the battle, Theseus
gives back the bones of the Seven to the Argives as a present from himself and
from the city, adding that they should remember this favour, and repeat his words
to their children, handing down the memory of the help they received from
Athens from generation to generation (vv. 1169–73). The Argives agree. They
are already leaving when Athena appears and stops the proceedings: Theseus
should not give away the bones so lightly. He must first require a sworn oath from
the Argives that they will never attack Attica (1185–95). Not only that: sacrifices
must be made over a bronze tripod which Heracles once ordered Theseus to set up
at the Pythian shrine (1196–200). Over the tripod, Athena orders,
‘cut the throats of three sheep; then engrave the oaths within the tripod’s hollow belly
(ªªæÆł  ‹æŒ ı æ  K Œ ºøØ ŒØ); and then deliver it to the god who
watches over Delphi to keep, a witness and a memorial for Hellas of the oaths (ÅE
Ł’ ‹æŒø ÆææÅ Ł’  Eºº Ø).’ (Suppl. 1201–4)
The oral memory, with which Theseus had been content, an oral memory
entrusted to the Argives to preserve, is under Athena’s directions changed into a
ritual (oaths, sacrifice) of which the inscription on bronze forms an important
part. As for the order that the inscribed vase, by now symbolizing that memory, be
located in a Panhellenic centre, Delphi, following a procedure that was commonly
practised by the Greek cities, it serves the purpose of removing the memory of the

91
Stinton (in Collard 1975: 441–2); Stinton makes it clear that the distinction between agraphoi
nomoi (divine, natural, universal, international laws) and man-made, positive laws of particular cities is
very different from the point at issue here. On written and unwritten laws, and their authority, see
Harris 2006: 41–61 (unwritten esp. at 45; relationship between unwritten and written, 53–7). For the
uses of writing in Sparta, see Thomas 1992: 136–7, and below, 320–6.
92
Ephialtes: Anaximenes of Lampsacus, FGrH 72 F 13 (although note Wilamowitz’s sceptical
position on this, with Jacoby’s comments, ad loc.); analysis of disclosure formulae: Hedrick 1999:
410–11 (list in his appendix 2). One does not have to believe that this notion is true (and for instance
the issue of the authority ordering the writing down of the laws is not addressed; yet this is crucial), nor
that it was shared over all of Greece (it emphatically was not). But the law cited by Andoc. Myst. 83–6
(86: Iªæ çø fi ø
fi a Iæåa c åæBŁÆØ) attests to the centrality of written laws at Athens; later political
theory connected the development of law with the beginnings of writing, cf. Plato, Leg. 3, 680 a, and the
detailed discussion in Harris 2006.
93
Turato 1976: 181–3.
 The Athenian Dramatic Stage 213

deed from its initial Argive–Athenian context. Note also that the main purpose of
the writing does not seem here to be publicity, for the oaths are incised inside the
vase: the writing on the bronze of this very special, ‘historical’ tripod (once given
to Theseus by Heracles, and thus reflecting earlier connections between the two
cities) has a fundamentally sacred, ritual value.94
This ‘contemporary’ approach to writing is present in numerous other passages.
Both victory epigrams and funerary epigrams enter with Euripides the text of
tragedy, in their ‘written’, epigraphic form, and not any more as voices of passers-
by, but with significant modifications. In the Troades, a funerary epigram is
smoothly incorporated into the tragic text, its presence signposted by the verb
of writing that precedes the epitaph, by the reference to a poet as the author (the
term chosen for this is mousopoios), and by the term ‘epigram’ in the following
verse. In arranging the burial of Astyanax, Hecuba wonders
 ŒÆ   | ªæ łØ ¼   ı  Øe K  çø
fi ; | !e ÆE Æ  ’ ŒØÆ æªE 
  | Æ; ÆNåæe  PªæÆ ª’  Eºº Ø. (Tro. 1188–91)
‘What could a poet write upon your tomb? “This child once, out of fear, the Argives
killed”? Shameful epigram for Greece.’
The epigram is written with the language and the hallmarks of real epigrams (the
initial deictic gesture, the indefinite  ), although of course in iambic trimeters.
The irony resides in the text: in the contrast between the killing of a child and the
very gesture of erecting a monument.95 The epos had already provided a model for
some play with the conventions of epigram; in a striking passage of the Iliad, for
instance, the funerary monument had been made to serve the memory not of the
dead, but of the victor, in what was already a distortion of the genre.96 An equally
unorthodox use of generic features can be seen in other plays of Euripides. In
Alcestis, for instance, the chorus at the end of its song on Necessity provides an
imaginary monument for the eponymous heroine, and then proceeds to pronounce
the words that passers-by will say, on walking past her grave:
`oÆ  b æ ŁÆ’ I æ, | F ’ Ø  ŒÆØæÆ Æø· | åÆEæ’, t Ø’, s b Å. |
 Æ
fi Ø æ æ FØ çBÆ
fi . (Alc. 1002–5)

94
See Steiner 1994: 63–71; Day 2010: 106; on echoes of epigraphic writing in tragedy Day 2010: 59–62.
95
On the ‘void’ of the monument, its beauty, and its permanence as signifier of death in Euripides,
Pucci 2003 [1977] is fundamental; see also Segal 1993: 29–33. I keep with Kovacs the  of the
manuscript tradition, against the fiH proposed by Dobree, and the  Ø accepted by Diggle. On the
use here of mousopoios (one of the rare instances in tragedy of the use of the language of ‘making’
rather than that of ‘song’) see Ford 2002: 138. On deictic markers in ‘epigrammatic writing’ see Day
2010: 112–20.
96
Hector imagines spectators in future times, who on seeing the funerary monument of the enemy
he has killed will say: ‘This is the sema of a man dead long ago, whom once brilliant Hector killed
aristeuonta’ (Il. 7. 89–90). The most ancient Athenian public funerary epigram is dated to c.490–460
(CEG 1); the casualty lists were often closed by epigrams. Two other inscriptions, CEG 11 and 12, dated
to the mid-fifth century and to 433/432 respectively, offer instances of public epitaphs for an individual,
thus presenting a closer parallel to the situation of Astyanax here. Private funerary verse epitaphs are
attested of course much earlier: in Attica the earliest might be that of Tettichos, dated to c.575–550
(CEG 13); CEG 13–76 assembles Attic private funerary verse epitaphs dated from the second half of the
sixth century to the beginning of the fifth century. See Day 2010: 59 on Hector’s remarks, and passim
for epigrammatic writing.
214 Letter Writing and the Polis
‘ “This woman once ( ) died for her husband’s sake, and now she is a blessed
divinity.” “Hail, lady, and may you grant us good.” With such words will they address
her.’
The epigram conforms to contemporary examples (but for the metre), and it
shows well how commemoration works, even envisaging future reperformances in
their details; but immediately afterwards, Heracles will arrive, with Alcestis in tow,
defeating expectations. Hecuba’s epigram for Astyanax is, however, the only
funerary one, and moreover attributed to a  ı  Ø, a poet, for which an
inscription is explicitly imagined, if only fleetingly and paradoxically. One may
well wonder if here the pause, with the explicit allusion to a poet as a composer, is
not an allusion to Euripides himself and to the Trojan Women: after all, his play is
(also) an extended funerary epigram on the victims of the Trojan war (and, of
course, of the Peloponnesian war). Later, Hecuba will state that the shield of
Hector, mother of countless victories, will die with the boy, and yet not die
(1221–3); this part of the burial ends with the remark that if the divinity had
not overturned things, ‘being unknown we would not have been sung, providing a
theme for song to the Muses of men to come’ (IçÆE i Z PŒ i ŁÅ
i |  ÆØ I Ø a  æø æ H, 1244–5, a line that echoes Iliad 6. 357–
8). Song and lament are possibilities: one (lament) is actual, the other will be taken
up by men to come; but the writing of a funerary epigram is in this instance not a
possibility. The issue here is commemoration, and there may be in the stress on
writing a reflection on the dramatist’s own possibility of guaranteeing adequate
commemoration.97
A paradoxical victory epigram figures instead in the Phoenician Women. As
Jocasta tries to dissuade Polynices from his intent of destroying his own city, she
wonders what victory monument he will be able to erect to commemorate such a
victory:
ŒÆd ŒFºÆ ªæ łØ H K’ " å ı Þ ÆE; | ¨Æ ıæÆ    — ºıŒÅ Ł E |
I Æ ŁÅŒ; (Phoen. 574–6)
‘And what will you inscribe on the spoils98 by the streams of the Inachus? “Having set
fire to Thebes Polynices dedicates these shields to the gods”?’ (Trans. Kovacs)
Again, this is a clear allusion to contemporary practice: Jocasta’s idea is that
Polynices will bring the spoils back to Argos, and there erect a victory monu-
ment.99 This instance of victory epigram is written exactly as a real dedication

97
See Pucci 2003 [1977]: 159–62. Discussion of Tro. 1188–91 in Segal 1993: 29–33; of Alc. 1000–5 in
Segal 1993: 46–7. Segal sets up a fascinating contrast between the immobile statue in his bed, imagined
by Admetus, and the open, public, and spontaneous commemoration offered twice by the chorus, at
445–55 (Alcestis’ fame will survive in the songs of poets at the Carneia in Sparta and in Athens), and in
the last ode. More on the place of Orpheus’ writings in this play below.
98
On ªæ çø construed with the external object of the thing affected see Mastronarde 1994, ad loc.
99
As Steiner 1994: 76 stresses, the act of inscribing the spoils is presented as part of the ritual of
dedication. The earliest public Athenian dedicatory epigram for a victory epigraphically preserved is
CEG 179, celebrating the victory against the Boeotians and Chalcideans in 506 (cf. Hdt. 5. 77, with the
epigram as well). Personal dedications begin earlier, at the very start of the sixth century. These
compare better with Jocasta’s example, since the trophy would be erected by Polynices, an individual;
but although they are dedications of tithes, they are not resulting from a victory in war. CEG 256, the
dedication of Callimachus, the polemarch who died in Marathon, might be a good example of
 The Athenian Dramatic Stage 215

after a victory would have been, but for the trimeters instead of the customary
elegiac distichs; it is, however, undercut by the fact that the epigram is composed
(‘read’, as it were) by the person who will lose most from the supposed victory,
Jocasta, and imagined as erected by one of the other losers, Polynices. Here the
irony is situated outside the text of the epigram, in the tragic text. In both this
instance and that of Astyanax’ funerary epigram, the genre is by now familiar
enough to allow for some play with its conventions, and for an exploration of the
limits of the format—and of its traditional ideology.
Another gesture towards epigraphical writing is in fr. 923 inc. fab. Kannicht. It
is even stronger than those we have seen until now, since it refers not to a
‘possible’ and ‘plausible’ inscribed text, but to a ‘real’ one. Here, a character refuses
to make a pledge, stating that the letters inscribed at Delphi (probably ‘Pledge
brings ruin’) forbid him to do so (a —ıŁ E PŒ Kfi A  ªæ ÆÆ).100
Euripides is also the first to have brought ‘books’ onto the stage (or rather,
references to what must be understood, in most instances, as books). In the
fragmentary Erecththeus, for instance, the chorus tell of their desire to lay down
their spear and, in their old age, the head crowned with garlands, sing their songs,
after having hung their shields as dedications in the temple of Athena; ‘and may
I unfold the voice of the tablets in which the wise are celebrated’ ( ºø ’
IÆ ØØ ªB- | æı fi v  ç d Œº ÆØ, fr. 369 Kannicht). It is here difficult to
know exactly what kind of books are meant and who are the wise referred to, and,
in particular, whether the voice of the tablets is opposed or complementary to the
songs mentioned above. A similar reference to oral tradition on the one hand and
tablets of the Pierian Muses on the other is present in the Iphigenia in Aulis (798),
where, as we shall see below, it resonates with other aspects of the play.
One other vague reference to poetic writings (with Muses lurking in the
background) is in the Hippolytus: after she has heard of Phaedra’s love, the
nurse, looking for mythological precedent to comfort her mistress, states that
‹ Ø b s ªæÆç   H ƺÆØæø | å ıØ ÆP  ’ Nd K  ÆØ Id | YÆØ
b Zf u  ’ Mæ ŁÅ ª ø | ºÅ (Eur. Hipp. 451–4)
‘Those who know the writings of the ancients and are themselves concerned with the
Muses know that Zeus once fell in love with Semele . . . ’
The nurse is hardly a character from which one would expect an advanced
literacy (not so much because of her status, but because of the way her mentality is
portrayed in the play), so the reference to writings rather than to oral tradition

dedication by an individual (and his family) in connection with a victory. Outside Attica, Corinth and
Samos offer good instances of public victory dedications: respectively CEG 351 for the victory at
Tanagra, dated to 458–457 bc, and also mentioned by Pausanias 5. 10. 4; and CEG 421, for a naval
victory at Memphis, dated to 460–454 bc. Many more dedicatory epigrams commemorating victories
are reported in literary sources. Pausanias’ attempt to appropriate the common Greek dedication after
the victory in the Persian war (Thuc. 1. 132. 2) is a good example of the way an individual might deal
with the genre—and be punished for it.
100
Collard and Cropp 2008, ad loc., refer to Plato, Charm. 165 a, the scholia to which cite a
fragment of the comic poet Cratinus Iunior, fr. 12 K.–A.: A. ‘Having for three times given security,
I am dead’; B. ‘Wasn’t there in Delphi a writing, saying: “security, ruin” ’? ( hø ’ K ˜ºç EØ q a
ªæ ÆÆ | c KªªÅ ¼Å <ºª ’;>); A. ‘There was, but I have a friendly soul’.
216 Letter Writing and the Polis

appears surprising—to the point that the suggestion of understanding ªæÆç  as

paintings has been entertained.101 This reference to writings (and even if they
were paintings . . . ) is all the more surprising since the exempla that follow are
Zeus’ love for Semele and Eos’ love for Cephalos: well-known examples, one of a
male god, the other one of a goddess, falling in love with a mortal (a choice
reflecting the high opinion the nurse has of Phaedra, and the low opinion she has
of the ‘son of the Amazon’ Hippolytus); there was no need of books to know about
them—nor to know how badly both affairs ended.
If the nurse here stresses that it is cultivated people who know such stories,
stories transmitted in books, it must be to bring home to Phaedra that this is an
authoritative knowledge, the kind of knowledge common among Phaedra’s
equals. That Phaedra is acquainted with literacy the audience knows: for only a
few verses before, in a very complex passage, Phaedra had first identified as the
main cause for the ruin of mortals’ life not lack of judgement (for we know what is
best), but either laziness or giving priority to some pleasure. Among the latter she
had listed ÆN  (respect), adding that there are two kinds of it (or of pleasures as a
whole), one not bad, the other a suffering for the house, and that it is not easy to
distinguish between them. Phaedra’s musings, however unclear in themselves, had
ended with a very clear statement:
N › ŒÆØæe q Æç, | PŒ i ’ XÅ Æh’ å  ªæ ÆÆ. (Eur. Hipp. 386–7)
‘If what is appropriate were clear, there would not be two things designated by the
same letters.’102
This is a rather unnecessary display of literacy: for the issue here is that two
slightly different concepts have the same name (or that the same quality can work
in two ways, just as the good and the bad æØ at the opening of Hesiod’s Erga)—
and thus, having the same name, share the same letters; but the latter is a
consequence of the former. Mentioning the letters, rather than the names, how-
ever, gives Phaedra the chance to objectify her reasoning, rendering it ‘observable’.
Moreover, literacy in the play is important. Phaedra will commit suicide, as had
been her plan from the beginning, but she will also destroy Hippolytus through a
lying letter. Pointing early on in the play to her acquaintance with literacy and to
her habitual thinking in terms of it makes sense.103 The Nurse’s statement may be
considered part of this endeavour too. Interestingly, the literacy that had marked
Phaedra is, after her death, transposed by Theseus onto Hippolytus, whom he
accuses of hypocrisy:

101
Barrett 1964, ad loc.; ‘I have been tempted (with many editors) to take ªæÆç  as paintings’—but
as he shows, this make little sense in the context; so also Goff 1990: 97–8; Ford 2002: 154 and n. 90, and
Stieber 2011: 222, with further references (but note contra Easterling 1985: 6 n. 26; Hall 2006: 112 and
n. 50).
102
To the bibliography in Goff 1990: 96–7 (whose reading in this instance is not particularly
compelling) add Kovacs 1980, who thinks it is pleasure that is here referred to; and esp. Halleran 1995,
ad loc. Whether it is pleasure or aidos that are referred to here is of less import than the assertion of an
inherent doubleness in some concepts; the open reference may be intended, as Phaedra is here
expressing general considerations.
103
Slightly different in its connotations, but also part of Phaedra’s aristocratic portrayal, is Phae-
dra’s dislike of enticing speeches (ŒÆº d ºÆ ºª Ø, 487, and passim), a dislike which she expresses more
than once and which she shares with Hippolytus; see Turato 1976; Goff 1990.
 The Athenian Dramatic Stage 217
‘Now you may be proud, adopt a meatless diet and parade your food, and with
Orpheus for your lord engage in Bacchic rites, honouring the vapours of many
books! ( OæçÆ  ¼ÆŒ åø |  Œåı  ººH ªæÆ ø ØH ŒÆ .) For
you have been found out.’ (Eur. Hipp. 952–5)
Theseus is being heavily ironical;104 and as the audience knows, he is wrong in
his belief. It is even more ironical, from the audience’s point of view, that Theseus
should consider bookishness a proof of corruption, while at the same time putting
his entire trust in Phaedra’s letter.105 It is noteworthy that in order to pour scorn
on Hippolytus Theseus should turn to Orpheus and his books: the negative tone is
evident, expressing the perception of a corrupting influence, or at any rate, of
duplicity, of a screening of the truth. And although this is the contrary of what
Phaedra initially had tried to achieve by using letters in her reasoning, Phaedra too
finally turns to writing for her false accusation.106
We have just seen books mentioned in connection with Orpheus’ teaching.
‘Orphic’ writings form a specific category, well represented in the relatively
infrequent early references to books.107 ‘Orphic’ books appear in the Alcestis
(the earliest securely dated Euripidean play, produced in 438 bc), in a context in
which various forms of wisdom are put side by side: the chorus claims to have
‘soared aloft with poetry and with high thought’ ( Øa  Æ ŒÆd  æØ fi q Æ)
and to have touched on many discourses (ºø ±ł   ºªø); however,
ŒæE  P b  ªŒÆ | Åyæ  P  Ø ç æÆŒ  | ¨æfiÆØ K ÆØ, a | OæçÆ
ŒÆªæÆł | ªBæı, P ’ ‹Æ E  - | ŒºÅØ ÆØ  øŒ | ç æÆŒÆ  ºı Ø |
IØg æ  EØ. (Eur. Alc. 965–72)
‘I have found nothing stronger than Necessity, nor is there any cure for it in the
Thracian tablets set down by the voice of Orpheus nor in all the simples which
Phoebus harvested in aid of trouble-ridden mortals and gave to the sons of Asclepius.’
(Trans. Kovacs)
The comprehensive inventory of the chorus, covering poetry, astronomy, and
philosophy, ends with two further sources of wisdom and remedies, connected
between themselves:108 the Orphic tablets, relying on a rather extraordinary type
of writing (an orally dictated text!), and the Apollonian simples, relying, at least
initially, on oral transmission. Both are put on the same footing, as equally useless
against Necessity (which does not have to imply ‘useless’ tout court).
Finally, a fragment from the Pleisthenes preserves a reference to written collec-
tions of oracles, such as we know circulated already in the late sixth century:

104
Against many critics, who see in Hippolytus someone inclined to mysticism, and take this
reference to books seriously (e.g. Segal 1993, Steiner 1994: 205–6; Henrichs 2003b: 212–13 and 216, an
important discussion of this passage; Scodel 2011), see Barrett 1964: 342–5; Linforth 1941: 56.
105
Segal 1985: 219.
106
The heavy emphasis on writing in the Hippolytus may explain why Artemis uses for song the
language of ‘making’ ( ı  Øe æØÆ, 1428), when at the end of the play she announces future
ritual performances and songs for Phaedra and Hippolytus, so that their memory will not die: cf. Ford
2002: 138.
107
On Orpheus and writing see above, 66–8; Linforth 1941; Henrichs 2003b (passim, but esp. 212
n. 13 for the passage of Alcestis); Graf and Johnston 2007: 175–84; Scodel 2011.
108
Apollo, the father of Asclepius, is in some versions father of Orpheus; both Orpheus and
Asclepius had attempted to save mortals from death.
218 Letter Writing and the Polis
Nd ªaæ Nd ØçŁæÆØ ºªªæÆçE |  ººH ª ıÆØ ¸  ı ªÅæı ø.
(Eur. fr. 627 Kannicht)109
‘There are, truly there are, parchments inscribed with song, laden with many utter-
ances of Loxias.’
The transmitted ºªªæÆçE (‘inscribed with song’) is worth some attention:
on a basic level, ‘song’ can be explained as referring to the hexametric verse in
which oracular responses were couched; but on a deeper level, we have here a
compressed version of the synaesthetic image of the writing that speaks, or rather
sings, an image found, in a more complex articulation, in the Hippolytus (dis-
cussed below). The repeated, emphatic assertion of the existence of these texts
may be the result of a context in which the authority of Apollo’s oracles was
doubted; the writing would have served to reassert it. What further twists the plot
might have brought is impossible to know.
Euripides may also have been the first tragedian to bring letters onto the
stage.110 Remarkably, letters have the lion’s share among the various appearances
that writing makes in Euripidean tragedy, in terms of both their frequency and
their importance for the advancement or the resolution of the action.111
The first letter to appear among the surviving oeuvre of Euripides is probably
the deadliest, the one with which Phaedra accused Theseus’ son in the second
Hippolytus (performed 428 bc). Another such letter, written this time by a man,
but the result of the calumny of a spurned woman, had a central part in the plot of
the Stheneboea, produced in the early 420s; that it failed to engineer the death of
the intended victim was in no way imputable to epistolary error but simply to the
extraordinary courage and strength of the hero, Bellerophontes. A false letter,
written by Odysseus, played an essential role in the peripeteia of the tragic
Palamedes (probably produced c.415 bc), here too with deadly results. And letters,
two of them positively marked (but one proving in the event useless, the other not
reaching its intended addressee), a third one bearing a strong negative connota-
tion, appeared in the two plays on Iphigenia, the Iphigenia among the Taurians
(brought to the stage c.412 bc), and the Iphigenia in Aulis (first produced after
Euripides’ death in 406 bc). These overall rather negative connotations do not
apply as consistently to the other forms of writing; there is a clear difference in
tragedy between epistolary writing and writing tout court. However, it is important

109
The translation above is borrowed from Collard and Cropp (2008); see also their note ad loc.,
with references to the many oracular collections known to have circulated at Athens in this period; and
Kannicht ad loc. for full references, and a thoughtful assessment of Bergk’s correction ºÆªªæÆçE,
‘written in black ink’ (accepted in LSJ).
110
There were probably no letters, but only oral injunctions, in Sophocles’ Andromeda; the Iobates
may be later than the Euripidean Stheneboea; a written message may have played a role in Sophocles
Poimenes, and the play may have been performed c.460 (see above, 205–6 and n. 73); nothing can be
said of the date (nor of the larger context) of Soph. fr. 784 Radt (see above, 200).
111
It is worth noting here that terms implying writing are not evenly distributed in Euripides:
KØ ºÆ is attested eleven times, according to the following distribution: once each Hipp. Andr. Hel.
Bacch. (in the three latter plays with the meaning of oral injunction, as once in IT); IT five times, IA
twice. Out of eighty-nine overall occurrences of terms having to do with writing (list including only the
terms formed on ªæ çø and º ) fifteen come from nine different fragmentary plays; the remaining
seventy-four are distributed as follows: 1 Alc.; 1 Bacch.; 1 Her.; 2 Suppl.; 2 El.; 2 Hec.; 3 Tro.; 3 Phoen.; 3
Ion; 13 Hipp.; 22 IT; 21 IA.
 The Athenian Dramatic Stage 219

in this context not to forget that oral messages (and even face-to-face speech) are at
times just as untruthful and destructive as written messages.112 This means that at
stake are mainly issues of communication, forms and typologies of successful and
unsuccessful exchange, rather than a straight contrast between orality and
writtenness.113
Let us take the plays in their chronological order. The first preserved letter of
Euripidean drama is the one in the Hippolytus. The playwright envisaged this
device for Phaedra’s accusation only on a second occasion, when reworking the
plot: for the Hippolytus we have is a radical rewriting of an earlier drama, in which,
as seems most likely, Phaedra herself had first confronted Hippolytus with her
love, and then, after being rejected, had falsely accused him to Theseus.114
Tradition imputes the failure of the play to its having been ‘unseemly and worthy
of condemnation’: Phaedra had been portrayed in too blunt a way (Aristophanes
in Frogs 1043 has Aeschylus speak of her and Stheneboea as whores, æÆØ). The
task Euripides had then to face in his new version was to bring on stage a Phaedra
who, although still falling in love with her stepson, would resist her love and
behave in a more restrained way. The two main problems in the handling of the
plot were constituted by the revelation of Phaedra’s love to Hippolytus, and by the
accusation of rape: Euripides solved the first by having the old nurse reveal to
Hippolytus, against Phaedra’s order, her mistress’ love, and the second by having
Phaedra commit suicide, leaving a letter.115 The suicide-letter played a crucial
role. Phaedra could now die nobly, yet, thanks to the letter, the action could go on
towards the expected resolution. The play was a success: it won Euripides one of
his only four first prizes at the Dionysia.
The letter appears only in the second part of the play, unannounced, except for
some cryptic words of Phaedra (715–30, where she says that she has found a
stratagem, oæÅÆ, to give her sons a good reputation and gain something out of
the present misfortune);116 but its appearance had been prepared by the earlier
references to ªæ ÆÆ and to books discussed above. The first one to notice the

112
See e.g. Lichas’ initial message to Deianira in Sophocles’ Trachiniae; or Euripides’ Medea.
113
Griffiths 2007 offers a fascinating account of letters (and others written documents) as attempts
to control the course of future events; but the failure to distinguish between official, public writing and
personal, private writing limits the import of her conclusions.
114
Discussion of the relationship between the three Hippolytus plays known (two by Euripides, one
by Sophocles) in Barrett 1964: 10–45; Halleran 1995: 24–37. Both stress that this instance of rewriting is
unique among Athenian tragedians; Halleran makes the point that the earlier version must have shaped
audience’s expectations (24); see also 25–7 for a concise summary of the evidence, and in particular 26
for the unlikelihood that a letter played a role already in the first Hippolytus. For the iconography see
Linant de Bellefonds 1990: 404–64, and 1994: 356–9 (no traces of the story in monuments of the fifth
century; earliest representations on South Italian vases, privileging the death of Hippolytus. A tablet
given by the Nurse to Hippolytus, evidently a letter from Phaedra, appears in some mural paintings of
the first century ad; the scene will enjoy success in late antiquity).
115
Thus the oral message of the Nurse and Phaedra’s letter to Theseus are on the same plan: they
both ‘constitute a disastrously aggressive mode of telling women’s secrets to men in power’, Segal
1993: 90. As pointed out by Griffiths 2007: 282–3, the letter fails Phaedra as well: she may not have
meant revenge on Hyppolytus, and her reputation, her one concern, is ruined.
116
Note also the nurse’s exclamation, when she realizes the catastrophic results of her good inten-
tions: 670–1, ‘What arts do we have, what speech, once we have faltered, that can undo the knot our
speech has created?’ ( j F åÆ å  j ºª  | çƺEÆØ Œ ŁÆÆ ºØ ºª ı;). The solution
will be Phaedra’s tablet. See also åÆØ at 680, said by the chorus of the failed ruse of the nurse.
220 Letter Writing and the Polis

object is Theseus. He sees the tablet ( º , 856) hanging from Phaedra’s hand,
wonders whether it may want to indicate something new (ŁºØ Ø ÅBÆØ  ,
857), then assumes that his wife has written to him a message (KØ ºa |
ªæÆł, 858–9) with a request concerning remarriage and her children; this
is clearly the normal assumption. He opens the sealed wrapping, and looks at
what the tablet ‘wants to say’ (çæ’ K º Æ æØ ºa çæƪØ ø | Y ø 
º ÆØ º  l   Ø ŁºØ, 864–5).117 The action of Phaedra (writing a message)
is thus neatly enclosed between two references to the desire of the tablet to
speak: the tablet, first presented as hanging from Phaedra’s hand, just as Phaedra
had been hanging from a suspended noose in the description of her death
first given by the nurse,118 from now onwards functions as a substitute for the
writer.
In Theseus’ first perception, the tablet, once he has silently read its message,
‘screams, screams of grievous hurt’ ( fi A  fi A º  ¼ºÆÆ, 877);119 the actor
impersonating Theseus here switched to song. In the following verses Theseus
expresses his suffering through a synaesthetic combination of sound and visual-
ization, stating that ‘such a song I have seen, wretched me, sounding forth through
the writing’ ( x  x  r  ªæÆçÆE º  | çŁªª  º ø, 879–80).120 The
superposition of sound and image here conveys powerfully how much Theseus
has been affected by the message, a message that he cannot hold any more within
the gates of his mouth (so much has this become part of himself), and that he
proceeds to publicize, uttering it in trimeters. Theseus’ statement that Hippolytus
has dared to touch with violence his marriage bed is all the audience will ever
know of the content of the tablet. Theseus is by now set on avenging the deed: in
what he sees as an utter confusion of every fundamental norm, Theseus is asking
for clarity. From his father, Poseidon, he wants to know whether he has truly
granted him ÆçE Iæ , curses that can be relied upon (890). As for men, he
dreams of a test of friends (H çºø ŒæØ  | Æç, 925–6) that would allow
him to know men’s minds. His wish is that men would have two voices ( Ø  
çø ), one just, the other as happens, so that the just voice would convict the
other, making deceit impossible (928–31). This closely recalls both Phaedra’s own
musings on the two forms of aidos bearing the same name, thereby making it
difficult to distinguish one’s course, and Hippolytus’ own wish that women would
live with animals deprived of speech (¼çŁ ªªÆ, 646), so that they would not speak

117
The opening of the seal, and the connotations (ominous) of this specific type of seal are well
analysed in Jenkins 2006: 82. Segal 1993: 97 stresses the growth in the quasi-magical power of the tablet:
what were signs at v. 857 has become speech by v. 865. See also Steiner 1994: 38–9.
118
As suggested by Loraux 1979: 53–4 (see also Goff 1990: 38), the use of MæÅÅ (857) here for
the tablet echoes the use of the same participle for Phaedra (779).
119
The verb  ø had twice been used earlier. In the scene in which the chorus and Phaedra,
outside, hear Hippolytus, inside the house, upbraid the nurse, the chorus asks Phaedra Æ Łæ E
ÆP ; Æ  fi A ºª ; (571); Phaedra answer, ‘The son of the Amazon screams, Hippolytus’ ( fi A |
 Iºı , 581–2). The noun   is also heavily employed. In this same passage, Segal 1993: 96 has
pointed out the presence in the words of the chorus addressed to Phaedra,  º  º  d  (586),
of an anadiplosis and a personification of speech that finds a parallel in the anadiplosis and personifi-
cation of the tablet of v. 877.
120
On the ‘paradox of the silent tablet that shouts and the song that has its utterance in writing’, a
paradox corresponding to the deeper one of an absent Phaedra who can still produce such an effect, see
Segal 1985: 219.
 The Athenian Dramatic Stage 221

nor be addressed in response at all.121 In any one of these ways, with two distinct
voices or with clearly separate (written) concepts, or even with no speech at all,
things would be defined. This deep preoccupation with the difficulty of reaching
clarity, a difficulty the three main characters of the play become aware of at
different moments, links them together.
In the ensuing confrontation between Theseus and Hippolytus, two things, in
Theseus’ mind, damn Hippolytus: the death of Phaedra (945: ‘convicted by the
dead woman . . . ’; 958: ‘She is dead . . . and this convicts you most’; 972: her
‘corpse, witness most reliable’) and the tablet, invoked as the final argument
against Hippolytus’ pleas (1057–9). Against these two arguments, Hippolytus
first proposes a rational defence (983–1035); then, as this proves useless, he
appeals to time, which reveals everything (Åıc åæ , 1051), and to the test
of oaths, sworn testimony, and the words of seers ( P ’ ‹æŒ  P b Ø P b
 ø | çÆ Kºª Æ ¼ŒæØ  KŒƺE  ªB; 1055–6). Against this last
appeal, Theseus pits the tablet:
 º  l  ŒºBæ  P  ªÅ | ŒÆŪ æE  ı Ø ·  f bæ Œ æÆ | ç ØHÆ
ZæØ ºº’ Kªg åÆæØ ºªø. (Eur. Hipp. 1057–9)
‘This tablet, containing no divination by lot, accuses you convincingly; as for the birds
that fly above my head, I wish them joy.’
Ominously, the king chooses to trust the writing, lightly dismissing oaths,
testimony, and seers in a sentence whose conclusion recalls the words of Hip-
polytus against Cypris, the very words and the very attitude that had set the
tragedy in motion (‘As for your Cypris, I wish her joy!’, c c b ˚æØ ºº’
Kªg åÆæØ ºªø, 113). At the end of the play, Artemis will accuse Theseus of
baseness, because he did not wait for testimony (Ø), the words of a prophet
( ø ZÆ), or time (åæø fi ÆŒæfiH, 1320–3).
In this agon, the tablet, and whoever trusts it against the traditional oral proofs,
is the villain. This impression is strengthened by the last two references to writing
in the play. The first occurs in the speech of the messenger reporting Hippolytus’
death: after having narrated the gruesome facts, the messenger states strongly that
he would never be able to believe in Hippolytus’ guilt,
P ’ N ªıÆØŒH A ŒæÆŁÅ ª  | ŒÆd c K ” fi Å ªæÆ ø ºØ Ø |
ŒÅ· (1252–4)
‘not even if the whole female sex should hang themselves, and fill with letters tablets
made with all the pine wood of the Ida!’
In other words, the messenger transforms into a hyperbole the two main grounds
on which Theseus’ accusation had rested, and then makes no count of them. The

121
Hippolytus’ statement that ‘the tongue swore, the mind is still unsworn’,  ªºH O å ,  b
çæc I  , 612, also testifies to a split sense of things; as also, at the end of the confrontation with
Theseus, his wish, dictated by despair, to go outside of himself to look at himself (to become a
spectator): ‘Would that I could stand apart and look at myself so that I might weep at the misfortunes
I am suffering!’, 1078–9. Stieber 2011: 257–62 offers an excellent discussion of the Euripidean
mismatch between external appearance and internal reality, and shows that it is associated with images
and metaphors linked to coinage (åÆæÆŒæ is the stamped image on a coin) and to the notion of
folding/unfolding (of a written text, through the use of forms of verbs such as IÆø).
222 Letter Writing and the Polis

second and last reference to writing is in the words pronounced by Artemis when
she appears to explain the events at the end of the play:
 ’ N ºªå  c fiÅ ç  ıÅ | łı E ªæÆça ªæÆł ŒÆd غ | º ØØ e
ÆE ’, Iºº’ ‹ø Ø . (1310–11)
‘Phaedra, fearing lest she be put to the proof, wrote a false letter and destroyed your
son by guile, but even so persuaded you.’ (Trans. Kovacs modified)
Here, the message is conveyed though semantics and rhetoric (the figura
etymologica and alliteration: me- pese-i/epeise, dio-lesen/doloisi). And this is final.
Because of the death of Phaedra, writing in the play cannot be tested; already in
Phaedra’s own words, it had proved unreliable in distinguishing between positive
and negative concepts; and the nurse, a rather negatively presented character, had
appealed to its authority in a speech that was clearly in a bad cause.
Of the Euripidean Stheneboea, produced in the early 420s, only a few fragments
are preserved, none of them mentioning any kind of writing. The only clues to the
importance of a letter in this tragedy are the argumentum, which explicitly
mentions a tablet (a º ) given by Proetus to Bellerophontes, carrying instruc-
tions for Iobates that were meant to cause the death of Bellerophontes himself;
and some relatively early Italiote vases (the earliest is dated to 420 bc) in which a
man is depicted giving a letter to another man.122 The identification is assured by a
Paestan hydria dated to c.350 bc, on which an old man, labelled Proetus, and
dressed in what seems to be a theatrical costume, hands a letter to Bellerophontes.123
Since no representations of the scene earlier than 420 bc are known, although
the story was already part of the Iliad, the conclusion that these depictions are
influenced by Euripides’ play seems reasonable. The reception of the play by
the painters shows that they recognized in the letter an object with remarkable
dramatic potential and significance. Because of the almost complete loss of the
Stheneboea, it is impossible to know how exactly Euripides recreated the Homeric
story: the name of the heroine changed from Anteia to Stheneboea, Proetus was
presented as king of Tiryns rather than Argos, and Bellerophontes arrived at the
court of Proetus seeking purification for an involuntary murder committed at
Corinth. Moreover, the plot was considerably expanded, so as to include the
return of Bellerophontes from Lycia and his revenge on the couple: death for
Stheneboea, affliction for Proetus.124 The argumentum and the testimony of the
vase-paintings show, however, that the folded tablet with baneful signs still played
a central role.

122
Argumentum: Kannicht iia; Collard and Cropp 2008, test. iia. Vases: Panathenaic amphora from
Ruvo, now in Naples, Mus. Naz. 82283, LCS 44.128, Pisticci Painter, c.420 bc, Kahil 1994 no. 1; Apulian
bell-crater in Bonn, Akad. Kunstmus. 80, RVAp i, 14. 49, 420–400 bc, Kahil 1994 no. 2; Apulian
stamnos of c.400 bc, Boston Museum of Fine Arts 1900.349, in front of a doorway that suggests
theatrical inspiration, Kahil 1994 no. 3. Cf. Kannicht *iii; Taplin 2007: 201–3.
123
Hydria Paestum Mus. Naz. 20202 signed by Asteas, dated to 350–330 bc, RVP 84, 134 pls. 5–56,
cf. Kahil 1990 no. 5; also linked to the theatre because of the costume worn by the man (a Samnite
dress) is another vase, a Campanian bell-crater from a private collection in Geneva, LCS Suppl. 2, 221.
281a and Suppl. 3, 198. 281a, group of Naples 3227, dated to c.330 bc, cf. Kahil 1990 no. 6. Here too,
Proetus is giving the letter to Bellerophontes; cf. Taplin 2007: 203.
124
On the plot of the play, as far as it can be reconstructed, see the introduction in Collard and
Cropp 2008: 121–5, wth further bibliography.
 The Athenian Dramatic Stage 223

Another play centred on the story of Bellerophontes attests to the significance

of the story, the Sophoclean Iobates, possibly produced shortly after the Sthene-
boea.125 The action of the Iobates took place in Lycia, and again a letter probably
played a role: for practical reasons, because somehow Iobates must have been told
that he should send Bellerophontes to his death, but also because of a number of
Italiote vases in which Iobates (and not Proetus) is depicted in a central role
(sitting on a throne), while Bellerophontes arrives on Pegasus, a letter in his
hands.126 Evidently, the death-bearing letter struck the imagination of the Italiote
painters, who saw in it a central moment in this well-known story. Yet bringing
back any depiction (and even the explicitly theatrical ones) to a specific play is a
risky endeavour. Moreover, all of this tells us something about how Bellero-
phontes’ story, and the role played by a letter in it, was perceived by Italiote
painters, but it does not tell us much about the mainland Greeks’ perceptions.
What is at any rate clear is that in the plays built around the story of Beller-
ophontes, a letter is used to transmit an order to kill; in the first part of the story,
the striking aspect of the affair resides not in the content of the letter, but in the
fact that the intended victim is the bearer: only a letter, that is, a sealed piece of
writing, can achieve this. In the second part (the arrival in Lycia), the same
element takes a slightly different slant: the letter, once open, becomes a ‘piece
for the accusation’, in the absence of the mandants (Proetus and Stheneboea).
Unfortunately we do not know, not having the plays, whether the text of the letter
was quoted and how exactly it was presented; there may have been an agon in
which Bellerophontes opposed his words to the text of the letter.127
We are on firmer ground with Euripides’ Palamedes: it is almost certain that a
letter played in it a role very similar to that of Phaedra’s suicide note, the ‘absent
person’ being this time its supposed sender Priam, who for obvious reasons could
not be consulted. There was a further twist, however, for the letter in question,
although seemingly coming from Priam, had been in fact written by Odysseus.
Thus the lie involved not only the contents, but also the provenance of the letter.
Here too, the written document, central to the tragic action, carried conviction
against the words of Palamedes.128

125
Radt, ad loc. makes no comments on the date, but cites Sellner’s dissertation (Jena 1910) for the
idea that the Stheneboea of Euripides stimulated Sophocles to write the Iobates. No argument survives,
and the three very short fragments we possess do not allow us to reconstruct the plot in any details.
126
See Berger–Doer 1990, nos. 1 and 3, both by the Darius Painter, both c.350 bc; 4, a loutrophoros
in a private collection, Baltimore painter, c.320 bc; 6, bell-crater, LCS 415, 360, Iobates sitting on a
diphros, in ornate (theatrical) costume, and reading the letter; behind him, his daughter Philonoe, who
looks at Bellerophontes; 7, lekythos from a private collection in Naples, LCS 302, 357, Laghetto Painter,
c.340 bc: Iobates receives the letter. 8, lekythos in Naples, Mus. Naz. 147868, LCS 334, 781, Aversa
painter, c.340 bc: Iobates with sceptre, oriental dress, and tiara reads the open diptych, held in front of
him by Bellerophontes.
127
For the continued life of the ‘Bellerophontes’ letter’, with explicit reference to Bellerophontes, see
Plaut. Bacch. 810–11; Plin. NH 13. 27. 88, doubting the statement of Mucianus, consul in ad 52, 70, and
75, that he had seen in a temple in Lycia a letter on papyrus, a charta, written by Sarpedon while he was
at Troy: for, as Pliny remarks, why would the others have used folding tablets such as those given to
Bellerophontes in Homer, if chartae had been available. See further Monaco 1965.
128
Discussion of the Palamedes above, 72–88.
224 Letter Writing and the Polis

Among the other surviving plays of Euripides two more feature letters, the
Iphigenia among the Taurians, and the Iphigenia in Aulis. In these plays, new
aspects of epistolary writing are explored.
The Iphigenia among the Taurians has been defined an ‘escape-play’, a tragedy
with a happy ending. Notwithstanding the initial focus on cruelty, human sacri-
fices, and remoteness from the Greek world, there is no overwhelming sense of
anguish, but instead a slightly oneiric, unreal dimension.129 And yet, even in a
‘light’ play, there are depths and complexity.130 It has often been remarked that
the play presents a very tight structure, with the various scenes leading one into
the other in a remarkably coherent way. This strong structure serves as a foil for
an exploration of the instability of reality, and more specifically, of the unstable
nature of the various individuals’ identity.131 This effect is achieved in various
ways: through the temporary inversion of the previous pattern in a person’s life
(this is Iphigenia’s case, for whom the letter is, in her new life, the tangible
memory of her former identity);132 through a split between name and reality;133
through the presentation of pairs and in general a high frequency of words for
‘double’;134 through the implications in terms of credibility arising out of the
rewriting of past stories;135 and through metadramatic references to rituals, for
these strengthen the impression of a gentle playing with tradition.

129
Ferrari 1988: 5–27. Discussion of definitions in Wright 2005: 6–55. Cropp 2000: 31–43 stresses
the element of tragic suffering, but also points to the changes in mood, the mock-combats at beginning
and end, the misunderstandings, and the folktale elements. He also notes the metadramatic references
to aetiologies (such as the Choes, 940–80), to tragic conventions, and to previous versions of the story
(especially Aeschylus’ Oresteia). All of the above combine to colour the perception of the story.
130
Well summarized in Wright 2005: 226–337 (‘Tragedy of ideas’), although his conclusion (that
escape-tragedies are among Euripides’ most pessimistic tragedies) seems to me excessive; see Ferrari
1988: 24.
131
Ferrari 1988: 5–34.
132
The dichotomies include the following. Iphigenia was sacrificed in Aulis / she presides over
sacrifices among the Taurians (a contrast made explicit at 336–9 by a herdsman and by Iphigenia
herself at 354–60.) She thought she would marry Achilles (25–7) / the name of the king of the Taurians,
Thoas, mirrors the fact that he is famous for his swift feet ( y ªB I Ø Ææ æ ØØ  æÆæ  | ¨Æ,
n TŒf  Æ ØŁd Y  æ E | K  h  qºŁ    øŒÆ å æØ, 31–3), just as the Homeric
Achilles was swift of feet,  Æ TŒ. She is utterly disconnected from the real world (cf. 220–30, and
esp. 220: ¼ªÆ  ¼Œ  ¼ ºØ ¼çغ ), yet she has moments of clear-sightedness (380–91). When
Orestes regrets that his sister will not be able to perform the funerary rites over his body (631),
Iphigenia pities him for his vain wish: ‘she dwells far off from this barbarian land’; but she declares
herself ready to accomplish that office instead, describing in detail what will happen to Orestes’ body,
down to the pouring of honey on his cinders (635).
133
See Ferrari 1988: 19–20, who highlights the opposition onoma–soma (a variant of the sophistic
antithesis onoma–pragma), explicit at 504; reappearing, in modified shape, at 765: e HÆ Æ  f
ºª ı Ø K ; Wright 2005: 293–4 (and below). The game of mirrors had begun early: Orestes
sees Erinyes in the cattle and slaughters them (296–300, with Wright 2005: 287 for the double layer of
illusion). See also the substitution for Orestes’ name of a ‘real’ one, ˜ııå, 500 (but compare 697);
Orestes lives P Æ F ŒÆd ÆÆå F, 568.
134
ıå Ø ÆÆØ is used three times for Orestes and Pylades (242, 474, 1289),  ı Ø once (456),
Ø  also once (264). Cf. also the Içº ªÆ  ıÆ (655), double arguments of the chorus, uncertain
whether to cry for the stranger that will leave or for the one who will be sacrificed. Wright 2005: 285–8
stresses the high number of terms relating to perception and appearance in the IT (forms of dokeo,
‘seem’, appear twenty-five times).
135
See Wright 2005: 283–5.
 The Athenian Dramatic Stage 225

The letter is central to the plot, as the instrument that engineers the recognition
of Orestes and Iphigenia: it occupies centre stage, both in the references made to it
by the actors and as an object brought physically on the stage and handed round,
from v. 582, when Iphigenia first mentions it, to v. 793, when Orestes discards it as
unimportant. Paradoxically this letter, viewed by Iphigenia as the instrument for
communicating at distance with her kin in Argos, will actually make possible
direct, face-to-face communication (above Pylades’ head, as it were) with Orestes,
who has been captured with his friend by the Taurians.
Iphigenia first mentions a letter when she learns the Argive origin of the
prisoners. She can spare one of them, and so proposes to Orestes (who has refused
to give his name) to save his life if he is willing to bring news of her to Argos:
‘If I spared your life, would you be willing go to Argos and bring news of me to my
friends there, and deliver a tablet, which a captive wrote having taken pity on me, since
he did not think that mine was the murderous hand, but that he was dying under the
law, considering these rites of the goddess legitimate. For I had no one who might bear
a message back to Argos and, on being saved, would convey my letter to one of my
friends. But you—for as it seems, you are not hostile to me, and know Mycenae and
those I love—save yourself there, and have this reward, not a shameful one: safety for
the sake of a light piece of writing (Œ çø . . . ªæÆ ø).’136
Before the dramatic action began, Iphigenia had thus asked someone to write a
letter for her, even though she did not have anyone who could deliver her message:
a letter singularly devoid of purpose, more a token than anything else. Yet the
story of how it was written gives pause: the letter is written in blood, as it were.
This is all the more striking as there was no need to introduce a former victim as
the writer: Iphigenia could have given Pylades a letter written by herself; nothing
in the play requires her to be illiterate. This is thus in a pregnant sense a letter ‘for’
Iphigenia, and it remains such until Iphigenia finds a carrier. Finally, at least as
remarkable is the way in which the letter is defined in the sentence that closes
Iphigenia’s entreaty: ‘light’, or also ‘vain’, letters, Œ FçÆ ªæ ÆÆ. Of course a
deltos is not heavy, and the grammata written on it, in their ‘incorporeal’ shape,
are even lighter; they will prove vain as well, in the course of the play (the near
homophony of Œ Fç  to Œøç, ‘deaf, mute, dumb’, said of writing and speech,
might have played a role in the choice of the term).
Once it has been decided that Pylades and not Orestes will carry the letter to
Argos, Iphigenia enters the temple to fetch her tablet. As she comes back with it, a
new fear has crossed her mind: that the person who will bear the letter to Argos
may not think much more about her request once he has left the land of the
Taurians (731–3). A letter is as safe as its carrier’s honesty: an oath is necessary. In
the mirror game of the play, Orestes then asks that she too takes an oath, that she
will let the messenger leave safely the land of the Taurians: a clearly absurd

136
IT 582–94: Łº Ø ¼, N ÆØ  , IªªEºÆ   Ø | æe @æª  KºŁg  E K E KŒE çº Ø, |
º   KªŒE, X Ø NŒæÆ Kb | 585 ªæÆł ÆNå ºø , Påd c Kc | ç Æ  Çø åEæÆ,
 F  ı o | ŁfiŒØ, a B Ł F   ŒÆØ ª  ; | P Æ ªaæ r å  ‹Ø IªªºÆØ  ºg |
K @æª  ÆsŁØ,   < > Ka KØ ºa | 590 łØ øŁd H KH çºø Ø. | f —r ª æ, ‰
 ØŒÆ, Påd ıc | ŒÆd a #ıŒÆ rŁÆ å G Kªg çغH— | ŁÅØ ŒE, ØŁe PŒ ÆNåæe
ºÆ, Œ çø ŒÆØ ªæÆ ø øÅæÆ. On the extremely sophisticated use of º , ªæ ÆÆ,
and KØ º, depending on what Iphigenia wants to stress, see above, 17; Kyriakou 2006: 200.
226 Letter Writing and the Polis

request, for as Iphigenia mildly remarks, ‘This is only fair: how could he deliver
the message otherwise?’ (740). At any rate: both Pylades and Iphigenia swear
never to reach home if they break their oaths, but a further worry this time crosses
Pylades’ mind:
K Ææ  Ø e  , X Ø ÆF  ŁfiÅ, | åM º  K Œº øØ åæÅ ø Æ | IçÆc
ªÅÆØ, HÆ KŒø  , | e ‹æŒ  r ÆØ   ÅŒ  . (IT 755–8)
‘Give me this exception: if something happens to the ship and the tablet disappears
with the goods in the waves, and I rescue only my own body, let this oath not be
binding.’
Letters and goods may disappear by accident, even while the messenger sur-
vives. This is where Iphigenia discovers the advantages of double transmission,
and adopts the procedure that was common in Greek official communications:
she will tell the messenger the content of the document (and announces this with
marked redundancy):
Iºº r Ł n æ ø·  ººa ªaæ  ººH ŒıæE. | IÆ ŒIªªªæÆ K º ı
ıåÆE | ºªø
fi çæ ø  Ø   IƪªEºÆØ çº Ø. | K IçƺE ª æ· j b KŒfiÅ
ªæÆç, | ÆPc çæ Ø تHÆ IªªªæÆÆ: | j K ŁÆº fiÅ ªæ Æ IçÆØŁB fi
 , | e HÆ Æ  f ºª ı Ø K . (IT 759–65)
‘But I know what I will do; for many attempts hit many targets. All that is contained
and written in the folds of the tablet I will tell you in words, to report to my loved ones.
That way lies safety: if you preserve the writing, it will itself speak silently its message;
but if the document is lost in the sea, by saving your body you will save for me my
message.’
Through being written, words acquire a HÆ, a solidity and a corporeality
linked to the tablet, a stable sign;137 but this body/tablet can be lost (or preserved);
a human body may survive (or be sacrificed: in all this, Orestes is due to be killed,
his death being the counterpart to the safe delivery of the tablet); a letter may
silently communicate, even as a living being (Orestes) may decline to state his
name.138 To ensure that her message arrives, Iphigenia proceeds to ‘read’ the
letter, with Pylades as the main addressee, and Orestes ‘eavesdropping’ on it.
But she is interrupted immediately after her first words ‘Tell Orestes, son of
Agamemnon . . . ’ (¼ªªºº Oæfi Å, ÆØ d IªÆ   . . . , 769), in the first of
a series of interventions by Pylades and then by Orestes, who slowly begin to
realize who she is.139
After having given the name of the addressee, following epistolary procedure,
Iphigenia states the name of the sender in riddling terms that again point to her
strange situation (living but dead): ‘She who was sacrificed in Aulis sends you this,

137
Wright 2005: 336.
138
See further on the word-plays in this passage Cropp 2000, ad loc. and Wright 2005: 294, both
stressing the pair man/chremata (consequence of the earlier mention of the loss of everything, and of a
possible word-play chremata/grammata); Cropp, however, interprets this in the legalistic direction,
Wright underlines the unsettling effect of this passage and in general the complexity of the relationship
between reality and illusion.
139
I follow here the line order (very much in accordance with the light character of the scene)
proposed by Jackson in 1955 and accepted by Kovacs in his Loeb edition, inserting after 769 the vv.
780–781–777. Diggle and Cropp 2000 retain the traditional order, as does Kyriakou 2006: 259.
 The Athenian Dramatic Stage 227

Iphigenia, alive, although no longer alive for people there’ (  `Pº Ø çƪE
KغºØ   | ÇH "çتØÆ,  E KŒE P ÇH Ø, 770–1). An indetermin-
acy of another type applies to the next sentence of the letter, which she adds, after
one more interruption: ‘Bring me to Argos, o brother, before I/you die’ (˚ØÆ
 K @æª , t ÆØ, æd ŁÆE, 774): the infinitive ŁÆE leaves it open
whether what is meant is her own death in foreign lands (that must have been
the meaning she had in mind) or Orestes’. Of course Orestes has to be alive if he is
to save her, but the play has been building up until this moment towards his death
as a sacrificial victim at the hands of Iphigenia.140 The letter closes with an Iæ , a
malediction: ‘or I shall be a curse upon your house’ (778) that reinforces the effect
of realism conveyed by the entire letter (the names of both the addressee in the
dative and the sender as well as the verb KغºØ combine to make this a very
real-sounding letter; although it is unlikely that the formula valedicendi at the end
of a letter was a stable feature of fifth-century letters, it is still striking to find here
an inverted one). After some further details to the messenger that make clear how
much she is already, in her mind, in Argos (the letter does indeed here function as
one half of a conversation: the simple fact of repeating it makes Iphigenia feel that
she is already there), Iphigenia marks the end of her ‘reading’ and of her instruc-
tions with a final: ‘Those are my commands, this is what is written in the tablets’
(Æ¥ KØ ºÆ, |  Kd I º ØØ KªªªæÆÆ, 786–7). Upon this Pylades
delivers himself directly of his oath by passing the tablet straight to Orestes:
N , çæø  Ø º  I  ø , | OæÆ, B  B ŒÆتÅ  æÆ.
(Eur. IT 791–2)
‘See, I bring you a tablet and give it to you, Orestes, from your sister here.’
But something interesting happens at this point: Orestes accepts the letter, as
one accepts a presage ( å ÆØ, 793);141 but he then immediately discards it,
choosing action (he tries to embrace Iphigenia) over both folds of letters and
words (Ææd b ªæÆ ø ØÆıåa | c  c æH’ P ºª Ø ƃæ ÆØ,
793–4). Iphigenia hesitates, however: having kept this letter for so long, she now
wants to retain it and the image she had formed of her liberation, and cannot
accept the new situation without a test. Orestes must give a proof of his identity
(ŒæØ , 808), and he does so, dividing the proofs in two groups. First, what he
knows based on what he had heard from Electra (IΠ, 811). This part comprises
recalling the dispute between Atreus and Thyestes, a story that Iphigenia wove ‘in
a fine-textured web’—in other words, a story registered in another kind of
writing;142 the bridal bath she received before leaving for Aulis; and a lock of

140
Cropp 2000 translates ‘I’ and does not comment; Kovacs translates ‘you’; Kyriakou 2006: 260
notices, and states that ‘the subject is ’. Clearly the ambiguity here is intended. (In the same way, at
771  E KŒE, ‘the people there’, is also sometimes used euphemistically for the dead: Wright 2005:
268).
141
We have to imagine the corresponding gesturing of the two actors; the deictic B  makes it
easy. For å ÆØ in the sense of ‘to accept an omen’, see Kyriakou 2006: 264.
142
The very rare hÅ  at 814, in the question of Orestes (ÆF s çÆ rŁ K P Ø
çÆE;) is striking: it appears here; at IT 312, for the finely-woven peplos with which Pylades covers
Orestes after his attack of folly; and at the very end of the play, in the prophecy of Athena, at 1465:
Iphigenia will establish a cult in Halai, she will be buried there, and the finely-wove peploi of women who
have died in childbirth will be dedicated on her tomb ( y ŒÆd Ł łfi Å ŒÆŁÆ FÆ, ŒÆd ºø | ¼ªÆº
228 Letter Writing and the Polis

hair that Iphigenia sent back to her mother as a memorial for the tomb, instead of
her body (ÅE ª Id Æ   P F, 821). Iphigenia is here defined
through her weaving/writing, and through absences, the missed wedding cere-
mony and the empty tomb. Orestes moves now to the second type of proof, those
that he himself saw (L r  ÆP,   çæ ø ŒæØÆ, 822). This provides
the final, decisive proof, derived from something that Orestes saw in Iphigenia’s
own bedroom: the spear of Pelops with which he conquered Hippodamia. Only
now is Iphigenia convinced. The procedure may have been closely modelled on
the Athenian judicial system.143 But in the way these proofs are presented we can
also, with Segal, see something of Euripides’ self-awareness: Euripides’
tragic version of the story of the Atreidae finds its place between the new
technology of writing, represented by the letter, the old and traditional forms of
transmitting the story, symbolized by the weaving, and, I would add, the imme-
diacy of the direct, personal visual experience, as experienced specifically in
drama.144
Once the letter has served its purpose, it is forgotten, discarded as useless, just as
no more proofs will be needed to convince Iphigenia. At least for the modern
reader, one question lingers: Iphigenia repeated the content of the letter—but she
cannot have known what was actually written in it, as she had had to ask someone
else to write it for her; Orestes discarded the letter. For all the emphasis on this
document, we shall never know what exactly the prisoner who was going to be
sacrificed wrote; these are indeed Œ çÆ ªæ ÆÆ, ‘light letters’. In this sense, it
may not be an accident that Iphigenia proudly presents the letter, when she first
brings it physically onto the stage, with: ‘Here, foreigners, is the tablet of many
doors and many folds’ ( º ı b Æ¥   ºŁıæ Ø ØÆıåÆ, |  Ø,  æØØ,
727–8). The emphatic, elaborate presentation is understandable: as long as she
remains among the Taurians, this letter is Iphigenia’s link with her Argive royal
past.  ºŁıæ  is used only here for a tablet; it can be understood as modelled on
Łıæ , a term that could be used of the two leaves of a tablet. Yet the redundancy
of both ‘folds’ and ‘many leaves’ to characterize this tablet, added to the unique use
of  ºŁıæ  for a letter, cannot but surprise.145 Aristotle (Rhet. 1407b34) quoted
it as an instance of the lofty style. It is quite possible that these words have been
put in Iphigenia’s mouth to stress the contrast between the girl (illiterate: she had
to ask someone to write for her) and this instrument, which she sees as almost all-
powerful. Because of her illiteracy, this is truly an object of many doors
( ºŁıæ ).146 But the letter can be said to have many doors also in another,

 Ø Ł ıØ P ı ç ). In all of Greek literature hÅ  is attested again only in a citation of Eur.
IT 312 by [Luc.] Amores 47 (but Lucian’s codices actually have PŒ ı ç ) and in Aelian. NA 7. 57.
—ÅEØ is attested as an epithet of Athena, in Aelian (the same passage), and in Anth. Pal. 6. 289
(Leonidas, fittingly a dedication of a kerkis). Could this be another Euripidean aetiological allusion?
143
Ferrari 1988: 21; cf. the legalistic phrasing of other passages, as with K Ææ  at 755, with Cropp
2000, ad loc. Good, detailed discussion of the recognition in Wright 2005: 305–7.
144
Segal 1985: 222–3. The lack of overt reference to the theatre in Greek tragedy, as opposed to
comedy, is discussed in Hall 2006: 105–11; among the earlier bibliography, note Taplin 1986.
145
See Cropp 2000 ad loc.; Kyriakou 2006: 248–9; Jenkins 2006: 96–7.
146
Kyriakou 2006 has an interesting discussion of the shape of the tablet, but fails to see that her
final dismissive remark might reflect a point Euripides wanted to make (‘There is no reason to assume
with Bruhn and others that Iphigenia’s letter actually had many leaves. Apart from the fact that it
 The Athenian Dramatic Stage 229

metatheatrical sense. Firstly,  ºŁıæ  applies very well to the possibilities open
for reformulations of the mythical legacy, and to Euripides’ awareness of them: in
this play Euripides is giving new shape to traditional material (note the aetiologies
at the end of the play: local cults also undergo a similar reshaping). Secondly, as
Jenkins suggests, this tablet of many doors will give to all characters a way out of
the land of the Taurians.147 But the tablet of many doors also suggests, just at the
beginning of the recognition scene, the always present possibility of alternative
outcomes: the unread letter contains in its folds many paths, and the three main
characters, Iphigenia, Pylades, and Orestes, turn around it; but the letter, for all of
the insistence on writing in this part of the play, is out of reach or fundamentally
unstable. If this play (as a number of others) shows that ‘language is fundamen-
tally inadequate’, the inadequacy involves spoken words, but also written ones.148
Such an interpretation of the role of the letter is supported by the conclusion of
the play. Athena, appearing ex machina, is needed to save the situation; the goddess
first addresses Thoas and orders him to desist from his pursuit. Then, she addresses
Orestes and Iphigenia, who, however, are no longer on the scene: as the Taurian
messenger has just reported, they are all on board the Greek ship, but, because of the
waves, cannot leave the harbour. Athena thus sends epistolai to Orestes:
ÆŁg , OæÆ, a Ka KØ º — | ŒºØ ªaæ ÆP c ŒÆæ P Ææg
ŁA— . . . (IT 1446–7)
‘Having learnt, Orestes, my orders—for even not being present you hear the voice of
the goddess— . . . ’
One wonders whether the use of KØ º  here is not one of these Euripidean
features that made Aristophanes qualify him as ‘frigid’: a letter (or rather, the
paraphrase of the contents of a letter, containing Iphigenia’s cry for help, ad-
dressed to someone reputedly far away, but actually within hearing range) has
played a role earlier in the drama; the term is now used by Athena for her own
very different, authoritative, and direct (oral) message, sent, however, to someone
who is truly at some distance, Orestes.
In this instance too, the letter struck the imagination of vase-painters: a scene
representing Iphigenia handing the letter to Pylades is painted on an Attic calyx-
crater dated to c.390–380;149 six more Apulian vases, as well as a Campanian
amphora, are similarly decorated. The fact that these images do not closely follow
all the details of the Euripidean play is all the more interesting, because it shows
that in the story as it was perceived and transmitted the letter kept playing an

makes nonsense of Aristotle’s example [not really], many tablets for such a short letter would be the
equivalent of a modern letter written or printed on a single page and followed by several blank pages’,
249). There might have been some dramatic play with the object Iphigenia brought on the scene.
147
Jenkins 2006: 96–7. Note that at the very beginning of the play, Iphigenia had sent Orestes
libations (something more in her range), because, as a consequence of her dream, she feared that he
might have died (‰ çŁØø fi    Ø ø, 171). The parallelisms between the two sendings might
have been enhanced through action on the stage.
148
Wright 2005: 307 (the citation); 307–16 for the more general point.
149
Crater Ferrara Mus. Naz. T 1145, ARV2 1440, 1, Iphigenia Painter, cf. Kahil 1990, no. 19. At 717,
Kahil notes that this is the only Attic vase to present the story of Iphigenia among the Taurians: it is
thus all the more significant that exactly this moment was chosen.
230 Letter Writing and the Polis

important role, somehow characterizing Iphigenia (as we have seen, its practical
use is minimal).150 Aristotle too praised the first part of the recognition scene
(Poet. 1455a16–19) as the best of the many through which recognition can be
achieved, because it is a logical consequence of the situation: Iphigenia would have
naturally wanted to send a letter (he liked the second part of the recognition, Orestes’
proofs, much less, and catalogues it among the contrived ones,  ØÅÆØ, Poet.
1454b30–5).151 While this is important evidence for how letter writing was seen in
the fourth century, we should not unquestioningly impose on fifth-century percep-
tions the assumption that sending a letter is the natural thing to do.
Letters figure prominently also in the other Iphigenia play. We cannot know
how the original Iphigenia in Aulis began.152 In the version we have, a letter
opens the action. At the beginning, Agamemnon enters holding a tablet that he
wants to entrust to an old man. We learn more about the tablet and the
difficulties it is causing through the words of the old man, who pointing to it
( º   ªæ çØ |  , "` 35–6) wonders at Agamemnon’s activities: he is
writing by the light of a lamp, erases what he has written (ÆPa  ºØ ªæ ÆÆ
ıªåE, IA 37), seals the tablet and breaks its seal, throws the tablet on the
ground weeping . . . 153 Clearly this unusual writing activity foreshadows a prob-
lem, and we learn of the problem in the following rhesis by Agamemnon: the
army cannot sail because of unfavourable winds, and Calchas has prophesied
that only the sacrifice of Iphigenia will allow the Greeks to sail. The first reaction
of Agamemnon was (he recounts) to order that Talthybius, the herald, proclaim
the dismissal of the army, ‘with a high-pitched (or straight) proclamation’
(OæŁøfi ŒÅæªÆØ, 94). This is the honourable, ‘straight’, open reaction. How-
ever, persuaded by Menelaus, Agamemnon has instead written a letter, a folded
tablet, to his wife, asking her to send Iphigenia to marry Achilles. The wedding
is, of course, a lie: and Agamemnon’s choice of words, when he reveals that he
has deposited the writing in the folds of a tablet (ŒI º ı ıåÆE ªæ łÆ |
łÆ, IA 98–9), may be significant. There is certainly a gesture towards the
ªæ ÆÆ ºªæÆ of Iliadic memory, written on a folded tablet (ÆŒØ
ıŒH);154 the statement may also be intended to contrast with the directness
of the herald’s hypothetical proclamation.
Having explained the reason for his sorrow, Agamemnon goes back to the deltos
he has at hand, in which he has written his revised decision (ƪæ çø, 108),
and entrusts it to the old man.155 But, before letting the messenger go and bring

150
Kahil 1990, no. 10–24, and 25, dated to the period between 350 and 320 bc. For a further vase
not in LIMC see Taplin 2007: 152 and n. 99. The number acquires consistency when viewed against the
total number of vases depicting some part of the story.
151
Discussion in Wright 2005: 298. Sketch of the reception of the play in antiquity in Cropp 2000: 62–4.
152
For the textual problems that beset the Iphigenia in Aulis see Kovacs 2003. Jenkins 2006: 87
n. 13, Foley 1985: 102–5, and Knox 1972 accept the text as it stands. Good discussion of the play in
Foley 1985: 65–102.
153
For the idea that the confused way in which Agamemnon here proceeds foreshadows the
confusion this letter will cause see Jenkins 2006: 88–9.
154
Pinax is not attested in Euripides; the letter mentioned first (chronologically second) had been
defined by the old man simply as a deltos, the object of Agamemnon’s writing ( º   ªæ łÆ, 35).
155
ƪæ çø (and with it vv. 106–10) has been rejected by Page among others with the argument
that this verb occurs only here in poetry; it is defended by Knox 1972: 287 as ‘a good fifth century word:
 The Athenian Dramatic Stage 231

his orders to Argos (Iºº xÆ åæØ   KØ ºa ºÆg | æe @æª , 111–12),
he proposes to tell ‘in words’ to the old man, and thus to the audience, ‘what this tablet
hides in its folds, everything written therein (L b ŒŒıŁ º  K ıåÆE, | ºªø fi
çæ ø  Ø  Æ IªªªæÆÆ, 112–13)’. The procedure is exactly the same as
that followed in the Iphigenia among the Taurians, but here the reading is not really
necessary, from a theatrical point of view, since the audience by now already knows
the gist of the letter.156 The reaction of the old man,
ºª ŒÆd ÆØ’, ¥ Æ ŒÆd ªºfiÅ |  Æ  E  E ªæ ÆØ ÆP H (IA 117–18)
‘Tell me, explain, so that I may express with my tongue things in tune with your writing’,
goes some way towards explaining this decision; moreover, this corresponds to the
procedure commonly followed in interstate official communications.157 But the
letter is read also because of the letter scene in the earlier Iphigenia play. Clearly
this is an intertextual allusion to that other ‘reading’ scene, as shown by the
remarkable closeness between IA 112–13 and IT 760–1, both incidentally marked
by extreme redundancy, even as they introduce a redundant message (note also
the mimetic redundancy in the words of the old man at 117–18, above). That letter
had brought about recognition, and ultimately salvation, for both Iphigenia and
Orestes; this letter would have brought salvation to Iphigenia, if read to the right
person, namely the addressee, Clytaemestra. But it is read instead to the old man;
it will not be delivered—and Iphigenia will be sacrificed.
Agamemnon’s message to Clytaemestra is the only letter ever quoted in extant
drama (in the IT, Iphigenia had actually redictated her letter):
ø  Ø æe ÆE æŁ | º ı, t ¸ Æ æ , | c ººØ a a r Ø æe |
a Œ º Å æıª’ ¯P Æ | `sºØ IŒºÆ. | N ¼ººÆ uæÆ ªaæ c | ÆØ e
Æ  Æ ı. (Eur. IA 115–23)
‘I send you a letter in addition to my earlier one, daughter of Leda: do not send your
daughter to Aulis with its bays, protected from waves and jutting out toward Euboea. We
will make the wedding feast for our daughter’s marriage another time.’ (Trans. Kovacs)
The contrast with Iphigenia’s own letter is remarkable: there, plain language and
iambic trimeters; here, very elevated language and anapaests, a rather surprising
choice for the form of communication supposedly closest to everyday speech.158
There may be an intended contrast between the type of language a young person
such as Iphigenia had used, and that of a king such as Agamemnon;159 but in the

Thucydides uses it of the same circumstances and in the same sense—a change in the text of a letter
(1.132.5), and it is used by Xenophon of changing the text of treaties and by Demosthenes of changing
the text of juridical verdicts.’ See also Jenkins 2006: 90–1 (‘Metagrapho is the most important word of
the passage’, 90). At any rate, even if metagrapho were non-Euripidean (the passage is considered
spurious by both Diggle and Kovacs), it inserts itself perfectly into a series of terms prefixed with meta-
to indicate changes of mind (343, 346, 388, 500, 502; cf. Ferrari 1988: 34).
156
One of the main problems of introducing letters onto the stage is that their content must
somehow be made known to the audience; reading them is one way, cf. Rosenmeyer 2001: 69–70.
157
Cf. Knox 1972: 285–6; Jenkins 2006: 91–2.
158
Fraenkel 1955; cf. Knox 1972: 286.
159
Note that the contrast name/reality (here onoma /ergon), so present in the Iphigenia among the
Taurians, is brought out here by Agamemnon just after this letter, in answer to the question of the old
232 Letter Writing and the Polis

context of this play, the letter also underlines the overt contrast (and deep
parallelism) between Agamemnon’s use of a lofty tone in a letter, which fails to
convey any sense of urgency, and his inability to deal with important decisions
in face-to-face encounters. The feverish nightly writing of Agamemnon is his
attempt at rewriting history, at being a different person—with the arrival of
Iphigenia and Clytaemestra at the camp, Agamemnon’s attempt will be ended.160
One more element characterizes this letter: the fact that ostensibly it is sealed. The
seal is first mentioned by Agamemnon as the element that will give credibility to the
old man’s errand (note that in the words of the old man oral and written messages
are fused into one, predominantly oral, message: he speaks of his own credibility
‘when talking’, as a messenger bearing an oral message would, using the same verb,
çæ Çø, that Agamemnon had used for his recitation of the letter at v. 113):
(Old man): Øe b çæ Æ   H  ÆØ, | ºª, ÆØ d Ł B
fi B
fi  Iºåø
fi ; (A.):
çæƪE Æ çºÆ m Kd ºøfi |   Œ ÇØ. (IA 153–6)
Old Man: ‘But tell me, how shall I look trustworthy to your child and wife, when
announcing this?’ A.: ‘Keep unbroken the seal, which you carry on this letter.’
This letter, however, will not be delivered; its privacy will be violated, its seal
broken. During the song of the chorus, Menelaus (off stage) intercepts the letter
before it can reach the intended recipient, opens it, reads its contents, and
confronts Agamemnon with it. Thus, this letter signally fails in its purpose.
To compound Agamemnon’s plight, the first letter too in a sense fails: it reaches
the intended addressee, Clytaemestra; but, as Agamemnon will soon learn, the
queen does not follow the request precisely, and instead of simply sending
Iphigenia, she accompanies her daughter. Thus, of the two letters sent by Aga-
memnon, the deceitful one he sends first reaches its destination and attains its
goal, but does not command respect, or only in part, from its addressee; the
truthful one does not reach its destination, and renders his own situation danger-
ous, since his change of mind is objectified in a text whose authorship he cannot
deny. The inverted parallelism between these two letters was certainly intended:
the messages could, in both instances, have been conveyed orally. Letters appear in
this play as bad instruments for communication; if they are so marked, it is
because in this play letters function to emphasize Agamemnon’s inability to
engage in open, face-to-face communication.
This comes close to being stated explicitly in the course of the confrontation
between Agamemnon and Menelaus. The old man, having lost his fight over the
letter (303–13), alerts Agamemnon, calling him out of the tent with a sentence
showing clearly, by the use of KØ ºÆ rather than º  for the object stolen,
that it is Agamemnon’s orders that are endangered. But the first words of
Menelaus (who is holding in his hands the tablet throughout the scene) to
Agamemnon are ‘look at me’. Verbs of ‘seeing’, ‘opening’, and ‘showing’ are
repeatedly stressed in the following verses:

man ‘How will you deal with Achilles?’: Z ’, PŒ æª , Ææåø åغf | PŒ r  ª  ı, 128–9 (‘It
is his name, not himself that Achilles is lending: he does not know of the marriage’).
160
Foley 1985: 93–5 convincingly proposes to see here also an allusion to Euripides’ own rewriting
of the myth.
 The Athenian Dramatic Stage 233
#. ºł  N A, ¥ ’ Iæåa H ºªø ÆÆ º ø. | `ª. H æÆ PŒ
Iƌƺłø ºçÆæ , æø ªª; | #.  ’ ›æfi A º , ŒÆŒø ªæÆ ø
ÅæØ; | `ª. N æH· ŒÆd æHÆ ÆÅ H I ººÆ  åæH. | #. h, æd i
 ø ª ˜ÆÆ E AØ IªªªæÆÆ. | `ª. q ªaæ r Ł’ L   ŒÆØæe N ÆØ,
Ææ’ I; | #. u ’ IºªFÆ ª’, I  Æ L f Œ Œ’ Mæª ø º ŁæÆ
fi . (IA 320–6)
Me. ‘Look at me. I want this as the starting point of my words.’ A. ‘Shall I, a son of
Atreus, be unable to raise my glance from fear?’ Me. ‘Do you see this tablet, the servant
of the vilest message?’ A. ‘I see it. And first you must let it out of your grasp.’ Me. ‘No,
until I show its content to all the Greeks.’ A. ‘What? Do you know what you should not
know, having broken the seal?’ Me. ‘Yes, to your great chagrin, I have exposed the
mischief you were secretly doing.’ (Trans. Kovacs modified)
Menelaus contrasts his own straight, direct way of dealing with facts with
Agamemnon’s stealthy use of a letter, a ‘servant of the vilest writing’. Agamemnon
protests his right to manage his own house (e Ke NŒE r Œ  PŒ K  ÆØ;
331), that is, he continues to consider his letter as a personal document, as if his
decision did not have consequences for the entire army. But this is not accepted by
Menelaus. In the following rhesis, the latter proceeds to expose what he sees as two
damning changes in his brother’s behaviour, changes that parallel the move from
open speech to letter writing. According to Menelaus, at the time when he was
hoping to be chosen to lead the expedition of the Greeks against Troy Agamem-
non kept his doors unlocked for everyone (ŒÆd ŁæÆ åø IŒºfi  ı fiH Łº Ø
Å H, 340) and was ready to enter in conversation even with those that did not
want it, trying through this behaviour to gain advancement from the multitude
( E æ Ø ÇÅH æÆŁÆØ e çغØ  KŒ  ı, 342). As soon as he won the
office, he changed his manners (Œfi p , Kd ŒÆå Iæå , Æƺg ¼ºº ı
æ ı, 343) and became ‘hard to approach and scarce within doors’
( ıæØ  ø  ŒºfiŁæø  Ø , 345). It is difficult not to compare this
with other instances of tyrannical behaviour in which the use of letters for
communication goes together with the imposition of a distance and the refusal
of communication on the same level. According to Thucydides Pausanias made
himself difficult to approach, behaving so arrogantly that no one was able to come
near him ( ıæ   ÆPe ÆæEå . . . u Å Æ ÆŁÆØ æ ØÆØ,
Thuc. 1. 130. 2); in that same context Thucydides tells how Pausanias turned to
letter writing. The behaviour of Deioces may also be compared, who was open
before, but imposed communication through writing once he became king of the
Medes (Hdt. 1. 96 ff.). Much later Plutarch will state the same of Nicias
( ıæ  q ŒÆd ııŒ  . . . ŒÆÆŒŒºØ , Plut. Nic. 5. 1–2), who
too, fearing difficulties with the assembly, chose to innovate by sending a letter.161
Menelaus is not presented in the play as the kind of character in whom one
would want to put excessive trust (just like Agamemnon, he is unable to face
reality, and he likewise undergoes a radical change of mind, something fairly rare
in Greek tragedy);162 yet his description of the ‘other Agamemnon’ is very
compelling. He reinforces his point about the change in Agamemnon’s behaviour

161
Cf. Stockert 1992: 295–6, with further examples of this specific change in behaviour; and above,
88–9 for Deioces, 140 and 142–6 for Pausanias and Nicias. The emphasis on open and closed doors
here (IA 340, 345) may actually have been intended to parallel the folded and sealed faces of the tablet.
162
See Ferrari 1988; Foley 1985: 96; Knox 1966.
234 Letter Writing and the Polis

by pointing out a second change, with a formulation that lends weight to my

hypothesis of a connection between difficulty of access and epistolary writing.
Menelaus states that upon learning from Calchas that the sacrifice of Iphigenia
would appease Artemis and allow the Greeks to sail, Agamemnon freely decided
to write to his wife that she should send his daughter on the false premise of a
wedding to Achilles (360–2); ‘but then, after all, having turned round you have
been caught changing to other signs’ (Œfi pŁ’  æłÆ ººÅłÆØ Æƺg ¼ººÆ
ªæÆç , 363). The two verses, 343 and 363, are clearly worded so as to enhance the
parallelism: the same adverb is used to connect them to what precedes, Œfi pŁ’; the
same verb, in the same tense, Æƺ, is in both cases construed with accusa-
tive plurals: ¼ºº ı æ ı, ¼ººÆ ªæÆç .163
Agamemnon will resist the attack, and will restate (in ‘brief, clear, and easy to
understand’ words, 400) his refusal to sacrifice his daughter; but the arrival of
Clytaemestra and Iphigenia (reality overtaking writing) will seal the girl’s fate.
After admitting the failure of his tentative attempt to send his wife away (away
‘from my eyes’, K O ø Ææ I EºÆØ, 743), Agamemnon exits the
scene. At this point, the chorus sings a visionary song: the Achaeans will go to
Troy, where Cassandra—‘as I am told’ (¥ IŒ ø, 757–8)—tosses her blonde hair;
Ares, trying to fetch Helen, will sack the city utterly . . . Their song (vv. 751–800)
closes with the statement:
Øa , a ŒŒ ı ºØåÆå  ª , | N c ç Ø ı  u |  Œ ZæØŁØ
Æøfi <¸ Æ>, | ˜Øe ‹’ Mºº åŁÅ Æ, Y’ | K º Ø —ØæØ | FŁ Ø  ’ K
IŁæ ı | XªŒÆ Ææa ŒÆØæe ¼ººø. (IA 794–800, ed. Kovacs)
‘And all on account of you, child of the long-necked swan, if the tale is true that
<Leda> bore you to a winged bird, Zeus in altered shape—or it may be that in the
tablets of the Pierian Muses these tales are born to men falsely and to no purpose.’
(Trans. Kovacs)
Here the chorus wonders about the reliability of the traditional accounts of
Helen’s birth, and of the Trojan war (still to happen!) more generally. Fascinat-
ingly, when debating over the truth and falsehood of this traditional story, the
chorus locates the first in unattributed oral tales (ç Ø) and the second in
written poetic accounts (accounts of the Muses, themselves associated with
oral transmission and not with books!) Of course, both will be either true or
false; but the hint at the (re)writings of poets on the tablets of the Muses sounds
like a striking self-reference to Euripides’ own rewriting of the story in this play, as
well as a reference to Agamemnon’s epistolary attempts to cancel his writing
through a new writing.164
The stasimon is followed by Achilles’ encounter with Clytaemestra, a scene full
of misunderstandings that will be solved by the old man: the second message thus
arrives, but too late, and in an oral version. The two letters are mentioned in this
context for the last time ( º  fiTåÅ çæø  Ø æe a æd ªªæÆÆ, ‘I
had gone to bring you a letter concerning the earlier writing’, 891). But they are

163
See Stockert 1992: 295, and esp. 301. Note that Menelaus’ reference to Calchas’ prophecy at v.
358 is also introduced by Œfi pŁ’: the adverb marks the turning points of Agamemnon’s behaviour.
164
Zeitlin 1994: 170–1 suggests that this passage may echo in another key the epistolary theme of
the play; she proceeds to connect this with the importance taken by visualization.
 The Athenian Dramatic Stage 235

reported in absentia, since the first one is in Argos and the second one has been
intercepted by Menelaus. And yet, without letter and without seal, the old man has
no difficulty in convincing Clytaemestra of the truth of what he is saying.
What emerges from this survey of writing in Euripidean drama is the disap-
pearance of private (metaphorical) writing, such as the inscription on the tablets
of the mind. This is replaced by a much greater attention to public, ‘documentary’
writing, designed to establish a public and common memory; to books, seen as
markers of authority, but also potentially as carriers of corrupting influences; and
to epistolary communication, often—but not always—seen as problematic.

5.1.4. The Alphabet on Stage

In the last quarter of the fifth century, the nuts and bolts of writing in the form of
specific alphabetic letters make their appearance on the stage. We shall examine next
an ensemble of fragments, having in common the fact that they all come from two
passages of Athenaeus’ Deipnosophistae, and that in them alphabetic letters were
brought on the stage in a messenger’s report, and also visually, as images to be decoded.
In the tenth book of his Deipnosophistae Athenaeus preserves a sequence of
extracts from plays that may have had in fact rather different characteristics. He
found them collected in Clearchus’ book On Riddles (composed c.300 bc). The
overall picture may thus be skewed by the fact that these fragments had all come
from a collection of riddles, which also explains their relative homogeneity. The
first text is a comedy by Callias, the ‘so-called’ (as Larensis, the deipnosophist
speaking, says) ˆæÆÆØŒc ŁøæÆ (Spectacle of Letters)—elsewhere, Athenaeus
uses ˆæÆÆØŒc æƪøØ Æ, Lettered Tragedy: in both cases, this may have been
the name given by Clearchus to the part of Callias’ play he was using in his
book.165 Callias’ piece was (most likely) a comedy inspired by the discussions
surrounding the reform of the alphabet that took place in Athens in 403 bc under
the archonship of Euclides. In his prologue, Callias had introduced the letter
names, in their alphabetical sequence, in iambic metre ( . . . <¼ºçÆ>, BÆ, ª Æ,
ºÆ, Ł F ªaæ r , | ÇB’, qÆ, ŁB’, NHÆ, Œ Æ, º  Æ, F, | F, E, e s, E, ÞH,
e ªÆ, ÆF, <e> s, | Ææe çE åE  fiH łE N e t); there was also a chorus of
women, which sang (possibly to a melody from Euripides’ Medea) a lyric passage,
consisting of couples of alphabetic letters (stoicheia: , Æ, Æ, , , , and so on).166
In the course of the play, Callias (or rather the poets he brought on the stage)

165
Callias’ work is first mentioned in Athen. 7. 276 a; it is again mentioned in 10.453 c–454 a,
where passages of the play (see Callias test. *7 K.–A.) are actually quoted or summarized. Smith’s
interpretation (2003) of the relation between the play by Callias, the Medea, and the Oedipus Rex
supersedes all former discussions; Svenbro 1993: 183–6, Rosen 1999, Ruijgh 2001, and Slater 2002
should, however, also be mentioned here. Gagné 2013 is an important contribution but appeared too
late to be taken into account in what follows.
166
Athen. 10. 453d–e; cf. Athen. 7. 276 a. Chorus of women (not necessarily the letters of the
alphabet): Smith 2003: 325 n. 37 (against the interpretation of Svenbro 1993: 184, Wise 1998: 15, Ruijgh
2001: 262, Slater 2002: 126). The notion that Callias’ play provided the model for Euripides’ Medea is
clearly a misunderstanding, to be explained along the lines proposed by Smith 2003. For school
exercises in the form of syllabaries closely resembling the sequence in Callias’ prologue cf. IG II2
2784; see IG II2 2783 for a compendium of tachygraphy (?).
236 Letter Writing and the Polis

further analysed in terms of incorporation of letter names some iconic tragedies;

the example given in Athenaeus comes for instance from Sophocles’ Oedipus Rex.
Callias’ play thus may have been mainly a comic review of earlier plays, to be
imagined along the lines of Aristophanes’ Frogs; but there was a strong focus on
pronounciation, as is shown by a further quote concerning an exchange between
an actor and the chorus, focused on the correct enunciation of the seven vowels
(Athen. 10. 453f–454 a). On the whole, at issue were the sounds of the letters, not
their shape—and thus not writing: Smith has convincingly shown that these
passages were not meant as riddles, and that their riddling aspect resulted from
the transcription of a text made for oral delivery (the theatre script, which did not
present any riddles) into a text written for reading, Clearchus’ book.167 A further
fragment, however, presents a riddle of a different type, based on the shape of
letters and not on sounds: a woman claims to be pregnant with a baby whose
name corresponds to that of two letters whose shape she proceeds to describe: a
tall one with two arms reclining, departing from the middle; and a round letter
with two small feet, in other words, probably $ and %.168 Whatever its exact
meaning, this riddle shows that some letters were described in detail, and that the
visual, graphic description of the alphabet thus had a role in the play.
Next, Athenaeus lists Euripides’ Theseus. In this play, produced sometime
before 422 bc, an illiterate herdsman in what must have been a messenger-scene
described the shape of the letters composing the name of Theseus.169 Athenaeus
cites from the play a long series of thirteen iambic trimeters; the first two allow us
to know how the speaker characterized himself: ‘I do not know the letters, but
I shall tell the shapes and manifest signs’ (Kªg çıŒÆ ªæÆ ø b PŒ Y æØ, |
 æça b º ø ŒÆd ÆçB ŒæØÆ); the rest was given over to a description of the
individual shape of the various letters, mostly in terms of abstract lines (ªæÆÆ),
straight or crooked, although comparison with a curling lock of hair is used for
the . The context makes it impossible to know whether the name seen by the
herdsman was inscribed on a ship (the prow, or the sail), or on some personal
object (such as a sword or a shield), maybe marking a dedication; thus, the context
and the function of this ‘reading’ remain obscure.170 It is also unclear how the
illiterate herdsman himself was presented: his description made it possible for
at least part of the audience to recognize the name; yet it is unlikely that in a
tragedy (and moreover one concerned with Theseus), this scene was meant to

167
Smith 2003.
168
Athen. 10. 454 a. This riddle is not discussed by Smith 2003; for suggestions as to its meaning,
see Svenbro 1993: 185, who imagines that the graphism of a long letter and a rounded open one was
played out obscenely on the stage (but it is not certain that the chorus was indeed formed by
alphabetical letters); Slater 2002: 127 plumps for an abbreviation of łøº, ‘penis with retracted
foreskin’, and imagines that the pregnant woman who is speaking would have ‘flashed’ a huge penis
from under her dress; Rosen 1999: 156 reports Dalechamp’s (1583) interpretation, that two letters
mean łHÆ, ‘foetidus ventris crepitus’—an interpretation that fits comedy exceedingly well. L. Holford-
Strevens (per litteras) makes the excellent point that $ and % are two of the letters absent from the pre-
Euclidean alphabet; and that if we are indeed to think of łøº, it is interesting that ¸ had acquired a
new shape and H a new function.
169
Eur. fr. 382 Kannicht = Athen. 10. 454 b–c. The terminus ante for the date is given by a notice of a
scholion to Aristophanes’ Wasps, 313, asserting that these verses parodied Euripides’ Theseus. Cf. Jouan–
Van Looy 2000: 146–47; but the play could have been produced anytime between 455 and 422 bc.
170
For the first hypothesis, Rosenmeyer 2001: 345; Turner 1975: 9. Slater 2002: 119–21 considers
this unlikely, and canvasses the other two possibilities. None imposes itself.
 The Athenian Dramatic Stage 237

make city-dwellers have fun at the expense of humbler characters living in the
country.171 There is another possibility, besides that of fun at the expense of the
illiterate or that of a serious messenger-report by a humble but dignified character,
and that is to think of satyrs. But it is not certain whether the Theseus was a satyr
play or a tragedy; the latter seems more likely.172
The next instance quoted in Athenaeus is Agathon’s Telephus (or Tlepolemus),
probably a satyric play, in which, in a game of emulation with the previous piece,
an illiterate again described with a shorter series of six trimeters the letters
composing the name of Theseus; this time, the letters were not described
in abstract terms, but were compared to objects of similar shape, such as the
Scythian bow to describe the , or the straight rod (kanon), mentioned twice to
illustrate the H and the Y.173 This is the only fragment we have of the Telephus,
and Athenaeus does not give any other information; it is thus not possible to
say what the social status of the person describing the letters was, nor in which
connection Theseus was named, nor how important a place this ‘reading/
recognition scene’ occupied in the plot. The recognition of Theseus’ name is
also the focus of the next passage in Athenaeus, eight verses of the playwright
Theodectas of Phaselis, active in the first half of the fourth century. The
speaking character was, once again, an illiterate rustic (agroikos);174 and it is
very clear that Theodectas is indebted to both his predecessors. The title of the
play is not known, but it is likely that this one too was a satyr play.175
Callias’ comedy and these three plays must have followed one another in
Clearchus’ book. Callias’ comedy may have been a sort of literary review; what
went on in the other plays may be easier to understand if we assume that they were
satyr plays (with the exception of Euripides’ Theseus, at the beginning of the
series), because of the satyrs’ usefulness for exploring culture. Puzzle-solving is
almost a topos in a satyr play, and the instances just discussed fit well with it.
Satyrs tend to have an interesting relationship with everyday life: in their curiosity
and youthful enthusiasm, they pose afresh questions that men might have forgot-
ten to ask. The questioning of satyrs involves the relationship between reality,
image, and writing (so, for instance, in Aeschylus’ Theoroi), and it might have just
as well involved the shape and functioning of alphabetic characters.176 However,
this is not a necessary assumption; what must be admitted is an intertextual play
between the three passages: otherwise, not only is it difficult to explain why
Theseus’ name is the only one chosen for the ‘recognition’, but the recognition
itself becomes a rather boring affair once the spectators already know who the

171
Harris 1989: 109 n. 194; Slater 2002: 118. Torrance 2010: 245–6 attempts an interpretation in
terms of elitism and social tensions linked to literacy.
172
Sutton 1978 considers it probably satyric; Jouan and van Looy 2000: 151–2 prefer to see in it a
tragedy, as also, decidedly, Kannicht, TrGF ad l., and Slater 2002: 117–22; Cipolla 2003 hesitates, but
tends towards tragedy, also because of the lack of resolutions.
173
39 F 4 Snell = Athen. 10. 454 d. Satyric: Cipolla 2003: 300, because of the unusually high number
of resolutions when compared to the other fragments of Agathon (and when compared to Euripides’
Theseus, the ‘model’), and because of the theme; see Cipolla 2003: 299–302 for a commentary. Slater
2002: 122–4 considers it a tragic fragment, and gives a good overall discussion, with some interesting
hypotheses as to the content of the play.
174
72 F 6 Snell = Athen. 10. 454 de. See on the play Slater 2002: 124–5; Cipolla 2003: 311–12.
175
So Cipolla 2003: 311–12; the verses again present a very high number of resolutions.
176
Excellent discussion in Voelke 2001: 273–99.
238 Letter Writing and the Polis

subject is. If recognition had been the main point, different names would have
been chosen. With the assumption of intertextual play, the focus shifts from the
‘recognition’ process itself to the clever variants in the description.177 Conversely,
if no other names are mentioned in Athenaeus, probably this means that there
were no other instances of this kind of ‘lettered description’ of a hero’s name in the
course of a messenger-report.
The hypothesis that some or all the plays by Euripides, Agathon, and Theo-
dectas mentioned in connection with the recognition of Theseus’ name may have
been satyric is supported by the example with which Larensius (the Deipnosophist
speaking) rounds up this part of his dicussion: Sophocles’ satyric play Amphiar-
aus. According to Larensius, Sophocles in it ‘attempted to do the same’, i.e.,
presumably described the word ‘Theseus’, by bringing on stage someone who
danced the letters (a ªæ ÆÆ Ææ ªø Oæå  ).178 The passage of Athe-
naeus does not allow us to understand how such a riddle could have been
verbalized – it may have been simply mentioned in Clearchus’ book. What is
however evident is that here the letters were not described but performed.
Earlier in the tenth book, but already in the context of his discussion of riddles,
Athenaeus had mentioned two more works that brought writing onstage in yet a
different way: Antiphanes’ comedy Sappho and the satyr play Iris by the tragedio-
grapher Achaeus. At issue in the Sappho was not the shape of the letters but their
nature and place, as we shall see later. From the Iris Athenaeus quoted a passage
describing how ‘a litharge flask full of unguent was suspended alongside the
Spartan inscribed tablet, on a double peg’ (e ÆæØ Å ªæÆe K غfiH
ºø
fi | ŒæØ, 20 F 19 Snell), chosen in order to instantiate Achaeus’ ‘obscure and
enigmatic’ mode of writing. Athenaeus explains that what Achaeus meant here
was a white leather strap, on which a silvery oil flask hung; but to describe it, he
employed an expression twice removed from the real thing, using ‘Spartan inscribed
tablet’ for the Spartan skytale, and the latter metonymically for the leather strip
carrying the message that was wrapped around it.179 Alphabetic letters were thus
not brought onstage in the Iris, nor did this play necessarily include any riddles.180
But the preserved fragment is definitely interesting in its association of Iris (the
messenger of the gods), Athens (implied by the use of the word kyrbis, a particular
type of triangular support, on which Solon’s laws famously had been inscribed;

177
Slater 2002: 125 suggests that the name of ‘Theseus’ may have been chosen for these recognitions
because ‘tragedy taught the audience, sometimes quite literally’, and ‘an ability to recognize his
[Theseus’] name as a label on monuments might have been desirable’ (125). Drama taught, it is true,
but I find it difficult to believe that a playwright would go to such lengths for something not directly
functional to his plot. The first name recognition (the Euripidean one probably) must have been
necessary for some reason, and the others, whether they happened in satyr play or tragedy, simply built
on the previous one. See Turner 1975: 9 for a sensible evaluation of the meaning of these scenes for
literacy; and Torrance 2010: 246–8 for some good observations on the move from abstract lines and
geometric shapes (in Euripides) to material objects of everyday life (in Agathon) to a combination of
the two (in Theodectas).
178
Sophocles F 121 = Athen. 10. 454f.
179
20 F 19 Snell = Athen. 10. 451cd. There still remains the question of the double-faced wood or
peg, which is explained by Cipolla 2003: 206 in reference to the fact that two sticks were used for the
coding of the message. See Cipolla 2003: 204–8.
180
Although the entire passage is said to come from Clearchus, the silent insertion of details or even
entire discussions from another source is certainly possible.
 The Athenian Dramatic Stage 239

comedy makes frequent reference to real or metaphorical kyrbeis), and Sparta, a

polis that, if it was famous for its skytale, the messenger’s stick, was also well-known
for its refusal to puts its laws in writing.181 Iris in Attic iconography is often
represented as assaulted by satyrs; the skytale may have been a (rather peculiar)
attribute of the goddess in her role of messenger, and possibly the satyrs, having got
hold of it as a result of an assault, comically misused it.182
The second group of texts bringing letters onstage occurs in Athenaeus’ eleventh
book, in the context of a discussion of the ‘lettered cup’ (ªæÆÆØŒe ŒøÆ, i.e. a
cup bearing inscribed letters). Here, the source is not Clearchus (or at any rate: we
are not told so), and the focus not anymore on riddles. Athenaeus puts together
examples from comedy (an unknown play of Alexis, the Neottis of Eubulus), from
a satyr play (Achaeus’ Omphale), and from prose commentaries (in particular, a
piece of information deriving probably from Asclepiades of Myrlea and concerning
a cup said to have belonged to Nestor, conserved at Cuma and inscribed with
letters—possibly Homeric verses).183 Achaeus’ Omphale is the oldest text in this
series. In the passage of the Omphale, individual letters inscribed on a cup beckon
the satyr, who tries to piece together their meaning:
› b Œç    F Ł F ŒÆºE  ºÆØ | e ªæ Æ çÆø· º’, NHÆ ŒÆd æ  | s, F
 ’ s  æØ, Œ PŒ I ıÆ | KŒ  PŒØÆ a  ’ s ŒÅæ 
(Achaeus, fr. 33 Snell).
‘The skyphos of the god has been calling me for a long time, showing its inscription;
delta, iota, and in the third place, o; nu and u are present; and in what follows, it is not
an absence that san and ou proclaim.’
We may imagine the satyr turning the cup around in his hands, reading out the
letters one by one, until the final two ‘proclaim’ the name of the god and owner of
the cup to those, onstage and in the audience, able to put together the separate
letters/sounds into a meaningful whole. Was the point here the necessity of
connecting individual letters into a whole to produce sense, or rather the connec-
tion between naming (vocally or in writing) and essence, reality, epiphany
(‘performative’ writing)?184 At any rate, the kind of object described here was
fairly common: silver cups with gilt letters are well attested, as are ceramic ones
with painted dedications to Dionysus, Hygieia, and Zeus Soter, to whom at the
beginning of the symposium a libation was owed.185 Here, the name on the cup is

181
Plut. Lyc. 13. 1–3 states that Lycurgus prohibited the writing of laws. Apart from two inscrip-
tions, there is no epigraphical evidence for the use of written laws in Sparta before the second century
bc. Cf. Boring 1979: 27–31.
182
The little that is known of the Iris is discussed by Schloemann and Krumeich 1999, with references
to satyric assaults on Iris at 526–7, and to the role of the skytale with its secret messages at 529.
183
Alexis: Athen. 11. 466 de, and also 481f; Eubulus: Athen. 11. 467bc; Achaeus: Athen. 11. 466 ef,
cf. 498d; cup of Cuma, Athen. 11. 466 e, and also 489 c.
184
For Svenbro (1993: 177–8) the text plays on the contrast between the ‘backward’, uneducated
satyr, who reads aloud, and the speaking letters, which imply the practice of silent reading (by then
shared by most of the literate members of the audience). For a summary of the very little that is known
about the Omphale see Pechstein and Krumeich 1999; Cipolla 2003.
185
On comic cups see Arnott 1996: 761–2 (list of inscribed cups, of epigraphical references, and of
literary references to inscribed cups; note the one in Plaut. Rudens 478, inscribed with the name of
Venus); Wilkins 2000: 231–41; at 235, Wilkins points out that comedy shares with satyr play the
interest in the cottabus; the same, it would seem, applies to lettered cups. Arnott 1996: 760 stresses the
imitation of tragic rhythm in Alexis’ passage dealing with the lettered cup; of course this does not imply
in and of itself a connection with satyr drama.
240 Letter Writing and the Polis

in the genitive: the cup is the property of Dionysus. Again, we cannot know what
this meant in the plot; the audience may have enjoyed recognizing the name in
advance of the satyr—although the satyr already knows that this is the cup of the
god: we have here a ‘literate’ satyr, if still hesitant.186
In discussing the ‘lettered cup’, Athenaeus also cites two comic fragments. As
these fragments are likely to have been influenced by the earlier satyr plays, they
are best discussed here. In Alexis’ fragment (fr. 272 K.–A.), a character describes
to someone else a cup, old and small, crushed in the handles, and with letters
inscribed all around it (å  ŒŒºø fi  ªæ ÆÆ)– ‘eleven letters in gold’; to this the
other replies, ‘dedicating the object to Zeus Soter’. Clearly the cup was a treasured
antique, a silver cup with gilt letters, meant to prepare a recognition scene;187 but it
was not ‘read from’, nor were the shapes of the letters described.
The same applies to Eubulus’ fragment, the only one surviving of his play
Neottis.188 Here, a speaker seems to recognize the object: ‘I have always hated
more than anything a lettered cup. And yet how like this one is to the little phiale
my son took with him when he disappeared . . . ’ to which the second character (a
courtesan) answers philosophically: ‘So many things look alike’.189 A recognition
probably followed, but a staged deciphering scene seems unlikely. Some reading
may, however, have occurred in an earlier play of Aristophanes, for in a fragment
preserved in the Etymologicum magnum (fr. 634 K.–A.), a character shows (?) to
another character an inscribed cup: ªªøŒ e ¼ºØ  ŒÆd a ªæ ÆÆ,
‘recognize the cup and the letters’. How this functioned in each instance we can
only guess, but it is clear that there was much intertextual play going on; it is also
clear, however, that long scenes of letter recognition, especially when the letters
were inscribed on cups, were not central to the plots.190

5 . 2 . T H E A T R I C A L L E T TE R S : C O M E D Y

Although, as we have seen earlier, fifth-century comedy may have alluded to

official letters in connection with Cleon, and although writing in general figures

186
Svenbro 1988: 183–4; Voelke 2001: 208–9, 279.
187
See on the fragment (possibly from the play Hippeus, the Horseman) Arnott 1996: 760–2.
188
On the plot of the Neottis and on its similarities with Antiphanes’ and Anaxilas’ Neottis plays see
Hunter 1983: 159–60.
189
ØH Œ ŒØ  ªæÆÆØŒe Œø’ I· | Iaæ ‰ ‹ Ø  e ıƒe fiþå | åø çØ ºØ  fiH .
B:  ººa ªÆØ | ‹ ØÆ, Eubulus fr. 69 K.–A.
190
Lettered objects (or even letters) could also be used as recognition tokens: cf. Adesp. fr. 1084. 29–
33 K.–A., Plaut. Rudens 1156–65, and Hunter 1983: 160. D’Angour (1999: 113 and n. 29) suggests
interpreting Aristophanes fr. 71 K.–A. (Æø › B KØ· ‰  ºıªæ Æ ) from the Babylon-
ians, produced by the didaskalos Callistratus in 426 bc with reference to the 24 letters of the Ionian
alphabet, ‘prominently displayed on the masks of the twenty-four members of the chorus’: if the
didaskalos was a Samian (a Samian Callistratus is mentioned in connection with both the play in
Andron’s Tripod: Phot. 498.15 = Suda σ 77, who preserve the fragment, and the reform of the alphabet
in Ephorus FGrH 70 F 106: see Jacoby, ad loc.), this might be a homage of the playwright to his didaskalos.
The traditional view is that fr. 71 alluded to the punishment of branding or tattooing inflicted on the
Samians, and that the chorus was made of allies treated as slaves; but for well-founded criticism see
Norwood 1930, who offers a reconstruction in which the chorus of barbarian slaves were tattooed or
marked with letters. If his hypothesis that the slaves were followers of Dionysus is correct, we might have
here a further instance of the strong connection between theatre and writing (above, 183–6).
 The Athenian Dramatic Stage 241

rather conspicuously in the plays of the archaia, the mese, and the nea, letter
writing does not feature in a noticeable way in what remains of fifth-century
comedy.191 A passage of Cratinus (incertae fabulae fr. 316 K.–A.) mentioning an
KØ º should be interpreted as referring to an order rather than to a letter:
Zonaras, who is the source of the fragment, quotes it as proof of the fact that
KØ º was also used instead of K º.192 Another fragment of Cratinus (fr. 243
K.–A.), although in itself so short as to defy interpretation, might have referred to
a letter, if we can trust Harpocration’s comment. In discussing the KØŁ ı
 æ  (non-traditional festivals, or also those instituted by a vote), Harpocration
(E 97 Keaney) looks for other uses of KŁ  in prose, and finds that ‘Lysias in
his speech Against Thrasybulus [fr. 158 Carey] mentions KØŁ ı KØ º ,
meaning letters given to others in order to be carried; for it was common practice
to say “he gave a letter” instead of “he delivered” ’ (ºªØ ªaæ q NŁØ 
‘KŁÅŒ KØ º’ Id  F Ææ øŒ), as Demosthenes in the speech for
Chrysippus against Phormio.193 Cratinus in the Trophonius (˚æÆE  K
!æ çøø fi )’. Assuming that the lemma is complete, this means that Cratinus
used the same expression. However, interpreters agree that the last phrase was
followed by words of Cratinus that have been lost: the adversative-connective
points in this direction, and might moreover imply that the next quote referred to
another aspect of the use. But whatever the specific meaning Cratinus might have
attributed to KŁ  here (and any of ‘additional’, ‘fictitious’, or ‘entrusted for
conveyance’ would fit the bill in a comedy), if Harpocration was following some
logical train of thought, then it is likely that Cratinus in the Trophonius had
mentioned a letter (an oracular pronouncement?).
And, of course, a scene of letter writing figures in the Thesmophoriazusae, in
Aristophanes’ parody of Euripides’ Palamedes: Mnesilochus, caught by the
women, resorts to writing on votive tablets, in order to send a message to
Euripides:
çæ  s <i> ¼ªªº  | łÆØ K ÆP; r Kªg ŒÆd c æ  | KŒ  F
—ƺÆ ı: ‰ KŒE , a º Æ | Þłø ªæ çø. Iºº P  æØØ ƃ º ÆØ. |
Ł s ª Ø ¼  Ø º ÆØ Ł; <Ł;> |  i N Æ d Iª ºÆ Id H
ºÆH | ªæ çø ØÆææ ØØ; ºØ   º. | º  ª  Ø ŒÆd ÆFÆ ŒIŒE q º .
| (sings) t åEæ KÆd | KªåØæE åæB æªø
fi  æø
fi . | ¼ª c Ø Œø H º Ø, |
 ÆŁ ºÅ ›ºŒ f | ŒæıŒÆ KH åŁø: Y Ø |  ıd e ÞH  åŁÅæ: | åæØ
åæØ.  Æ ÆhºÆŒÆ; |  Œ Kª  Æ ŒÆŁ › f | ŒÆ fi ÆÆfi : Æåø åæ.
(Ar. Thesm. 768–84)

191
For a list of comic passages referring to ‘reading, writing, books, inscriptions, letters, and waxed
tablets’ see Wise 1998: 17 n. 6 (forty-one passages, in a very incomplete list: the Letter(s) plays of
Euthycles, Alexis, Macho, and Timocles go unmentioned, as do Menander and the adespota comica
discussed below, and Antiphanes’ Sappho).
192
Pace Rosenmeyer 2001: 64. Text above, 17 and n. 65. Since we do not have the context, we
cannot know what exactly the entole/epistole would have been in Cratinus; the comparison with
Aeschylus fr. 293 Radt, ¼Œ ı a Ka KØ º  (possibly, following a hypothesis of Hermann,
words of Helios addressed to Phaethon climbing the chariot) speaks for understanding epistole in
both passages as an oral injunction because of the evident lack of spatial distance between speaker and
addressee.
193
Dem. 34. 28: the speaker says: P b a KØ ºa I øŒÆ ÆP E, L wØ KŁŒÆ. The
example shows how freely Harpocration refers to the original.
242 Letter Writing and the Polis
‘Let me see, what messenger could I best send to him? Actually, I know a method, from
the Palamedes; like him, I will write on oar-blades and throw them out. But there are
no oar-blades here. Now where could I get oar-blades from, where . . . ? What if I were
to write upon these votive dedications instead of on oar-blades and then throw them
away? Much better! And they are even wooden, like those oars. (sings) Oh! my hands,
it is necessary to set at work on an inventive and safety-providing task. Come, sheets of
smooth tablets, receive the traces of my stylus, heralds of my sufferings. Oh! oh! this
R is troublesome! Go on, go on. What a furrow! Come, hasten in all directions, this
way, that way; speed is the word!’
Incidentally, here Aristophanes, besides parodying the Palamedes and fooling
with the topic of long-distance communication, also plays with the shape of a
specific letter by highlighting the rho: we may imagine that the relative has
difficulties with the round part of the letter (he is possibly writing åÆæØ, or
has reached the rho in ¯PæØ Å).194 But apart from these examples, in Aris-
tophanes and more generally in the poets of the archaia terms of writing refer
mainly to books, lawsuits (ªæÆçÆ), oracles, decrees, and the like. While these are
exceedingly frequent, so much so that it does not make sense to review them here,
there do not seem to be other traces of letters.195 Thus, in a sense, the picture
offered by the archaia is rather different from that of Euripidean tragedy, in which
letters played a (comparatively) major role. What is similar, however, is that the
very frequent references to other types of writing are made incidentally, as passing
comments, but are usually not central to the plot.196 Scrolls may, however, appear
as dramatic props, as in Knights and even more pointedly in Birds—a large part of
the scenes following the agon is actually built on the arrival of individuals who rely
on books and writing for their activities: an oracle seller (960–92), and a decree
seller (1035–57), who both actually handle writings and read from them. The
characterization of writing we find in Aristophanes is also interesting: as pointed
out by Slater, books and writing bring with themselves the empowerment of those
who possess them (that Aristophanes views this negatively is another matter).197
Epistolary writing seems, however, to have featured rather prominently in
the mese and in the nea: the attested titles show that some comedies took their
title from a letter, which consequently must have played an important role in the

194
See Austin and Olson 2004: 261. A kind of spelling game might also be going on, such as the one
present in Callias’ Lettered Tragedy: for if, as Austin and Olson point out, there is an echo between
åŁø and  åŁÅæ, then it is striking that the letter highlighted as problematic is a æ, i.e. exactly the
suffix that distinguishes the adjective from the noun (both from  åŁø).
195
Note however Adespota Comica fr. 456 K.–A., discussed below, 321 n. 65. The pinakis whose
letters a character asks another to read out, in Philyllius’ Poleis fr. 10 K.–A., is unlikely to be a letter (KŒ
A ØÆŒ  ØÆæø ‹ Ø ŒÆ ºª Ø | a ªæ ÆŁ , æı). A survey of literacy in old comedy is
offered by Slater 1996. References to books appear in comedy at approximately the same time as they do
in tragedy, with Eupolis fr. 327 K.–A. (before 412 bc, Eupolis’ death) and Aristomenes fr. 9 K.–A. (this one
from a play Sorcerers: in this context, a reference to a bibliopoles might have entailed a specific type of
writing, associated with oracles and magic).
196
Cratinus’ Pytine presented probably the old poet as writing: see Cratinus fr. 208 and 209 K.–A.,
with Slater 1996: 108–19. See also Slater 1996 for a discussion of the passage from mainly oral to written
scripts for comedy, a change he situates at the beginning of Aristophanes’ activity—a question that
need not detain us here.
197
Slater 1996: 101–5. The second old woman in Eccl. 1050, who cites the law (H ªæÆ ø
NæÅŒø), may have brandished a copy of the decree: Slater 1996: 101 n. 8. The evidence from Birds is
relevant, but also complex; cf. Dobrov 1993 for possible reflections of the tragic Tereus (and of literacy
issues) in Birds. Detailed, important discussion of the oracle-book in Birds, from the point of view of
writing and religion, in Henrichs 2003b: 216–22.
 The Athenian Dramatic Stage 243

action. Thus Euthycles composed a play @ø Ø j ¯Ø º (Profligates, or, The
Letter); Alexis (fr. 81 K.–A.) and Macho (fr. 2 K.–A.) both wrote an ¯Ø º,
while Timocles authored a play ¯Ø ºÆ (frr. 9 and 10 K.–A.). Unluckily, in
none of these instances do the fragments allow an understanding of the kind of
use made of letter writing in the play.198 Letters moreover feature in some plays
of Menander, as well as in a few unattributed fragments. Thus in fr. 238 K.–A. of
Menander’s Misogynes a ªæÆÆ Ø  is qualified by Łıæ , a term interpreted
in a similar context by both Hesychius and Pollux as meaning ıå , ‘folded in
two’—that is, probably, a letter, or at any rate a closed, confidential document.199
A letter (in the form, however, of a will) certainly played an important role in
Menander’s Sicyonians: through it the foster mother of Stratophanes, before
dying, informed him that she and her husband were not his real parents; through
the tokens given to him, the hero managed to find his real parents.200 In the
Adespota comica fr. 1084 K.–A., 30–1 ªæ ÆÆ (here probably a letter) found in a
jar probably played a role in the final anagnorisis; letters are mentioned also in frr.
Adesp. 1096 K.–A. (a ªæÆÆ Ø  appears three times, at ll. 25, 44, and 48,
without its being possible to understand what is going on because of the condition
of the papyrus) and 1139 K.–A. (again a ªæÆÆ Ø  at l. 4, while what follows,
although very fragmentary, might indicate that the letter was read on stage). And
in a fragment of the playwright Philemo, probably to be dated, because of the
mention of king Magas, to the second quarter of the third century, a character
claims that a letter has arrived from Magas; even though the other replies:
‘Impossible, he can’t write’, clearly that piece of writing (possibly a fake by some
ingenious slave) must have served some purpose in the plot.201
The important presence of letters on the comic stage, obvious even from the
scanty fragments of the nea, is confirmed by the fact that letters featured also in a
number of Latin plays. Of course this cannot be taken directly as evidence that
letters were important in Greek comedy: still, it is a fact that should be kept in
mind.202 Here too, some plays took their name from the use of a letter (and it is
only reasonable to suppose that the letter giving its title to a play will have had an
important role in it). Caecilius and Afranius wrote an Epistula (respectively a

198
Jenkins 2005: 362–3.
199
Hesych. 1775 Łıæ  ªæÆÆ Ø · ıå ; Pollux 10. 57: ªæÆÆ Ø  Łıæ  j
æıå  j ŒÆd ºØø ıåH. Compare the  ºŁıæ Ø ØÆıåÆ of Eur. IT 727–8 (above, 228).
200
Cf. Men. Sic. 130–41, and 141–4 for the tokens. The document is said to be a will at 248. Another
letter is mentioned by the main character Stratophanes at 136–7 (the word is ªæ ÆÆ), as having
brought to him notice of the death of his father at a time when he was in Caria. Detailed discussion in
Scafuro 2004: 2–5.
201
Philemo fr. 132 K.–A., from Plut. De cohib. ira 9, 458 a: A. Ææa  F Æغø ªæ ÆŁ’ lŒØ  Ø
# ªÆ. | B. # ªÆ ŒÆŒ ÆØ ; ªæ Æ’ PŒ KÆÆØ. ‘There is a writing for you from king Magas.’
B. ‘Magas, you rogue? He can’t write’.
202
Scafuro 2004 suggests that out of six letters read out on the Plautine stage, three were present in
the Greek model, while three would be Plautine inventions. Nadjo 2004, Scafuro 2004, and Jenkins
2005 identify specific Plautine aspects of the use of letters in the plays: their ‘tone’ and the practice of
interrupting their reading (Nadjo); their composition or reading aloud on stage (Scafuro); a metathea-
trical reflection on the ‘realities’ of speaking and writing (Jenkins). These Plautine elements may have
been present already in the nea. (Note Scafuro’s remark, 2004: 7 and n. 23, that letter writing ‘has left
only faint traces elsewhere in republican Rome’—but see, for instance, the fascinating fragments of a
letter attributed to Cornelia, the mother of the Gracchi, preserved in Cornelius Nepos fr. 59 Marshall =
Cugusi, Epistolographi Latini minores, 124. 3–4).
244 Letter Writing and the Polis

palliata and a togata), while Novius authored an atellana bearing the title of
Tabellaria; letters played a role in Turpilius’ Philopator.203 The characters in the
preserved plays of Plautus do not hesitate to make use of letters, for their love
affairs, in writing to their family or friends, or for business, so that about half of his
plays present references to letters, real or imaginary; in four of them (Pseudolus,
Bacchides, Curculio, and Persa) the text of a letter is actually read out.204
If the scant remains of the Greek comic plays do not allow the same type of analysis
of the presence of letters that is possible for tragedy or for the Latin comic theatre,
there are nonetheless some very informative passages. We shall look in detail at two, a
fragment of the comedy Sappho by the fourth-century playwright Antiphanes and a
fragment of the poet of the nea, Philemo (an older contemporary of Menander).

5.2.1. Antiphanes’ Sappho: Private Writing, and Public Speech

Antiphanes’ Sappho contains rather intriguing information relevant to our con-

cerns: it situates epistolary writing in relation to women on the one hand, and to
the polis with its rhetors on the other.
The plot of the play is not known (only two fragments survive); but it is fairly
certain that the main character was the poetess Sappho, and that at some point in
the play, the eponymous protagonist began proposing riddles, to which other
characters gave answers. In fragment 194 K.–A. we have one such riddle, to which
two solutions are given: the first one Sappho rejects—it is a masculine solution
that brings into prominence the activity of the rhetors in the polis; in contrast, the
solution she gives herself, the ‘good’ one, so to speak, shows the familiarity of the
poetess with writing, and, more specifically, epistolary writing.
This text thus allows us to look at the role played by epistolary writing in
comedy, and to propose a comparison with the situation in tragedy, a genre much
more studied from this point of view. Moreover, this fragment offers the oppor-
tunity of looking at the relationship between the oral speech of the rhetors and the
epistolary genre: for even if Sappho considers the first solution a wrong one, it
must have had some degree of plausibility or her interlocutor would not have
proposed it. Finally, the historical context is interesting: the Sappho is dated to the
years 370/350 bc, at a moment in which letter writing has become a relatively
common and widespread means of communication. Antiphanes’ play gives us the
possibility to see how epistolary writing positions itself in respect to other genres.
The fragment is preserved by Athenaeus in the tenth book of his Deipnosophis-
tae, dedicated to the discussion of riddles. Athenaeus probably did not read the
Sappho: the source for the passage seems to be Clearchus’ treatise æd ªæçø, On
Riddles. It is thus impossible to know whether the comments with which the
citation is introduced are simply inferences or if they rest on a sure knowledge of

203
Caecilius: fr. 31–4 Warmington = 34–6 Ribbeck3; Afranius: Daviault 1981: 169–70; Novius: fr. 86
Ribbeck3; Turpilius: x–xi and Philopator frr. 12–13 Rychlewska = 195–8 Ribbeck3. See further Monaco
1965 and Nadjo 2004.
204
Letters in Plautus: Scafuro 2004: 20–1 lists sixteen mentions of letters; ten of these are carried on
stage, and out of these, six are read aloud: Plaut. Bacch. 728–47, 996–1035; Curc. 426–36; Pers. 500–27;
Pseud. 41–73 and 998–1014. See Scafuro 2004; discussion in Nadjo 2004; Jenkins 2005.
 The Athenian Dramatic Stage 245

the plot of the play, and in particular, to ascertain whether in this play Sappho
proposed a series of riddles or this one only.205 The text is as follows:

K b Æç E › Øç Å ÆPc c  ØæØÆ In Sappho, Antiphanes represents the poet
æ  ºº ıÆ  ØE ªæç ı   e æ , propounding riddles in the following manner,
I ºı  ı Øe oø.  b ª æ çÅØ· while somebody solves them thus. Sappho says:
. Ø çØ ŁºØÆ æçÅ fiÇ ı’ e Œº Ø S. There is a feminine being that keeps its babes in
its bosom,
ÆB, ZÆ ’ ¼çøÆ  c ¥ ÅØ ªªøe safe; and even though voiceless they raise a
sonorous cry
ŒÆd Øa Ø  r Æ ŒÆd Mæ ı Øa  Å over the waves of the sea and across all the dry
land
x  KŁºØ ŁÅH,  E ’ P b Ææ FØ IŒ Ø for those they want among mortals; and even
those who are not present
 Ø, Œøçc ’ IŒ B ÆYŁÅØ å ıØ. may hear; their sense of hearing, however, is dull.
ÆF Ø Kغı çÅØ· Someone solves this by saying:
B.  b çØ ªaæ m ºªØ Kd ºØ, B. The being of which you speak, that is the polis,
æçÅ ’ K ÆB
fi ØÆæçØ  f Þ æÆ. the children which she nurtures, the rhetors.
y Ø ŒŒæƪ b a ØÆØÆ These by their bawling draw over here receipts
IŒ B Æ ŒÆd Ie ¨æfi ŒÅ ºÆÆ across the sea, from Asia and from Thrace.
CºŒ ıØ Fæ .  ø b ºÅ  And while they split the revenues and brawl
forever on,
ÆPH Œ ŁÅÆØ º Ø æ ıø ’ Id the people sits near them, and doesn’t see nor
› B  P b h’ IŒ ø hŁ’ ›æH. hear.
. a l k l H ªaæ ª Ø’ ¼, t  æ, S. [What nonsense!] How could then, father, a
rhetor
Þøæ ¼çø ; B. j ±ºfiH æd ÆæÆø. be voiceless? B. If three times convicted
of unconstitutional measures.
[ a l k l ]206 ŒÆd c IŒæØH fiTÅ [– – –] I thought I knew what you were talking
KªøŒÆØ e ÞÅŁ. Iººa c ºª. about. Now tell me yourself.
ØÆ  ØE c Æçg Øƺı Å e ªæEç  Then, Antiphanes represents Sappho as solving the
oø· riddle thus:
. ŁºØÆ  ı Kd çØ KØ º, S. The feminine being, then, is an epistle,
æçÅ ’ K ÆB fi æØçæØ a ªæ ÆÆ the babes in her, the letters it carries round;
¼çøÆ ’ ZÆ <ÆFÆ>  E ææø ºÆºE they, though voiceless, talk to whom they desire
x   ºŁ’· æ  ’ i åfi Å Ø ºÅ  when far away; yet if another happens to stand near
g IƪتŒ   PŒ IŒ ÆØ. when it is read, he will not hear.
(Antiphanes fr. 194 K.-A. = Athenaeus 10. 450 e -
451 b)

205
Eustathius, in commenting on the Æ ıŒ sent with Bellerophontes to Lycia (Hom. Il. 6.
168–70, see above), cites, through Athenaeus’ Epitome, the fragment of the Sappho, with very minor
differences in respect to Athenaeus (Eust. in Hom. Il. (6.169) 632. 10: 0EØ N  ŒÆd › ıŒe
Æ . . . c Ææa  E æ Ø ź E KØ º, N m ªæØç ıÆ  Æçg ªæÆł oø· . . . (note
the word-play ªæØç ıÆ / ªæÆł), ‘Moreover it should be said that the folded tablet . . . points to
what later is going to be the letter, concerning which in a riddle Sappho wrote as follows: . . . ’; but
according to the archbishop, Sappho herself would have been the author of the riddle!)
206
At l. 13, Jacobs proposed P b ºªØ (and Kaibel ºÅæE åø, rejected by Cobet as unseemly in
a daughter answering back to her father). Something of this kind must have stood here. L. Holford-
Strevens suggests (per litteras) that we assume at 15 (where we have exactly the same kind of lacuna,
and where no solution has been proposed, apart from Cobet’s own ÆÇØ åø), a quick change of
246 Letter Writing and the Polis

This text has been discussed more than once. Bernard Knox, in an important
article published in 1968, referred to it as one of the main, and earliest, sources for
the practice of silent reading in antiquity.207 A number of studies on Sappho and
on the implications of her gender for the reception of her work refer to Anti-
phanes’ fragment, usually concentrating on the second solution of the riddle.208
Jesper Svenbro has also made use of the second solution, the one proposed by
Sappho, to support an allegorical interpretation of Sappho fr. 31 V. (the famous
poem on jealousy, çÆÆØ  Ø ŒB  Y  Ł ØØ), in the context of a larger study
on the anthropology of reading in ancient Greece. In Svenbro’s allegorical
reading, the ‘I’ of the poem is Sappho herself; the feminine ‘you’ is the written
poem, while the third person designates the reader of the poem; Sappho’s jealousy
depends on the fact that she will have to separate herself from the poem, which is
thus to be understood as an allegory of writing. Antiphanes’ fragment, in which
Sappho defines the letter as a feminine being, carrying in her womb children who
are signs, can of course support such an interpretation.209 However, Svenbro’s
fascinating reading presupposes a culture of literacy which it is difficult to accept
for seventh- and sixth-century Lesbos. While the possibility of an anachronistic
reading of fr. 31 by Antiphanes cannot be excluded, Antiphanes may not have
thought of fr. 31 when composing this part of his play: for the reference to Sappho’s
fr. 31 explains only one answer to the riddle, the second one, but not the first. The
same tendency to focus on the riddle and on its second, ‘good’, solution, ignoring
altogether the first one, characterizes Rosenmeyer’s discussion.210
In contrast, those scholars who are interested in the relationship between the
text and fourth-century Athenian politics, in particular the issue of corruption,
show the reverse tendency of quoting only the first solution, omitting Sappho’s
own. Thus Wankel for instance focuses exclusively on the corruption of the
politicians and on the relationship between the fragment and the Athenian
political situation, not even mentioning the second solution.211 Both approaches,
the purely literary one, and the purely historical one, ignore one part of the text,
and cannot thus offer a really satisfying interpretation.
In a recent contribution, however, Richard Martin has pointed to the way
forward, by offering an interpretation of Antiphanes’ fragment in which he
carefully uncovers the system of opposition underlying it: masculinity, connected
with orality and publicity, versus femininity, connected with epistolary writing
and privacy.212 But he too is mainly interested in the female voice and in Sappho’s

interlocutor; Sappho might have again stated P b ºªØ (or, in both situations, ÆÇØ åø; Cobet’s
proposal too requires a change of interlocutor). See the apparatus of Kassel–Austin, ad loc.
207
Knox 1968: 421–35 (esp. 432–3); see also Gavrilov 1997: 56–73 (esp. 68).
208
See e.g. Williamson 1995: 14–16.
209
Svenbro 1993: 145–59 (and for Antiphanes’ fragment, 158–9). The reading ϝ Ø of Apoll. Dysc.
Pron. 1. 59 Schn. instead of  Ø in the first line would not affect Svenbro’s argument. Most 1995
presents a good discussion of the interpretations offered of Sappho in general and of fr. 31 in particular.
Traill 2005 has convincingly shown that Plautus, Miles 1216–83 is a spoof of fr. 31, probably mediated
by the unknown author of a Greek play º Çø (cf. Plaut. Mil. 8–9); however, in Plautus (and
presumably in the º Çø) the fragment is read literally.
210
Rosenmeyer 2001: 96–7.
211
Wankel 1991a: 34–6.
212
Martin 2001: 73–4. Olson 2007: 200–3 offers a line-by-line commentary on the passage, good on
details but otherwise rather flat. Yatromanolakis 2007: 300–5 looks at the riddle mainly for what it can
 The Athenian Dramatic Stage 247

answer, and finds it to be the ‘right’ one. In what follows, I will offer a detailed
interpretation of the fragment along the lines sketched by Martin; but rather than
looking for the right solution, I shall try to explore the meaning deriving from the
tension between the two answers. The first step must, however, be a reconstruc-
tion, inasmuch as one is possible, of the original literary context of the fragment:
the plot is not known, but it is still possible to define what ‘type’ of comedy the
Sappho was.

5.2.1.1. The Generic Context of the Sappho

According to the anonymous author of the De comoedia, Antiphanes would have
been, with Alexis, the most illustrious poet of Middle Comedy; his activity
spanned most of the fourth century bc, and he won thirteen victories in Athens,
five at the Dionysia and eight at the Lenaea.213 Our sources attribute to him the
composition of a remarkably high number of comedies (260 according to the
anonymous De comoedia, 280 or 365 according to the Suda). Whatever the exact
number, it is clear that not all of these plays can have been composed for the
Athenian stage. Antiphanes must have produced plays ‘on paper’ for itinerant
companies, or—maybe—even simply for reading. The titles of his plays (140 titles
have been transmitted) show a production typical of the authors of the mese:
mythological travesties are frequent; more than twenty plays took their title from
the personal names of characters (for instance hetaerae), or their origin (B Øø,
Œ , `NªØ Ø); more than thirty titles seem to have concerned ‘types’ (either
moral types, such as the #Ø Åæ , or professional ones, such as the
Zøªæ ç ). Interestingly, Antiphanes wrote also a !æØƪøØ (fr. 207 K.–A.),
in which scenic artists were brought to the stage; and he spoke of his art in a long
fragment from the play — ÅØ (fr. 189 K.–A. = Athen. 6. 222a). All this
combines to produce the image of a ‘literate’ and at times self-reflexive playwright.
What kind of play was the Sappho? The poetess from Lesbos was not a
character unknown to the comic stage: at least six comic authors, from the
archaia to the nea, wrote plays bearing the title Sappho, whose content, however,
it is impossible to reconstruct, with the exception of Diphilus’ play; to Sappho’s
biography may also be attached most of those plays (numbering seven) bearing
the titles of Phaon or Leucadia.214 It is of course not necessarily the case that all
the plays taking their name from Phaon or from the rocks of Leucas brought
onto the stage the story of Sappho and Phaon.215 But even if they concerned

tell about the reception of Sappho in comedy (see 308 for the clearest statement of his way of viewing
the processes whereby her character is invested with new meanings, here and in the other plays linked
to her).
213
De comoedia 3. 45, 10 Kost. See the testimonia in Kassel–Austin, PCG, ii, 312–13; according to
the Suda, Æ 2735, he was born in the 93rd Olympiad (between 408 and 404 bc), began his career as a
comic poet in c.388, and died aged 74, i.e. around 330 bc. See Nesselrath 1996: 781–2; Konstantakos
2000: 173–98; and Ceccarelli 2004.
214
Detailed discussion and comparison with other poets’ presence on the comic stage in Ceccarelli
2004: 15–16; Yatromanolakis 2007: 295–9. Traill 2005 offers a splendid discussion of the presence of
Sappho in Greek and Latin comedy, and in particular in Plautus’ Miles.
215
Arnott 1996b: 224–5 insists on severing any links to Sappho in the plots of Menander’s and
Turpilius’ Leucadia plays; on the same lines, but for Alexis, Arnott 1996a: 394–5 (he finds, however,
248 Letter Writing and the Polis

everyday people and their vicissitudes, they may have mirrored in their plot the
legendary story—at any rate, they did refer to it, as lines 11–16 of Menander’s
Leucadia show. And Sappho’s love stories were certainly the subject of Diphilus’
Sappho.216
In Antiphanes’ Sappho, however, the poetess was presented as proposing
riddles. This points to a relationship with another type of comedy, in which
riddles and riddling characters featured prominently. ‘Riddle comedies’ are well
represented in Antiphanes’ oeuvre: they comprise plays such as the #ÆÆ, the
—Ææ ØÆØ, and the —æºÅÆ. For all the difficulties that we may have in
imagining how the plays worked, yet the plots of Mnemata and Paroimiai must
have somehow revolved around proverbs and memorable sayings, while the
solution of riddles must have been at the centre of the Problema.
For this second tradition, too, there were important precedents, starting already
with Epicharmus’ çª and with Cratinus’ ˚º  ıºEÆØ.217 On the basis of the
few fragments of the Cleobulinae we have (frr. 92–101 K.–A.) and of what is
otherwise known of the activities of Cleobulina and of her father Cleobulus, it has
been suggested that the plot centred on a group of women acting as ‘Cleobulinai’
(or, even better, accompanying a Cleobulina present on the stage) and proposing
riddles.218 The fragments do not allow one to know, however, who was supposed
to find an answer to the riddles. The best solution still seems that advanced by
Zieliński, who, basing himself on a well-known folkloric tale, suggested that the
women submitted riddles to their lovers.219 Such a hypothesis has the advantage
of linking together the riddles, an agon, and amorous stories. This type of comedy
centred on riddles seems to have grown in importance: it is certainly well attested
among the plays of the mese.220 The relatively important place of riddles and

Wilamowitz’s suggestion that mythical incidents might have been transformed into stories of ordinary
people ‘appealing’). While the fragments do indeed reflect vicissitudes of everyday people, the name
Dorcium for the girl in Turpilius’ play is interesting, for the name Doricha can be restored in Sappho fr.
15 (a minimal restoration, for whose plausibility see now Yatromanolakis 2007: 330–2), and tradition,
as represented by Strabo 17.1.33 and Athenaeus 13. 596 cd (who in turn quotes Posidippus), portrayed
Doricha as the lover of Sappho’s brother. The ‘everyday people’ might have been modelled very closely
on the ‘legendary’ figures. A contextualization of the stories involving Sappho, Charaxus, and Doricha/
Rhodopis is in Yatromanolakis 2007: 312–37.
216
˜çغ  › ŒøøØ Ø  Øe  ÅŒ K Æç E æ ÆØ Æç F KæÆa æåº å  ŒÆd
 IÆŒÆ, Athen. 13. 599 d. Sappho had had other lovers: Hermesianax fr. 7. 47–56 Powell presents
as lovers of the poetess Alcaeus and Anacreon. These texts are all of the Hellenistic period, but the
legend certainly began earlier, as the famous Attic black figure vase of the first half of the fifth century
(München 2416) representing Alcaeus and Sappho shows (Williamson 1995, 6–7, with fig. 2). On the
early reception of Sappho see now Yatromanolakis 2007, esp. 63–164 for the iconography (and for this
vase 73–9, with fig. 3a), 293–312 for comedy, and 348–61 for Hermesianax and the tradition he reflects.
217
The Sphinx can be dated to the first half of the fifth century bc; as for the Cleobulinae, both 450
bc (Wilamowitz 1899) and the years 439–437 (Pieters 1946: 144–6) have been proposed.
218
According to Diog. Laert. 1. 89, the tyrant Cleobulus of Lindus composed c.3000 verses of song and
riddles; as for Cleobulina’s status, scholars are divided: some see in her an invention of Cratinus (so
Wilamowitz 1899: 219–22), and put the three riddles which go under her name among Cratinus’ fragments;
others (Bowie 1999: 576) think that the riddles may have been attributed to a poetess ‘Cleobulina’
independently of Cratinus (her fragments are in West, IEG, ii, 50–1). Further on this Ceccarelli 2004.
219
Zieliński 1931: 31 and n. 16.
220
Cf. Eubulus Sphingokarion fr. 107 K.–A, preserving four riddles one after the other; riddles may
have occupied a prominent position also in Alexis’ Aesop. For the links connecting Aesop, Solon,
 The Athenian Dramatic Stage 249

oracular pronouncements, traditionally composed in hexameters, in the plays of

this period is one of the reasons that explain the overall rather important propor-
tion of the hexameter kata stichon in these same plays.221
The privileged context for the presentation and solution of riddles in these plays
seems to have been the symposium;222 as a long passage of Antiphanes’ Problema
shows, hetaerae will have been the main proponents of such riddles. This was also
the case in Alexis’ Cleobulina, whose fr. 109 alludes to a courtesan, Sinope; and it
is explicitly stated in fr. 22 K.–A. of Anaxilas’ Neottis, a long passage in which a
man twice compares women to Sphinxes speaking through riddles.223 A further
aspect worth stressing is that very often one of the solutions to these riddles points
to actual Athenian politics, while a second answer addresses personal issues: the
dual solution enabled the playwright to poke fun at women (for instance), while
keeping up the tradition of attacks on politicians that remained, even in the fourth
century, such an important part of comedy.224
Thus, Antiphanes’ Sappho should be located at the intersection of two comic
traditions, one bringing onto the stage the amorous private life of the poetess, and
one which gave riddles a central place in the play.
Before we move on to the interpretation of Sappho’s riddle and its two
solutions, however, we must ask a question concerning its status in the context
of the play: should we consider it as one of a series of many riddles proposed by
the principal character, or can we attribute to it a special status, a role in the plot?
Barring the discovery of new fragments, it is a question to which no final answer
can be given; however, the only other fragment from the Sappho which we have
(fr. 195 K.–A.) might incline the balance in favour of the second possibility. It is, as
luck would have it, a very short fragment: commenting on the word ØºØ ªæ ç ,
‘copyist’, Pollux (7. 211) says that the term was used in Antiphanes’ Sappho. The
plot referred thus at least twice to writing—and the riddle concerning epistolary
writing might have played an important role in it.225 It certainly was not a
traditional riddle, but rather an invention of Antiphanes, for no other versions
of it appear in the later collections of riddles (Athenaeus knew it probably from

Socrates, and Cleobulina in a sympotic context, see Ceccarelli 2004, with further bibliography; Sappho
too can be seen as close to Aesop.
221
On the importance of riddles in the mese see already schol. Dion. Thr. 479, characterizing the
mese as ÆNتÆ Å; Meineke 1839: 273–8; Pretagostini 1987: 249–51; Arnott 1996: 293 (on Alexis’
Cleobulina); Hunter 1983: 200–7 (on Eubulus’ Sphingokarion).
222
Fr. 122 K.–A. of Antiphanes’ Knoithideus or Gastron is particularly interesting, because it not
only contextualizes riddles in the symposium, but exemplifies these sympotic riddles with an instance
(‹ Ø çæø Ø c çæØ, fr. 122, 4) that may allude to fr. 101 K.–A. of Cratinus’ Cleobulinai (çæ ØŒ ),
thus supporting the hypothesis of a common riddling culture on which the comic poets built their
variations. See also Martin 2001.
223
Anaxilas fr. 22 K.–A., 22–31. Cf. also Callias fr. 28 K.–A. ‘Sphinx from Megara’, an expression
which became proverbial to designate hetaerae. Callias’ (?) Spectacle of Letters also brought together
women, riddles, and literacy (above, 235–6).
224
Pointed out by Meineke 1839: 278 for the Sphingokarion, but equally valid for Antiphanes’
Sappho; see also Ohlert 1886: 100.
225
Differently Knox 1968: 431, for whom the ‘bad’ solution, with its reference to actual politics,
would have been the starting-point, around which a riddle not having any further connection with the
plot would have been built. But see Olson 2007: 202, who finds the ‘political’ answer rather badly
constructed: ‘it may be that the explanation is not offered seriously and that the protest in 15–16 is
therefore made tongue-in-cheek’.
250 Letter Writing and the Polis

Clearchus’ treatise On Riddles; but even for Athenaeus, the riddle remained linked
to the comedy from which it originated).226
Some other comic fragments dealing with the life of Sappho present allusions to
writing. In the Phaon of Plato Comicus, a masculine character announces his
decision to find some quiet corner to read through a غ ; when, at the request
of another character, the first one reads a few lines, we realize that this is a cookery
book, written, however, in oracular hexameters (could this have been a parody of
Sappho reading her own poems?).227 Similarly, a fragment ex incertis fabulis from
a palliata by Turpilius, preserved by Jerome, presents a reflection on letters that
might be linked to the Sappho:
Turpilius comicus tractans de vicissitudine litterarum sola, inquit, res est, quae homines
absentes praesentes facit.228 (Letter 8, 31, ed. Ilberg = fr. 1 inc. fab. Rychlewska, 213
Ribbeck3)
The comic author Turpilius, in discussing the reciprocity of the letters, says that they
are the only instrument that can make present those who are absent.
This fragment has been tentatively attributed to the Philopator, a play in which
a letter was lost and furtively taken by someone else (as can be seen from frr.
12–13); but an alternative attribution to the Leucadia, a play that might have been
connected to the Sappho story, has been suggested.229 While the idea that the
letter has the capacity of rendering present those who are distant is a fairly
widespread one, there are undeniable similarities between Turpilius’ text and
the second answer to the riddle proposed in Antiphanes’ Sappho. If this passage
indeed came from a play bringing Sappho onto the stage, it would support the
hypothesis that a letter (or writing) played a part in some at least of the Sappho
plays. As we have seen, this would not have been the only instance of an important
role being given to a letter in the peripeteia of a comic play.

5.2.1.2. The Riddle

So far, I have tried to sketch the main lines of the tradition within which
Antiphanes’ Sappho should be located. But it is time to address now the riddle
itself, keeping in mind that riddles have in Greece a special epistemological status.
Riddles are a mode of knowledge built around a structural aporia, a dead end: no

226
Cf. Schultz 1912: 41: ‘der Brief als weibliches Wesen, das Kindlein unter seinem Busen bewahrt,
[ist] mir in der übrigen Rätsel-Literatur nicht vorbekommen und wohl auch erst durch überaus
künstliches Anknüpfen an das zufällig feminine Wort KØ º überhaupt möglich.’ A riddle having
as its solution ‘the letter’ is in the Veglie autunnali divise in cento enigmi per veglia di Fosildo Mirtuzio,
Accademico Brillante (Venice, 1796), veglia terza, no. 47 : ‘Già più di me non v’ha chi giri il mondo, | o
morta, o viva sia, son io sostegno, | per mia virtute, a cui non v’ha secondo, | d’ogni città, castel, e d’ogni
regno. | Volo, cammino, ancor scarsa di pondo, | necessaria a qualunque sia d’ingegno; | or già parlo
lontana, ed or vicina, | senza lingua né tosca, né latina.’ (Its origin is unknown.)
227
Plato com. Phaon fr. 189 K.–A.: Kªg KŁ K B fi KæÅÆ fi |  ıd غŁE  º ÆØ e غ  |
æe KÆı. B. Ø , IØ ºH ,  F ; | `. غ  ı ŒÆØ Ø OłÆæıÆ. | B. K Ø  ÆPc
lØ C . `. ¼Œ ı . | ¼æ ÆØ KŒ  º E , ºÅø Kd Ł . One would expect hexameters in
didactic epic, as in Archestratus; but the adjective ŒÆØ, in connection with the name Philoxenos,
might have recalled the famous writer of dithyrambs: a play on poetry is going on.
228
This is the text of Rychlewska 1971, who restores a trochaic septenarius; Ribbeck 1898: 130 had
suggested: Sola res quae homines absentes <nobis> praesentes facit.
229
Pastorino 1955: 43–5.
 The Athenian Dramatic Stage 251

riddle has a simple, clear solution, and typically the solutions offered oscillate
between banality and absurdity.230 As a result, we should resist accepting Sappho’s
indication and considering the answer of the ‘father’ wrong: both answers are
valid and wrong at the same time. To get the full meaning of the riddle, we must
explore the tension enacted in the riddle between the speech of the rhetors and
epistolary writing.
The text of the riddle itself, couched in hexameters, harks back to the genre of
epic and oracular parody; but the main literary referents for these verses are on the
one hand Aristophanes and the comic tradition, on the other, tragedy.231
The first hexameter opens in the way typical of riddles, with Ø followed
by the subject, as in Aristophanes’ Knights, but also, for instance, in fr. 94 K.–A.
of Cratinus’ Cleobulinai and in Eubulus’ Sphingocarion (Riddling Carion).232
Eubulus is worth closer examination, for the question asked concerns a creature
that lacks a tongue but speaks (Ø ºÆºH ¼ªºø ), whose female shares the name
with the male, a treasurer of its own winds, hairy, but sometimes hairless, saying things
that make no sense to the sensible (I Æ ı EØ ºªø), and making one law/
music follow on the other (  KŒ  ı ºŒø). (Eubul. fr. 106. 1–3 K.–A.)
Here too, two solutions are given—which are in fact one: Callistratos (a rhetor,
the ‘wrong’ solution); a æøŒ ºÆºÅØŒ, ‘a chattering arsehole’.233 Riddles
concerning beings that speak and do not speak are numerous;234 those concerning
specifically letters seem to go back all the way to fifth-century drama.
Three tragedies of Euripides are among the most important literary referents
for Sappho’s riddle: the Iphigenia among the Taurians, the Palamedes, and the

230
Cf. Jedrkiewicz 1997: 40–1; Pucci 1996, and especially 15–105 on the aporetic and tragic
structure of riddles; Martin 2001. For the terminology, Schultz 1914: 88–9; see also Gärtner 2001:
754–8, with bibliography.
231
A good instance of how these traditions overlap is offered by v. 3, ŒÆd Øa Ø  r Æ ŒÆd
Mæ ı Øa  Å: the first part recalls Ar. Av. 250, Kd Ø  r Æ ŁÆº Å (the song of the hoopoe
calling all his winged friends, in a passage alluding to Alcman fr. 26, 2–3 P.), which in turn is the result
of the association (‘hyper-heroical’: Dunbar 1995: 221) of Ø  r Æ (Eur. IA 704) and ±ºıæe
r Æ ŁÆº Å (Hom. H. Dem. 14).
232
Ar. Equ. 1037: Ø ªı,  Ø b º Ł ƒæÆE K ŁÆØ; Ar. Equ. 1058–9: Ø —º  æe
—º Ø ; Cratinus, fr 94 K.–A.: Ø ¼Œø ŒÆd çFæÆ ÆÆ fi hæØåØ ºø fi . This is a typically Homeric
verse opening: cf. Hom. Od. 3. 293, 4. 844, 15. 393, 19. 353.
233
On Callistratus’ status as rhetor cf. Dem. 18. 219; Hyp. 3. 1. See on him the short remarks of
Olson 2007: 221. The wording of v. 3, I Æ ı EØ ºªø, plays on Eur. Phoen. 1506, ı ı ı
ıe º  ªø çتª, as well as Eur. IT 1092, P   ı EØ  , said by the chorus of the
song of the halcyon; but also, as L. Holford–Strevens points out to me, on Pindar, Ol. 2. 85, TŒÆ
ºÅ . . . çøAÆ ı EØ, ‘arrows that speak to those who can understand’.
234
Hunter 1983: 201 mentions, besides Antiphanes’ Sappho, Euripides, Palamedes fr. 578 Kannicht;
Simias, Anth. Pal. 7. 193. 4 (a grasshopper); Kaibel Epigr. 234 (a gravestone). The fr. 96 K.–A. of
Cratinus’ Cleobulinai might be added; see also the following votive epigram in elegiac distichs, ‘said to
be by Sappho’: ÆE , ¼çø  K EÆ  Kø, ÆY Ø æÅÆØ, | çøa IŒÆ Æ ŒÆŁÆ æe  H,
‘Children, although voiceless, I answer when anyone asks, having set down at my feet a voice that is
never weary’, followed (vv. 3–6) by the ‘text’, i.e. the inscription engraved on the base of the statue
talking here (Anth. Pal. 6. 269). All these texts (including Antiphanes fr. 194. 2 (ZÆ ¼çøÆ  c
¥ ÅØ) and 5 (Œøçc IŒ B ÆYŁÅØ)) are built on games of oxymora. Discussion of the games of
opposites as typical of riddles in Schultz 1914: 67 (‘Gegensatz-Symbolik’), who quotes Arist. Poet. 22,
1458a26 : ÆNªÆ  ªaæ N Æ ÆoÅ K· e ºª Æ  æå Æ I ÆÆ ı łÆØ.
252 Letter Writing and the Polis

Hippolytus. As we have seen, in all of these plays a letter was central to the plot;
the fragment from Antiphanes is in close intertextual relationship with them.
The only link with the Hippolytus is in the use of   for the sound emitted by the
‘babes’ (compare the silent screams of the tablet in Eur. Hipp. 877). In the
Iphigenia, the eponymous character, when explaining how she wanted news of
her plight to be brought to Orestes in Argos, had pointed to a tablet, adding
(v. 763) ÆPc çæ Ø تHÆ IªªªæÆÆ, ‘by itself it will silently communicate
what is written’. At issue was communication at a distance, as in Sappho’s
solution; but needless to say, in the Iphigenia the recognition was brought about
not by the tablet, but by Iphigenia’s own loud ‘re-creation’ of it—had the tablet
remained true to its essence, Orestes would have been sacrificed. As for the
Palamedes, the echoes between the fragment of Antiphanes and fr. 578 Kannicht
are fairly close. Not only does the same word ¼çøÆ appear in both texts
(although with a slightly different meaning),235 also the idea that writing allows
communication at distance is exactly the same and is put across in a similar way.
Riddles are another element connecting the character of Palamedes with the
Sappho of this fragment: a fragment of Anaxandrides’ comedy Gerontomania
attributed to Radamanthys and Palamedes the invention of geloia (jests), which
included also riddles.236 Finally, within the field of interpretations created by the
riddle, Sappho positions herself on the side of epistolary writing; and Palamedes,
as we have seen, is the epistolary hero par excellence.237
But if Sappho took on the role of a Palamedes, then her letter writing may have
been felt as potentially problematic: after all, Palamedes is the hero who dies
because of a false letter accusing him unjustly, a victim of rhetors making
improper use of written communication. Possibly a letter played a negative role
in the plot of the Sappho—but this must remain a pure hypothesis. What is certain
is that all the tragic associations of this fragment are negative indeed.
In the Palamedes, Odysseus was the clever politician who pushed Agamemnon
against Palamedes, and who forged the letter that was the hero’s undoing. Who
are the rhetors in Antiphanes’ fragment? Obviously, politicians who probably
also held some military command; but it is impossible to put specific names to
them.238 It is, however, possible to give more precision to the rhetors as a group.

235
Stobaeus, who has preserved the Palamedes fragment, has aphona kai phonounta, which fits with
the oxymoronic tradition outlined above (some interpreters, including Jouan and van Looy 2000,
Scodel 1980: 91, and Jenkins 2006: 20, accept the correction çøÆ: see above, 81 and n. 76).
236
Anaxandr. Gerontomania fr. 10 K.–A.: e I º  yæ ªº ØÆ ºªØ  & Æ ÆŁı ŒÆd
—ƺÆ Å, the intellectual contribution replacing thus the contribution in food. For the nexus of
ideas behind this see above, 68–9, 79–81.
237
Detienne 1989; Ceccarelli 2002: 16–20; Jenkins 2005: 15–36; see above, 72–81.
238
Edmonds (1959: 263 n. h) goes much further than the evidence allows when he notes, concern-
ing the revenues from Asia or Thrace: ‘this, combined with the “epistle” below, suggests March 341,
after Philip had written to complain of the Athenian commander in the Chersonese, Diopeithes, who
was plundering ships and levying blackmail in the N. Aegean in order to pay his troops, and had held
his envoy to ransom, action which Demosthenes defended against Aeschines, who was thought to be
acting as an agent of Philip; about the same time the Persian king was privily sending money to
Diopeithes and perhaps also—this is uncertain—to Demosthenes.’ I doubt that the reference to a letter
in the play, in the context in which it is made, may have been as topical as that. See rather on the date
Wankel 1991a: 36 (367 as terminus post quem), with the apparatus of Kassel and Austin. However,
Wankel’s further suggestion that the end of the 360s should be excluded because at this time there are a
 The Athenian Dramatic Stage 253

A rhetor is someone who is ‘active as speaker in the assembly’, someone who

frequently speaks on policies concerning the city, in front of the city.239 But at
Athens, and at this moment in time, the term has a more specific meaning. In the
fourth century, the rhetors are a specific policy-making group, different from
magistrates (who must give accounts, euthynai, at the end of their term of service)
and from idiotai, private citizens. In fact, already in the fifth century the rhetors
had been able to control decisions by organizing themselves in informal groups,
and had thus attracted rather negative connotations, in comedy, but also in other
genres.240 The term is relatively frequent in comedy; moreover, Crates had written
a play Rhetores, and a comedy bearing this title may be attributed also to
Nicostratus, a comic author of the mese.241
Numerous passages of Aristophanes show that most of the points raised by
Antiphanes concerning the rhetors are not ad hoc inventions, but are part of a
discourse about the city and those who are supposed to direct its policies,
conducted mainly by the comic poets. Thus in Acharnians the contrast between
generations is presented in the following terms:
ƒ ªæ  ƒ ƺÆØ d çŁÆ B
fi ºØ· | P ªaæ I ø KŒø z KÆıÆåÆ
| ªÅæ  Œ Ł ç H, Iººa Øa  å ,| ¥ Ø ªæ Æ ¼ æÆ Kƺ
K ªæÆç  | e ÆŒø KA ŒÆƪºAŁÆØ ÞÅæø, | P b ZÆ, Iººa Œøç f ŒÆd
Ææ ÅıºÅ ı, . . . På ›æH P b N c B ŒÅ c MºªÅ. (676–81 and 684)
‘We, the veterans, blame the city. For we are not tended by you in a manner worthy of
our deeds on the sea, but suffer shameful abuse: you allow that we, old men, be thrown
into trials, where young politicians make fun of us, who are now deaf nothings who
have lost their music . . . not discerning anything if not the darkness of our case.’
In the larger context of an opposition between generations, here and in what
follows (until 718) Aristophanes develops aspects such as the old men’s voices,
now similar to the sound of an aulos that has been played beyond the limit,
their deafness, and their incapacity to see, which will reappear in the answer given
by the ‘father’, that is, a relatively old man, in the Antiphanes fragment;242

number of lawsuits against generals, while in the play the demos is portrayed as not seeing and not
hearing, seems to me to take too seriously a comic quip. See Harvey 1985 for a survey of corruption and
its vocabulary in fourth-century Athens.
239
Dover 1993: 242.
240
On the rhetors as a group with specific aims and connotations, see Canfora 1991: 11–17; Ober
1989: 156–91; and the studies collected in Hansen 1989. For the opposition rhetores–idiotai, see
Rubinstein 1998. The first epigraphical instance of the term is IG i3, 46. 25; already in the fifth century
the debate in the assembly was dominated by a relatively small group of rhetors: Olson 2002: 80. Critias
fr. 88 B 22 D.–K. (æ  b åæÅe Içƺæ   ı· | e b ªaæ P d i ØÆæłÆØ   |
Þøæ ÆØ , e ¼ø  ŒÆd Œ ø | ºª Ø Ææ ø  ºº ŒØ ºıÆÆØ) is particularly interest-
ing, because here a negative view of the rhetors overlaps with a negative view of the (written) law.
241
Fifteen examples in Aristophanes, and seven in other comic poets. The only fragment from
Crates’ Rhetores, fr. 30 K.–A., does not allow any inferences as to the plot. Nicostratus, fr. 24 K.–A.:
K øæØŒ ıåæE ƺØÆæ Ø, if it concerns rhetors (as suggested by Kaibel: see Kassel and
Austin’s apparatus), betrays a negative connotation.
242
Olson 2002, ad loc., interprets Œøç  in Ar. Ach. 681 as ‘mute’; but if so taken, the term would
double with Ææ ÅıºÅ ı, which is why I prefer ‘deaf ’. The term appears with this meaning in
Antiphanes fr. 194 K.–A., 5 (and deafness is stressed at v. 12).
254 Letter Writing and the Polis

conversely, the young men hit with their rounded words (æ ªªº Ø  E
ÞÆØ, 685); they set up verbal traps (ŒÆ ºÅŁæ ƒa KH, 686), and all
the while produce an unending chatter (º º Ø, 705 and 716).
The idea that the (old) demos is being duped by the rhetors (in the specific case,
Cleon) is at the core of the plot of the Knights; it comes to the fore explicitly at
the end of the play (vv. 1349–53). As for the screams, they are one of the hallmarks
of the rhetor: one may compare Peace 635–7, a passage in which Hermes describes
how the rhetors, with their ‘forked’ screams (ŒŒæ ªÆØ), had chased away the
goddess.243 Finally, polis, demos, and rhetors are joined together in Aristophanes’
Plutus: at vv. 567–70 Penia points out that the rhetors in the cities (K ÆE ºØ
 f Þ æÆ) behave justly towards the people and the city itself (æd e B 
ŒÆd c ºØ) as long as they are at the beginning of their career and thus poor; as
soon as they have gained wealth at the commons’ expense (Ie H Œ ØH) they
become unjust, and start intriguing and attacking the people itself.244
The fragments of other comic poets go in the same direction: thus for instance a
character in an unidentified play of Plato Comicus laments the proliferation of
rhetors and compares them to the mythical hydra:
‘for if one son-of-a bitch ( Åæ) dies, two rhetors grow back; for we don’t have an
Iolaus in our city who’ll cauterize their heads. You’ve let someone bugger you; that’s
why you’ll be a rhetor!’ (fr. 202 K.–A., trans. Olson slightly modified)245
Two more terms can yield insights into the connotations of letter writing and
the speech of rhetors in Antiphanes’ fragment: the adjective ¼çø , and the verb
ºÆºE. ¼çø , ‘voiceless’, appears three times in the fragment: in the text of the
riddle (v. 2), in Sappho’s comments to the first solution (v. 13), and in the solution
supported by the poetess herself (v. 19). In the interpretation proposed by the
father, the adjective indicates atimia, a technical sense of the term (normally
implying deprivation of civic rights) documented in Demosthenes.246 ¼çø  is
also used by Demosthenes, although in a non-technical sense, to accuse Aeschines
of remaining always silent when the safety of the fatherland is at stake: it is thus a
term used in the context of public and political speech.247 Finally, ¼çø  appears

243
Œæ Çø denotes the croaking of ravens, but it is used figuratively for human beings, with negative
connotations (loud screams deprived of meaning); cf. Ar. Equ. 287; Vesp. 103 and 415; Ach. 33; Plut.
722; and Antiphanes fr. 194 K.–A., 8 (ŒŒæƪ).
244
See also, for a negative characterization of rhetors in Aristophanes, Equ. 324–5 (their impu-
dence); 880 (the Paphlagonian bans male prostitution, fearing that the prostitutes may also become
rhetors; the same joke already at 425–6); Plut. 30, lumping together ƒæıº Ø, Þ æ | ŒÆd ıŒ ç ÆØ
ŒÆd  Åæ , and Plut. 379 (corruption); Aristoph. Daitaleis fr. 205 K.–A. (in the context of a contrast of
generations).
245
Short commentary in Olson 2007: 204. See also Eupolis’ Demoi fr. 96, and fr. 102 K.–A., stressing
the superiority of Pericles to the average run of rhetores (the term is repeated twice). Derogatory allusions
to rhetors are present also in authors of the mese: cf. Alexis, fr. 478 K.–A.; Theophilos fr. 4 K.–A.
(someone asks that a dish of fish that has become cold be taken away, adding ‘I don’t partake of rhetors’,
ÞÅæø P ª ÆØ).
246
Dem. 51. 12 (On the Trierarchic Crown): the proponent of an illegal decree, at the third condemna-
tion, is punished with a partial atimia and prevented from advising; cf. Hyperides, C. Phil. 4. 11 (1. 24. 28–25.
12 Blass), and D. S. 18. 18. 2 (for Demades). Discussion of the legal aspect in Wankel 1991b.
247
Dem. On the Crown (18) 191 and 198. The contrast between loquacity, used to abuse the right,
and aphonia is present already in Ar. Nub. 1320: the chorus anticipates that a moment will come when
 The Athenian Dramatic Stage 255

twice in the speech of Alcidamas’ On Those Who Write Written Speeches, or, On
Sophists, at 15 and at 25, in a context in which oral speech is positively contrasted
with writing; both times, the term is used to stress the fact that the use of writing
implies as a consequence total silence, the loss of expressive means.
As for ºÆºE, the term used by ‘Sappho’ in her solution (the ‘right’ solution)
to describe the activity of the letters, its connotations are not always entirely
positive—or rather, they may not be negative as long as the context remains a
private one; in the public arena, ºÆºE is clearly negative.248 The amplest, most
sustained use of the term throughout a play is found in Aristophanes’ Frogs. The
first hint is in an initial reference by Heracles to the existence of youngsters
(ØæÆŒººØÆ) who might be even greater blabberers (ºÆºæÆ) than Euripides
(91). From here onwards, ºÆºE will be used to connote the speech of young men,
of women, of sophists, or of Euripides himself, in contrast with Aeschylus and his
manly type of tragedy. Thus, Dionysus defines Euripides’ characters as ƒ F
ºÆº F (917), and contrasts them with the long Aeschylean silences (interest-
ingly, however, the example of a silent character picked upon is feminine,
Niobe); slightly later (948ff.) Euripides affirms that he taught his characters
(and the list comprises the master, but also the mistress, the maiden, the slave,
and the brothel-keeper) not to be idle, but to speak; the type of speech he
taught them is defined with ºÆºE (954), and in this context the term has
markedly negative connotations. This is later picked upon by Aeschylus, who
accuses him to have taught ºÆºØa KØÅ FÆØ ŒÆd øıº  (‘to pursue babble
and small talk’, 1069); finally, the chorus will use ºÆºE to connote the teachings
of Socrates (1492): and in fact, already in the Clouds the students of the
sophists had been presented as practising small talk (ºÆºØa IŒBÆØ, Nub.
931, cf. Nub. 1052).
But the negative connotations of ºÆºE are not limited to Aristophanic
comedy. A fragment of Eupolis presents a strong contrast between ºÆºE and
ºªØ: Phaeax, whose rhetorical abilities apparently did not match those of
Alcibiades, was depicted as ‘a prince of talkers, but in speaking most incapable’,
ºÆºE ¼æØ , I ıÆÆ  ºªØ.249 The same contrast will be used, much
later, by Aelius Aristides to describe Pericles as uninterested in small talk, but the
best in speaking.250 And while it is true, as pointed out by Dover, that in the fourth
century ºÆºE loses some of its negative connotations, becoming the normal word
for ‘making conversation’, ‘talking’, nonetheless the term still denotes talk con-
ducted on a private level.251 As Antiphanes’ fragment 194 K.–A. clearly inscribes
itself in the tradition of abuse of politicians going back to the fifth century, I would

Strepsiades will wish for his son, who has now learnt to successfully defend villainous causes, to be
¼çø . But aphonos is also used, in a non-technical sense, for those persons who first make noise and
then (because of a bribe) refuse to speak, as e.g. in [Dem.] Against Aristogeiton (25) 47.
248
See Olson 2007: 167–8 (‘in the fifth century ºÆºE carries a strong sense of subjective disap-
proval’); and the discussion in Dover 1993: 22.
249
Eupolis fr. 116 K.–A., preserved by Plutarch, Alc. 13. 2 (trans. Perrin); cf. Gellius 1. 15. 12.
250
Ael. Arist. 3. 52 (310. 3 L.–B.): º º  b lŒØÆ, r ÆØ, ºªØ b ¼æØ  NŒø K Ç . This
is not a quote, but Aelius Aristides uses much material from comedy.
251
Dover 1993: 22; Olson 2007: 167–8 ‘in the fourth century it is often merely a colloquial
equivalent of ºªø’.
256 Letter Writing and the Polis

argue that it also shares the negative fifth-century connotations of ºÆºE as

‘(useless) chatter’, and possibly feminine chatter.252
So what can we say of the riddle and of its two solutions? One thing seems fairly
clear: Antiphanes reaches back into a long comic tradition, discussing the activity
of rhetors in the city.253 He joins this strand with another tradition, which, for us,
is attested at its clearest in the Palamedes of Euripides and which is concerned
with the respective roles of (epistolary) writing and speech in the polis. The result
is typical of comedy, a system where both answers to the riddle point to something
which is going wrong: on the one hand, the deafness of the demos in front of the
screams of the rhetors vying to appropriate riches; on the other, the silence of a
writing which is markedly private (people nearby cannot hear, nor can the letters
themselves hear others), feminine, and thus also external to the polis. The ‘father’
in the fragment gives the male, traditional answer, marked by orality and publi-
city; he identifies in the children of the riddle a bad line of children of the polis,
who deal with public affairs in an incorrect way, mixing public and private, and
becoming wealthy through their political activity.254 Sappho chooses a female
answer—corresponding to the (ideological) female and private connotations of
epistolary writing. But her children too are not unproblematic.
In the field built by the riddle, the polis and the rhetors are opposed to the letter
and the characters. The latter do pass on a message but, as stated in the riddle,
have a dull sense of hearing, that is, they are not able to hear answers, to build a
dialogue. The rhetors have a similar problem: they too are marked by a shift of
what should be public towards the private in their appropriation of the public
wealth, while the demos, who does not see or hear, remains outside a system that it
cannot control. The fracture in communication highlighted in the letter is present
also among the rhetors who, if caught three times committing paranomia (trans-
gressing the law), may indeed lose the right to speak and become aphonoi.
Just as the rhetors are here negatively marked, so also epistolary writing is not
something positive for the polis—or at any rate, for Athens. But it is important to
add that in the case of the rhetors of the fragment from the Sappho, we look at a

252
Note that the womanly talk of courtesans is put on a par with that of demegoroi (orators, rhetors)
in a fragment from an unidentified play of Diphilus (an author of the nea), in which a man complains
about the unfaithfulness of courtesans: ‹æŒ  ÆæÆ ÆPe ŒÆd ÅŪæ ı·| Œ æ  ÆPH OØ
æe ‹ ºÆºE, ‘A courtesan’s oath is just like a politician’s; they each swear to suit the person they are
talking to’ (Diphilus fr. 101 K.–A., trans. Olson).
253
That Antiphanes and the other poets of the mese discussed the activity of the rhetors (and made
fun of it) appears from a number of fragments, collected in Athen. 6. 223 d–224 b, that all play on a
passage of the speech On Halonnesus, singled out by Aeschines in his On the Crown, 83 as well.
Antiphanes (fr. 167 K.–A.) in particular referred explicitly to Demosthenes; cf. also Plut. Demosth. 9. 5.
254
The pregnant letter should be compared with Theognis’ description (39–40) of a pregnant polis
that might give birth to a tyrant: ˚æ, ŒØ ºØ l ,  ØŒÆ b c ŒfiÅ ¼ æÆ | PŁıBæÆ ŒÆŒB
oæØ  æÅ (see also 1081–2, where the first line is identical, but the second strengthens the
tyrannical connotations: æØ, åƺB ªÆ  Ø ). The text continues with a protest against
the lack of distinction between private and public fortunes ( NŒø Œæ ø ¥ ŒÆ ŒÆd Œæ  , 46;
Œæ Æ Å ø fi f ŒÆŒfiH Kæå Æ, 50). On the polis as sexed entity in comedy see Eupolis’ Poleis, with
Rosen 1997, and esp. 167–8 for the opposition metropolis–patris: Rosen shows that when they are
thinking of their civic identity, the Athenians tend to conceptualize their city as Ææ (a term linked
to masculine descent), while when considered with respect to the outside world, the city is Åæ ºØ,
mother of other cities, of colonies.
 The Athenian Dramatic Stage 257

perversion of a situation and of a type of speech that in theory should have

positive connotations; along the same lines, the possibility may be left open of
positive connotations, in specific instances, of a letter writing that keeps to its
proper sphere.

5.2.2. Philemo, fr. 10 K.–A.

A rather enigmatic passage by a comic poet of the nea, Philemo, not only confirms
that interest in the invention of writing, and on the purposes of such an invention,
remained alive, but also offers an interesting angle on it. It must be stated right
away that if the text concerns communication at a distance, it is not at all clear that
epistolary communication is specifically meant. In fr. 10 K.–A., a short run of
iambic trimeters from the play Apollo (or possibly the Exile), someone (speaking
for a group) describes the uses of an invention by an unknown character:

Påd   y  yæ H ºÆº  Not only did he find how we could talk together
Æ E, Iå  ºf I Iºººø åæ , being far distant in time from each other,
P n æ  Å   Å b x   ºØ not only how never again one of us ever will
KغŁ H Å , KºłÆ   ı forget anything, but having a look will know
IŒ  F åæ ı  F Æe Y , Iººa ŒÆd events of all eternity, but also
łıåB NÆæe ŒÆºØ a ªæ ÆÆ. as doctors of the soul he left the letters.

The text presents some problems. First, the title: Apollo is what the manuscripts
of Stobaeus, our source for the fragment, have; the correction @ ºØ (the Exile)
has been suggested, without any serious grounds.255 It is difficult to see a
connection between Apollo and these lines, while the use of writing for keeping
in touch may sound appealing in a comedy revolving round an exile; yet, as this is
the only fragment of the play, it is best to accept that it is not possible to
reconstruct the plot, and avoid textual interventionism. Moreover, the fragment
does not really discuss the use of writing for keeping in touch over geographical
distance. What seems to be at stake in lines 1–2 is rather communication over
time (Iå  ºf . . . åæ ), in other words, the preservation of information
for later generations, a theme pursued in lines 3–4. Slightly problematic, however,
is the fact that lines 1–2 and 3–4 repeat the same concept with variations; and also,
the fact that the first person plural of line 1 (ºÆº ) and even more the Æ E
of line 2 imply a reciprocity. Hence, attempts have been made at emending the
first two lines; but none of the ‘interventionist’ solutions proposed until now
appears really satisfying.256

255
See the apparatus in Kassel–Austin. Caecilius wrote a play Exsul, but the two extant fragments
cannot be related in any way to this one. The same applies to the one fragment of Alexis’ Phygas.
256
Thus instead of  ºf . . . åæ  of l. 2,  ºf . . .   and  º ª . . . åŁ e have been
suggested; cf. the apparatus in Kassel and Austin. The first person plural ºÆº  given by the
manuscript S (a correction of ºÆºø) is justified by the presence at l. 4 of a first person plural pronoun,
H and by the metre; the manuscripts A have ºÆºÆØ or ºÆº. Æ E is a correction of
Meineke, universally accepted, for the ÆP E of the codices. But it makes sense only if we imagine
258 Letter Writing and the Polis

As writing is clearly central here, it has been suggested that the inventor should
be identified with Prometheus or Palamedes, or, less plausibly, with Cadmus. The
invention of writing is praised on three counts: writing allows a discussion
between persons who are distant in time; it is a remedy for forgetfulness; and it
can cure the soul. While the first two aspects are fairly traditional (and can be
reduced to one, permanency of written texts through the ages), the last one is
untraditional and new. Unluckily, the fragmentary nature of the text makes it
difficult to know how exactly writing was envisioned to cure the soul: in the form
of a letter from the absent beloved? But there is no hint of her. By means of a
message from the friends left behind by the exile? That requires emendation of the
title. Or by writing acting as a mediator of wisdom, as in the inscription that
according to Hecataeus of Abdera adorned the sacred library of Egyptian Thebes
(D.S. 1. 49. 2: łıåB NÆæE , ‘Healing-place of the soul’)?
At the level of terminology, the choice of ºÆºE for the type of exchange that it
will be possible to establish thanks to the good deeds of the discoverer points to a
type of language often marked as private and female: Philemo builds on the
traditional elements that we have seen at play in Antiphanes’ Sappho, but orients
his reflection differently. It is at any rate clear that the action of writing here
is placed entirely in a personal sphere: ºÆº  strikes the first note, with its
connotation of comfortable small talk; the KºłÆ   ı, ‘having a look’, points
to reading—possibly even silent reading, since the visual aspect is here strongly
underlined; in the same direction goes the mention of the soul at the end.257

5 . 3 . C ON C L U S I O N

It is time to attempt some conclusions. References to writing and written docu-

ments concern all sort of spaces, from the public and official to the personal and
private; they appear ubiquitously, especially in comedy, which is free from the
restraints on tragedy dictated by the necessity to avoid anachronisms. Such
references attest to the widespread diffusion of writing, at least in an Athenian
milieu, but also to an intense reflection on its uses. For practical reasons, each of
the great tragedians of the fifth century has been discussed separately in what is
the canonical order; a different group of fragments, some tragic, some satyric,
others comic, all presenting very specific characteristics, has been then tackled on
its own, and has prepared the way for the discussion of the presence of letters on
the comic stage, which also relies mainly on fragmentary texts. On the whole, this
order reflects, very roughly, a chronological order, allowing us to follow allusions
and references to writing and letter writing in drama from c.470 to c.250. Mapping

conversation among members of a group, over geographical distance. The transmitted ÆP E could be
maintained, notwithstanding the following Iºººø, if we imagine that the invention allows the group
to speak to some other group (posterity) mentioned earlier in the text.
257
Note that a fragment of a Palamedes attributed to Philemo survives, fr. 60 K.–A., preserved by
Stobaeus—but it does not look like a comic fragment; a number of critics have suggested Euripides’
Palamedes (cf. the apparatus of Kassel and Austin, and Euripides fr. 585 Kannicht). The most likely
explanation is that the passage of Philemo and the heading of the following passage in Stobaeus, with
reference to Euripides, may have fallen out in the process of textual transmission. A letter figures also in
a fragment from an uncertain play of Philemo, fr. 132 K.–A.; see above, 243.
 The Athenian Dramatic Stage 259

changes over time is difficult, however, since what we have is only a minimal portion
of the overall production of plays and since many of the plays are not dated, or only
conjecturally so. As we shall see, there is no ‘unitary’ view of writing: in most cases,
the context within which the reference appears orients the interpretation.
The image of writing on the tablets of the mind had been used as a metaphor for
permanent, strong recollection in Aeschylus and Sophocles (and before them in
Pindar).258 The metaphor will reappear in the Theaetetus as one of the explan-
ations for perception and will have a long history.259 Seen from the point of the
support (one’s own mind), this is obviously a very personal, and often secretive,
writing; but it need not be so, as the instances of its use by Dike and Hades show.
How should this metaphor be evaluated? In most of the Aeschylean and Sopho-
clean passages that employ the metaphor, the result is indeed an accurate remem-
brance; but the further development of the metaphorical imagery of writing, such
as the writing of oaths on water (or ashes, or wine, as in comedy) points to the
untrustworthiness of similar metaphorical writings.260 And that is indeed the way
in which the writing in the mind had been presented in its earliest appearance, in
Pindar’s Olympian 10. There, the text written in the mind, although still present,
had been forgotten: it needed a Muse to read it out and to make it reach its
addressee. As Segal puts it, ‘By the playful figure of an invisible writing, Pindar is
asserting that the oral kleos is more reliable a memorial than the written word.’261

258
The earliest datable instance of this metaphor is in Pindar (Olympian 10. 1–3, an ode composed
some time after a victory in 476 bc); the metaphor then turns up in Sophocles’ Triptolemus (468 bc?),
in Suppliants (463 bc), and Oresteia (458 bc), in the Prometheus Vinctus, and again in Sophocles’
Trachiniae and in the Philoctetes (409 bc). The Dike fragment, if connected with the Aitnaiai, would
provide an instance as early as Pindar’s, since that play celebrated the foundation of Aetna by Hieron in
476; but this is very hypothetical. I accept here the traditional date for Aeschylus’ Suppliants; but see
Scullion 2002 for a much earlier date (c.475 bc).
259
Plato, Theaet. 191 c–192d; for the later fortune of the metaphor, see Draaisma (2000: 24–48).
That the wax tablet of the mind is a gift of Mnemosyne (Theaet. 191d) ties in well with the fact that in
Greek tradition the purpose of the invention of writing is memory, rather than communication: see
above, ch. 3. Plato must have had in mind here the extraordinary passage from Aristophanes’ Clouds, in
which Strepsiades proposes to melt from afar through a crystalline stone the wax of the tablet so as to
avoid a suit (Ar. Nub. 770–2: › ªæ ç Ø c ŒÅ › ªæÆÆ, | Iøæø a z  æe e
lºØ  | a ªæ Æ KŒ ÆØØ B KB ŒÅ;). The verb used, KŒŒø, is also used in the Prometheus
Vinctus by the chorus for the fading of memory, in order to express their decision (PV 535: Iºº  Ø
 K Ø ŒÆd   KŒÆŒÅ), and in Critias B 6 (= Athen. 10. 432 d), ºBØ KŒŒØ Å Å
æÆ ø, a forgetfulness induced by excessive potations. The inverted metaphor (eating the wax of the
tablet and thus forgetting; but in the comic context this probably was presented as reality and not
metaphor) appears in Aristophanes fr. 163 K.–A., from the Gerytades: c  ºŁÆ KŒ H ªæÆÆø
XŁØ , ‘They ate the wax from their writing-tablets’, possibly, as hypothesized by Bergk, said of
ambassadors who out of greed (or hunger, as Kassel and Austin suggest) had eaten the wax of the
tablets that they were supposed to carry.
260
One may contrast statements emphasizing the ability to remember of persons ignorant of
writing, as in Cratinus fr. 128 K.–A., from the Nomoi: Iººa a ˜’ PŒ r ’ ªøª ªæ Æ’ P ’
KÆÆØ, | Iºº’ Ie ªºÅ çæ ø  Ø· Å ø ªaæ ŒÆºH, to be interpreted, following Kaibel, as
the answer of someone who has been asked to read a passage of Homer or Solon, and who answers
saying that he does not know the letters, but remembers.
261
Pind. Ol. 10, 1–6: ‘The Olympic victor read me (I ªø  Ø), the son of Archestratus, where it
has been written in my mind (ŁØ çæe KA ªªæÆÆØ). For I owed him a sweet song, and I have
forgotten (KغºÆŁ ). But you, Muse, and the daughter of Zeus, unforgetting Truth (º ŁØÆ), with a
hand that puts things right, keep from me the blame for lying, which wrongs a friend’, with Segal 1985:
212; the poem is analyzed at 210–12. See also Ford 2002: 118–19 and n. 25.
260 Letter Writing and the Polis

Interestingly, with Euripides metaphorical writing in the mind disappears, and

‘solid’ writing becomes, in a very positive sense, the guardian of common memory
and oaths against the practice of simply ‘remembering’, as in the elaborate finale
of Euripides’ Suppliant Women. This writing receives a peculiar ‘added value’:
from the support on which it is engraved (a bronze tripod of Heracles, left to
Theseus), from the ritual that accompanies the oath taking and the engraving, and
from the status of the location where the support and text are finally dedicated,
Delphi. Another form of ‘solid’ writing, attested in tragedy only in Euripides, is the
epigram (found in Troades, 415 bc, and in Phoenician Women, c.410 bc). The
presence of epigrams in tragedy is obviously linked to a reflection on forms of
commemoration. Their writtenness, as against the ‘spoken’ epigrams also attested
(e.g. in the Alcestis, 438 bc), may be explained by the desire to underline the
shocking character of these epigrams, whose content is impossible to utter—these
are epigrams that will not be written.
Besides these traces of a reflection on memory, commemoration, and writing,
we find also references to official documents of the polis, such as laws, decrees, and
official lists. References to such documents are rather infrequent in Aeschylus:
they are alluded to in his Suppliants (463 bc?) as something foreign (even sealed),
and as a foil for the ‘real thing’, the free public word from a free mouth, while the
reference in one of Aeschylus’ fragments to a grammateus (in a comparison: fr.
358 Radt) simply attests to the mediated presence of official writing. In Sophocles,
too, official writing figures only in a limited way: through the allusion in Antigone
(c.442 bc?) to unwritten laws, which somehow must imply written laws; and
through the references to a ‘real’ list (metaphorical writing in the mind had been
used for lists too), in the Achaion syllogos (fr. 144 Radt). In Euripides, discussion
of this type of writing acquires a remarkable importance. For Theseus (Suppliant
Women, c.423 bc?) written laws guarantee equality, in a positive sense; Hecuba
(Hecuba, c.424 bc) had highlighted their negative aspect. These very different
ways of viewing the same type of writing depend not on writing itself (it equalizes
in both instances, from above or from below), but on the perspective of the play as
a whole, and on that of the speaking character. Both points of view can be shared:
problematizing concepts is after all one of the hallmarks of tragedy, and a number
of plays respond to each other in an intertextual discourse, which becomes a kind
of commentary on the Athenian polis and its culture. Comedy enters fully into
this discourse, and if most of the references to writing in comedy have to do with
lawsuits, the writing of laws and decrees is also mentioned more than once, again
with ambivalent effects: decrees are on sale, but they are also used to give
authority; the very stelae on which they were engraved are mentioned, and at
the same time made little of, as an item against which one may lean and relieve
oneself; oracles are on sale, or brought on stage in boxes, to give a solid foundation
to some (usually wildly implausible) great enterprise; in the Women at the
Assembly, the ‘letter of the law’ is cited to support equality—in the sharing of
men by old and young women.262

262
Inscribed stelae as part of everyday conversation: Ar. Lys. 513; relieving oneself against a stele:
Ar. Av. 1050–4; for the other references, see above, 242 and n. 197, and Slater 1996.
 The Athenian Dramatic Stage 261

Another aspect that seems to have excited interest is the relationship of

painting/graphe and reality. Aeschylus and Euripides make important use of
ªæÆç for pictorial images; the fact that this use is not attested in the extant
oeuvre of Sophocles may or may not depend on the hazards of transmission. Most
instances fall into one of the two following categories: either the reference to the
image is made in such a way as to call attention to its status as an image, provoking
what Froma Zeitlin has defined as ‘hyperviewing’, or auditive and visual data are
paired as sources of information. Pointing to images qua images, in a medium
such as drama that relies not only on sounds, as had been the case for epic
performance, but also heavily upon the visual (actors and gestures, stage setting,
props), brings the audience as close as can be to a rupture of dramatic illusion
(notice, for instance, Aeschylus’ Theoroi), but at the same time forces consider-
ation of the status of the image in question. Tragedy (especially Aeschylus and
Euripides) thus pursues the reflection on the way images function, including
explicit comparisons of images to song, attested already from the end of the
sixth century by Simonides.263 The remarkable power of images is evident in
Aeschylus’ Theoroi and Agamemnon (where the references to images had also
served the purpose of linking together the characters of Iphigenia, Agamemnon,
and Cassandra); the best Euripidean instance is that of Hecuba asking to be
viewed as a painting, in the play that bears her name. In general, the use of the
term graphe for an image has the effect of calling attention to its artificial
character;264 the power of this depiction resides in the fact that painted images
are enhanced, crafted representations of reality (this has not to do with realism,
but with the fact that an image may mean much more than its simple primary
referent). Images (just as written words) speak, in Aeschylus as in Euripides. And
yet, already in the Agamemnon, Cassandra’s reflection on the ease with with
images can be wiped off had hinted at the instability—of images, of writing, and
of men—an instability stressed in Euripides too, for instance in a fragment of the
Peleus.265
As for the pairing of auditive and visual data as sources of information, this is a
specifically Euripidean feature.266 It produces less complex effects, but is no less
interesting: in most cases, seeing and hearing are treated as exactly parallel, and as

263
See for Simonides the careful discussion of Ford 2002: 93–112. For tragedy, Segal 1993; Zeitlin
1994; O’Sullivan 2000; Steiner 2001: 44–56; and Wright 2005: 318–25. A similar reflection on the
relationship between image, reality, and writing was being developed by the vase-painters: see above,
186–7.
264
Cf. Empedocles, 31 B 23 D.–K.: ªæÆçE mix drugs to form images resembling all things (N Æ
AØ IºªŒØÆ), with Segal 1993: 100–1.
265
Segal 1993: 101 suggestively states that ‘For Euripides’ contemporaries, the art of “inscription”,
graphike [Segal is here talking of paintings] embodies an ambiguous and paradoxical relation between
surface and depth’, referring to Empedocles (above), Gorgias (Helen 18), and Socrates in Xen. Mem. 3.
10. 1 (‘graphike is the figuration of visible things’, ªæÆçØŒ KØ NŒÆÆ H ›æøø). Comedy will
also pursue this discourse: cf. the plays ˆæÆç by Alexis (its fr. 41 K.–A. tells of a man from Samos who
fell in love with a marble statue, so possibly the protagonist fell in love with a painting; same topic in
Philemo fr. 127 incertae fabulae K.–A.); ZªæÆç  (‘Painter’) by Antiphanes (see fr. 102 K.–A.),
Diphilus, and Hipparchus; Zøªæ ç ı (‘Painters’) by Anaxandrides; and Pictores by Pomponius.
266
Note, however, the ‘inscriptions’ on the shields of Aeschylus’ Seven (467 bc): there, writing had
been used in the context of a messenger’s report, in synaesthetic combination with other communi-
cative codes, the overall message being the result of the various pairings of codes.
262 Letter Writing and the Polis

having the same value. This prima facie runs counter to the traditional position of
privileging the view (ZłØ) over the oral report (IŒ ), a position attested for
instance in Heraclitus (fr. 101a D.–K.: ‘for the eyes are more precise witnesses than
the ears’, OçŁÆº d ªaæ H þø IŒæØæ Ø  æıæ). The privileging of the
visual over the aural/auditive is also present in the prose of Herodotus, Thucyd-
ides, the sophists, and is important in the context of the contemporary Athenian
judicial system. Yet, as is clear from many of the Euripidean passages, ‘seeing’ the
painted image, or seeing the writing, cannot be (and will never be) considered as
equivalent to ‘seeing the thing’, as a form of ZłØ, because information transmitted
through writing is mediated communication.267 Hence, its pairing on the same
level with akoe, as another instance of mediated communication. Behind this is
clearly a lively a debate on knowledge, and on how it can be best attained; but this
is also obviously relevant to epistolary communication.
Critical examination of writing and its role in society also involves oracular
collections, i.e. the writing down of the god’s voice, and more generally, books.
Books are very frequently mentioned in comedy, and appear also more than once
in Euripides. If they have managed to infiltrate tragedy, they must by now be a
relatively important presence in Athenian society. Twice in the Hippolytus (428
bc) books are mentioned that are certainly not operating in a direction conform-
ing to the ideals of the polis, first in the words of the nurse to Phaedra, and then in
those with which Theseus insults Hippolytus.268 But otherwise, books are not
marked negatively: rather, they are presented as complementary to song, as
parallel to oral traditions. This is the case in the Erechtheus, as well as in the
Alcestis, the earliest surviving play by Euripides (438 bc), and in the Iphigenia in
Aulis (performed in 405 bc, after Euripides’ death). There may be a hint of the
metatheatrical in these references to books, for drama relied on scripts;269 Euripi-
des’ particular fondness for writing was not lost on Aristophanes, who in Frogs
had Euripides say that, when he inherited the art of tragedy from Aeschylus, he
took out the inflammation from it and ‘added extracts of fluencies filtered out of
papyri’ (943; cf. also 1407–10: Euripides can put his papyri on the scale, where he
will still prove lighter than two lines of Aeschylus). It is, however, worth noting

267
Segal’s excellent discussion (1993: 101) of view and sound in the Hippolytus is marred by the fact
that he does not distinguish clearly enough between view of facts or objects, i.e. direct view, and
mediated view; hence, his stress falls on the eye–ear asymmetries, rather than, as here, on their
symmetry. His conclusion at 150, that ‘autopsy, which in the fifth century thinkers regard as the surest
and clearest proof, is here [in the Hippolytus] deceptive’, can be accepted only with the proviso that in
the Hippolytus viewing (the viewing of a graphe) is not direct viewing: this is why it can be paired to
hearing.
268
Books may be similarly connoted in comedy: cf. Aristophanes fr. 506 K.–A., ‘This man, either a
book has ruined him, or Prodicos, or some babbler’, or the third-century comic poet Theognetus, fr. 1
K.–A.: ¥øfi ’ › Æø çغ çø fi ıfiŒØ. | KÆææ’ ÆŁ, t Åæ, ªæ ÆÆ· | Iæ ç  ı
e   a ıºÆ. But there are also numerous books that turn out to be simply cooking manuals, or
that are not supposed to have a negative effect on the reader.
269
Cf. Hall 2006: 46 for the discovery at Daphne (south of Athens, along the road to Sounion) of a
tomb dated to c.430–420 bc, in which a roll (the oldest known Greek papyrus), a lyre, a harp, a pair of
auloi, a bronze pen and tablets accompanied the burial of a man, possibly an actor or a singer. The
papyrus roll and the tablets have been now published in West 2013; for the archaeological context and
the other objects see the papers in Greek and Roman Musical Studies 1 (2013). See also the Attic hydria
and the volute-crater discussed above, 183–4 and nn. 2–4.
 The Athenian Dramatic Stage 263

that the terminology used in Euripides is non-specific: tablets (Æ, º ),
leather skins ( ØçŁæÆ—this is possibly the most specific word), or ‘writings’
(ªæ ÆÆ): º /º  is not attested in what is extant of either Sophocles or
Euripides, nor any word formed from that stem; in Aeschylus, º  (Suppl. 948)
is used for a type of writing marked as oriental, and is unlikely to mean ‘book’.
Comedy will use غ  unproblematically, possibly already with Cratinus, cer-
tainly with Eupolis and Aristophanes; as we have just seen, in Frogs Aeschylus offers
to weigh two of his verses against Euripides, his wife, his children, Cephisophon,
and all his books (غÆ, 1409).270
In general, books are connected to specific individuals, and are mentioned for
the effect they have on them (or the connotations they convey); fairly personal is
also the writing down of oracles, individually, or as collections, since the oracle,
once written, can be stored for later reuse by individuals. A tablet with the oracular
pronouncements of the Dodonaean oak plays a prominent role in the Trachiniae.
A reference to a written oracular collection is in Euripides (fr. 627 Kannicht). Both
historiography and comedy show that already the Pisistratids had treasured
written collections of oracles, and Herodotus presents the writing down of the
Delphic answers by people consulting the oracle as normal practice. It is thus not
surprising to find this kind of text in drama. Nor can the ambiguity of the oracle
be imputed to the medium (the writing), even if it is true that the audience of
Trachiniae may have left the theatre with a dim view of writing and its uses.
Finally, letters. These are definitely a personal type of writing. If we look at
epistolary writing against the background of the other types of writing, it becomes
immediately apparent that there is here an imbalance, both at the chronological
level, and, more importantly, in terms of quality. To tackle the chronology first.
Although the opening of the Oresteia (458 bc), with its description of the light
signals arranged by Clytaemestra, had shown a profound attention towards
language and communication, the first clear reference to a letter is in Euripides’
Hippolytus (428 bc); but from this moment onwards, a number of letters figure in
what is extant of Euripides. If we turn to comedy, for the fifth century we have one
possible reference to epistolary communication in Cratinus’ Trophonius, and the
parody of the Palamedes in Aristophanes’ Thesmophoriazusae. The situation
changes in the fourth and third century, when letters play an important role in
the plots, at least if we are to judge from the titles. This may have to do with the
hazards of transmission; more reliable is the second aspect, quality. While more-
or-less casual references to writing are scattered in numerous plays throughout the
fifth century, there are no casual references to letters. Letters are not mentioned

270
Cratinus 267 K.–A. (غØƪæ ç  refers probably to a sycophant active in the writing of decrees,
cf. Caroli 2012); Eupolis fr. 327 K.–A.; Ar. Av. 1288. Turner 1975 is a landmark study of the circulation
of books in fifth-century Athens, although some of its conclusions (especially that the books circulating
would have been mainly those of the sophists, in prose, 23–4) cannot be accepted sic et simpliciter; the
‘books’ referred to in Euripides at any rate are all ‘poetry’ (but then they probably would be); comedy
has references, difficult to evaluate, to books of both poetry and prose (on comedy and books, Caroli
2012). See also Pfeiffer 1968: 22–30 (optimistic); O’Sullivan 1996, stressing that the sophists in their
attempt to present their prose as the ‘successor to the poetry of traditional Greek culture’ had to ‘adopt
the book as their symbol’ (117, with reference to e.g. Ar. fr. 506 K.–A., mentioned in n. 268 above); a
very balanced, thoughtful assessment in Ford 2002: 152–7.
264 Letter Writing and the Polis

for the sake of a gesture towards ‘reality’; no character idly thinks of sending a
letter. When a letter is mentioned, it plays an active role in the intrigue.
In Euripides in particular, letters do not have only a communicative purpose,
they are central elements in the plot. They may help push the action forward, or
also slow it down, as with the recognition scene in the Iphigenia in Aulis; they may
be used to bridge situations in which a direct encounter is not possible or
desirable.271 They are thus an ‘active’ type of writing, as opposed to the other
types, mentioned but not ‘used’. Letters may also constitute an interesting scenic,
visual prop. Moreover, bringing letters to the stage allows the playwright to pursue
the exploration of narrative techniques, to play with intention, delayed communi-
cation, realization, to bring in touch a time past and a time future. But also, the use
of letters makes a subtle characterization of the protagonists possible: the charac-
ters that write letters are those that are not in a position to have a face-to-face,
open confrontation because of their geographical situation, or that choose not to
have such a confrontation because their intents are not acceptable to the wider
community. Since they do not want to (or cannot) remain inactive or yield, they
choose epistolary means. This applies to Phaedra, to Stheneboea and Proetus, to
Odysseus deviously plotting against Palamedes, to Iphigenia, who, from her
captivity in Tauris, has no choice but to resort to a letter, and to Agamemnon,
who cannot face telling Clytaemestra (or the army) the truth. While there are
nuances, and while it is the plot as a whole that inflects the connotation of the
letter, it is the case that in the majority of the situations epistolary exchanges are
negatively marked.
From Euripides, letters move into comedy, and in two ways that correspond to
the two main purposes they had had in Euripides’ plays. On the one hand, they
become elements of the plot, and all the more naturally, since letters from the
fourth century onwards indeed begin to circulate more and more abundantly.
Here, the numerous plays bearing a title ‘Epistole’ or ‘Epistolai’ are relevant; even if
we cannot know how letters functioned within these plays, the Plautine renditions
may give us an idea. On the other hand, references to letters in comedy serve to
prolong a reflection on public speaking, public writing, and their respective uses
within the polis, of which we may have some scant remains going back to fifth-
century comedy in the stories concerning the epistolary greeting used by Cleon for
his official communication after Sphacteria. Such a reflection remains very much
alive throughout the fourth century, both in comedy and outside it. In comedy, it
is best exemplified by the riddle of Antiphanes’ Sappho. In the next chapter, we
shall look more closely at how the letters and speeches of the rhetors interacted on
the Athenian political and legal scene.

271
Rosenmeyer 2001: 65: the letters become ‘agents in the plot, provoking reactions and directing
events kinetically’.
 6
Letters on the Legal and Political Stage

According to many scholars, the transition from the fifth to the fourth century
constitutes a watershed in terms of literacy (at least in an Athenian context), with an
apparent increase in interest in documents and their preservation. Starting with the
late fifth century, Attic orators, for example, cite documents and laws codified in
writing as part of the evidence. But a compelling case has been made that archives of
some kind existed in Athens even earlier and that we should reckon with a fairly
widespread, and ever increasing, use of writing throughout the fifth century.1 How
do letters, as a specific type of written document, fit into this picture?
From the mid-fifth century onwards, so Pébarthe has recently argued, the
‘letter’ becomes the accepted and customary instrument for communicating at a
distance. Thus, in a discussion of the letters mentioned by the orators in the course
of their speeches, he remarks: ‘dans cette deuxième moitié du cinquième siècle, la
correspondance écrite est banale’.2 Some later texts may seem to bear this asser-
tion out: for instance, a passage from the Apollodoran speech Against Timotheus,
delivered at some time in the late 360s, but referring here to the situation of
Athens in the year 373 bc, may be understood as presenting oral messages and
letters in parallel, as the two ways in which information concerning the army
could be conveyed home: publicly, through a report in the assembly, and privately,
through letters by each individual to his family.3
And yet, Pébarthe’s is probably an excessively optimistic view of the situation,
and would need qualification, at least from a chronological perspective. A very
different assessment of the situation has been made by Sian Lewis, who suggests
that ‘personal communication through writing letters never became common in

1
e.g. Sickinger 1999 for the archives; Pébarthe 2006 for literacy (against the more conservative
vision of Harris 1989, Thomas 1989). Excellent discussion of the way literacy and orality are integrated
in Athens (and in the early Greek world) in Faraguna 2006; Faraguna 2007 focuses on the interplay of
literacy and orality in legal situations.
2
Pébarthe 2006: 81–4 and 86. He bases this claim on a disputable analysis of early ‘letters’ on
ceramic and lead from Athens (see above, 44); and on Antiphon 5. 53–6 (discussed below).
3
[Dem.] 49. 13: ‘those who came from the army were reporting before the assembly
(Iƪªººø K fiH  fi ø) the distress and need that existed, and at the same time individuals
kept receiving word/letters from their relatives and friends (Ø KØ  ºH Œ  ı ıŁÆ  ı Ææa
H NŒø ŒÆd KØÅø) telling of their plight’. Note, however, that Ø’ KØ  ºH is ambiguous:
while L. Gernet (Démosthène. Pladoyers civils, iii (Paris, 1959)) translates ‘connaissaient la situation par
des lettres de parents et d’amis’, A. T. Murray (Demosthenes: Private Orations, ii (Cambridge, MA,
1939)) assumes it to mean ‘through messages’, since he translates: ‘kept receiving word’. The problem is
the ambiguity inherent in KØ ººø/KØ  º; but on the whole, by the fourth century bc KØ  º
seems indeed to mean ‘letter’ rather than ‘injunction’ or ‘oral message’.
266 Letter Writing and the Polis

Classical Greece’ because written messages were not felt to be ‘trustworthy

enough’.4 The practicalities of sending a letter in the ancient Greek world are
the main reason behind Lewis’s position.5 And yet notwithstanding these difficul-
ties, letters were possibly a relatively common instrument of communication
between individuals already in the late fifth century bc, and certainly so from
the fourth century onwards. The fourth century is indeed the moment in which
epistolography develops in a massive way, as references to letters in literary
sources indicate. If everyday letters from Greece appear to us to be almost non-
existent, this is simply because papyrus, the material on which they were written,
has not survived the test of time in Greece; by contrast, we have an enormous
amount of correspondence of all types from Hellenistic and Roman Egypt.6 Letters
were thus certainly used as a means of communication between individuals. What
of official communications?
Some Athenian decrees mention letters sent to the Council and the People by
foreign kings and dynasts;7 the corpus Demosthenicum shows that Demosthenes
could produce as evidence a number of official letters, which would have been
taken from the Metroon and read out by a clerk. About a half of the letters
mentioned by Demosthenes were sent by Philip of Macedon, to whom tradition
attributes an extensive use of letter writing for diplomatic purposes, as well as the
institution of an official chancery at least from 343 bc, when Eumenes was
nominated as ªæÆ Æ or secretary (Plut. Eum. 1. 2–3; Nepos Eum. 1). Alexan-
der continued in the footsteps of his father. Some of the letters he sent to the Greek
cities were published on stone by the cities themselves, as in the case of Chios and
Priene (Syll.3 283 and OGIS 1=Rhodes and Osborne nos. 84 and 86). Others, such
as the letter concerning the return of the exiles (D. S. 18. 8. 4), enjoyed an ample
diffusion in Greece and entered the literary tradition; yet more are known only
through passing references in literary sources, such as the letter sent to the
Athenians in 335 bc ordering them to hand over the leaders of the anti-Macedo-
nian party (Arr. Anab. 1. 10. 4), Alexander’s request to the Athenians for ships
(Plut. Phoc. 21. 1), and the letter sent to the Plataeans stating that he would rebuild
their city (Plut. Alex. 34. 1). From this time onwards, letters are the normal
instrument of diplomatic communication of the Hellenistic kings.8

4
Lewis 1996: 142 and 152. See also—on different grounds—Sykutris 1931: 217.
5
See on this 10–13 above.
6
For Egypt (and women) see Cribiore 2001; Bagnall and Cribiore 2006. For fourth-century Greece,
compare Aristotle’s remarks on the likelihood that Iphigenia would want to send a letter (Poet. 1455a
16–19), discussed above, 230.
7
Rhodes 1972: 43 n. 7 refers to Tod 133 = Rhodes and Osborne 2003, no. 33, a honorific decree for
Dionysius I of Syracuse and his son, dated to 368 bc, in which a letter (æd H ªæÆ ø, l. 8) sent by
Dionysius to Athens is mentioned; Tod 167 = Rhodes and Osborne 2003, no. 64, an honorific decree
for Spartocus, Paerisades, and Apollonius, dated to 346 bc, in which a message (æd z K غ) sent
by Spartocus and Paerisades is mentioned, together with the ampler exposition by the ambassadors; IG
ii2, 387 (grant of citizenship dated to 319 bc, mentioning in the motivation clause a message sent by
Polyperchon, [ . . . æd z — ]ºıæåø K[ ]- | [ƺŒ, ll. 8–9); and IG ii2, 486 (grant of citizenship
dated to 304/303 bc, alluding to a message: æd z › Æ Øº]- | f K غ E [ ıºE ŒÆd HØ
 øØ], ll. 11–12). Only the first instance, however, refers explicitly to a letter or written text; the other
texts all have simply a form of KØ ººø. See also Pébarthe 2006: 293 n. 17.
8
The letters of Philip will be discussed below. Demosthenes cites passages relatively often: cf. 9. 27;
18. 39, 77–8, 157, 166, 167; 19. 38, 40, 51, 161, 187) or refers to them (cf. 2. 7; 4. 37; 7. 1, 27, 32–3; 8. 16;
 The Legal and Political Stage 267

Philip and Alexander’s letters exemplify a particular type of official writing.

They are royal letters, originating from the Macedonian chancery and reflecting
the will of an individual, the king. In this sense, although they address ‘the many’,
they are close to private letters. Akin to the royal letters from this point of view
are those that we might call ‘open letters’, letters of specific individuals who
had apologetic or political purposes, such as some of the Isocratean letters and
the Demosthenic ones; for this type of writing, too, the fourth century marks a
decisive change.9
Things are not so clear-cut with letters and poleis. At the level of interstate
diplomacy, very little is known; but the impression is that most affairs were
discussed orally through embassies (ambassadors could of course request letters
that might facilitate their mission, as Hegesias of Lampsakos did in c.195 bc).
A certain reluctance to inscribe (and thus confer a monumental format on) the
writings of another state or polis within one’s own polis can be discerned even for
non-epistolary writings (excluding, of course, honorific decrees).10 This, however,
does not mean that writings were not received, simply that they were kept in the
archives rather than publicized. Similarly, although we do not have epigraphical
trace of them, letters appear to have been the normal means with which strategoi
on campaign or ambassadors on a mission would communicate, at least from the
fourth century onwards, with the polis. This emerges clearly from a passage of the
Athenaion politeia, stating that out of the four monthly assembly meetings, two
should be dedicated to dealing ‘with all other business, at which the laws prescribe
that three cases of sacred matters are to be dealt with, three audiences for heralds
and embassies, and three cases of secular matters’, and further specifying that
æ æå ÆØ b ŒÆd ƒ ŒæıŒ ŒÆd ƒ æ Ø  E æı Ø æH , ŒÆd ƒ a
KØ  ºa çæ    Ø I ØÆ Ø. [ . . . ] ([Arist.] Ath. Pol. 43. 6)
heralds and ambassadors meet the prytaneis first, and the bearers of dispatches give
them to these.

11. 1, 17; 12. 22; 18. 221; 19. 36, with Lewis 1996: 146–7. One such letter even entered the Demosthenic
corpus ([D] XII). The bibliography on Alexander’s letters is huge. For the false ones, which will be one
of the main sources of the Romance of Alexander, see Merkelbach 1954: 1–5 and 32–55; Rosenmeyer
2001: 171–92. On Alexander’s chancery, Berve 1926: i, 42–54. Collection of letters of Hellenistic
kings: Welles 1934. Kienast 1973a suspects a Persian influence behind the choice of this mode of
communication.
9
Isocrates’ letters: Nicolai 2004: 120–7. Demosthenic letters: Clavaud 1987: 3–68; Goldstein 1968,
who believes in the authenticity of the first four. The apologetic nature of the open letter in the fourth
century is well brought out by Goldstein (1968: 97–132); these letters can be classified as litterae
negotiales, a genre recognized as such by ancient theorists, who connected them with oratory (cf.
Demetr. Eloc. 234: ¯d b ŒÆd º Ø ŒÆd Æ ØºF Ø ªæç   ø Æ  ØÆFÆØ Æƒ KØ  ºÆd ØŒæe
KÅæ ÆØ ø). However, the ancient open letter was usually addressed to a foreigner, or a citizen
residing outside the polis, not directly to the public; and it concerned general matters rather than
specific ones (Goldstein 1968: 99). The Demosthenic letters are in this respect an exception, due to the
situation arising out of the orator’s exile.
10
Hegesias: see above, 169. Covering letters: see below, 312–15 and n. 49. A simple administrative
letter, besides its primary function of solving a problem or imposing a decision, will publicize the
authority from which it emanates, if monumentalized on stone; the propaganda value may sometimes
be more important than the content of the text itself. Hence the natural reluctance of Greek poleis
towards admitting the writings of the ‘other’ in their centre. See Bertrand 1985: 107–8; Funck 1996:
199–210; Ma 1999: 147–9.
268 Letter Writing and the Polis

Outside Athens, allusions to letters in literary texts, or for instance the mention,
in the Lindian chronicle, of letters sent by the Rhodian priests Gorgosthenes and
Hierobulus to the Rhodian boule and to the Lindian mastroi respectively, describ-
ing the objects that had been dedicated to the goddess Athena in Lindos, show that
letters were among the accepted instruments of internal communication in most
Greek poleis.11
Thus, it is important to differentiate sharply between official and personal
communication; between different modes of official communication (e.g. from
one polis to another or between a polis and one of its magistrates); and between
various types of letters and occasions—all the while keeping in mind the issue of
the public monumentalization of pieces of correspondence, as opposed to their
simple preservation in archives.
In this chapter, I shall try to map out the connotations of these epistolary types
through a detailed analysis of the letters (or mentions of letters) preserved within
the corpus of the Attic orators.12 The reasons for this choice are multiple. In the
first place, working on a homogeneous corpus of deliberative and forensic
speeches preserving various types of documents (decrees, laws, and letters) may
help highlight the specificity of letter writing in relation to other forms of writing.
Secondly, with the speeches of the orators we are at the interface of orality and
writing, in more than one sense.13 The speeches were delivered orally, and were
meant to give the impression of being composed on the spur of the moment, but
were in most cases prepared. Yet the documents were read out by a clerk, and this
shift from the orator to the clerk, creating a break, will have enhanced the written
status of the documents.14 At the same time, letters remain very close to oral
communication: they are a (formalized) speech, in which the person of the sender
is very markedly present.15 It may not be an accident that the first collection of
‘literary letters’ is probably that of Lysias; and that in the work of Isocrates there is

11
The use of letters in the Hellenistic period, as implied by the narratives of ancient historians, is
discussed in ch. 4. 4; as for the Lindian chronicle, see FGrH/BNJ 532 F 2, 5–9, with the commentary of
Higbie, ad loc. The opening of the text, reporting the decision to inscribe the list of all objects ‘from the
letters and from the public records’ ([K]Ø  ºA ŒÆd H åæÅ Æ[Ø H], F 1 lines 6–7), implies,
however, a distinction; and the fact that the letters concern the most ancient objects, while the public
records refer to more recent donations, attests to the generalized perception of the increasing value of
documentation (cf. Higbie, ad loc.). The letters of the priests may have been connected with the
destruction of the temple by a fire, in 392 or 391 bc: see again Higbie, ad BNJ 532 F 2, 5–9.
12
I will follow the traditional dating of the various speeches, as it is enough for my argument if the
speeches discussed belong to the fourth century bc; for the same reasons I do not enter into discussions
of authenticity.
13
See for instance Nicolai 2004: 20: ‘l’intersezione fra oralità e scrittura nell’oratoria è continua’.
Nicolai (at 22–3) also makes the important point, however, that early prose deliberately keeps its
distance from spoken language; imitation of spoken, everyday language is something that begins later
with Lysias and will reach its climax with Demosthenes. Isocrates was never part of this movement: his
text is created to be read, individually or in narrow circles.
14
I develop here an astute observation of Olding (2007), concerning the request made by Aeschines
in his Against Timarchos that the clerk read some poetic passages, a practice very unusual (normally the
orator would have read them), and here possibly used to give authority to passages that diverge from
the vulgata but support Aeschines’ case.
15
An illocutory tendency; cf. Stirewalt 1993: 5 (quoted above, vii n. 9) on parousia, the projection of
the official’s person and the transmission of his authority through the letter, and the discussion above,
3–4, 8–9.
 The Legal and Political Stage 269

no strongly clear-cut distinction between speech and epistle.16 The analysis of the
uses of letters in the speeches will confirm, but also nuance, the interpretation
proposed above on the basis of the fragment from Antiphanes’ Sappho.
In what follows, the focus will be mainly on letters; but I shall on occasion
contrast the reports of letters and the use made of them in the speeches with
that of other types of written document.17 I shall discuss the early logographers
(Antiphon, Andocides, Lysias, and Isaeus), Demosthenes and the corpus
Demosthenicum, and Aeschines, making only passing reference to the other
orators. At the end, a discussion of Isocrates’ work will address the issue of the
distinction between epistle, speech, and treatise. What will not be discussed here
are the letters attributed to the ancient orators (with a partial exception for
Isocrates): the reason, beyond the difficulty posed by the issue of the authenticity
of these documents, is that for the references to (or citations of) letters in speeches
we have a context, the speech, that may help to evaluate them; such a context is
lacking in the case of the letters.

6.1. ANTIPHON

Writing and written documents make a gradual entrance in the arena of the law-
courts.18 Thus, it is no surprise to find that no laws or decrees are cited in
Antiphon, although a number of witnesses are summoned. But in two passages
there are references to writing, and, in particular, to letters. One, from the speech
On the Murder of Herodes, delivered probably c.420 (at any rate between 420 and
412 bc), is a staple of all discussions of the perception of epistolary writing in
Athens. Here is a translation of the text:
The prosecution further allege that they discovered on board a note stating that I had
killed Herodes (ªæÆ ÆØ  æE K fiH º ø fi , . . . , ‰ I ŒÆØ Ø), which I had
intended to send to Lycinus. But what need had I to send a note, when the bearer
himself was my accomplice? Not only would he, as one of the murderers, have told the
story more clearly in his own words, but it would have been quite unnecessary to
conceal the message from him, and it is essentially messages which cannot be disclosed
to the bearer that are sent in writing. 54. Then again, an extensive message would have
had to be written down, as its length would have prevented the bearer remembering it.
But this one was brief enough to deliver—‘The man is dead.’ Moreover, bear in mind
that the note contradicted the slave who was tortured, and the slave the note. The slave

16
Lysias: Blass 1887, i, 422–3 (the letters ‘grösserentheils wenigstens echt gewesen sein mögen’); the
very few fragments can be easily consulted in Carey’s (2007) edition. Isocrates: Too 1995: 195–9;
Livingstone 2001: 5–8; Nicolai 2004: 118–27; and below. On epistolary collections see Trapp 2003: 12–33.
17
All passages in which epistole is used have been examined, as well as most of those mentioning the
exchange of some form of written note, be this designated as grammata, or by a verb of writing. A note of
acknowledgement has, however, a different status from a letter: it is simply a written statement of something
also discussed orally by the interested parties. On the use of written documents and witnesses as evidence in
Athenian courts see Thomas 1989: 41–3; Bonner 1905; Harrison 1971: 133–54; Sickinger 1999; Thür 2005.
On the role played by writing in the trial, see Bernard 1999: 205–17; Faraguna 2007 and 2009; Gagarin 2009.
18
On the gradual ‘encroachment’ of writing in legal contexts throughout the fourth century see
Thomas 1989: 41–4.
270 Letter Writing and the Polis
stated under torture that he had committed the murder himself, whereas the note
when opened revealed the fact that I was the murderer. 55. Which are we to believe?
The prosecution discovered the note on board only during a second search, not during
their first one; they had not hit on the idea at the time. It was not until the first witness
had said nothing to incriminate me when tortured that they dropped the note in the
boat, in order to have that, if nothing else, as a ground for accusing me. 56. Then, once
the note had been read and the second witness, when tortured, persisted in disagreeing
with the note, it was impossible to spirit away the message read from it. Needless to
say, had the prosecution supposed that they would induce the slave to lie about me
immediately, they would never have devised the message contained in the note. Call
me witnesses to confirm these facts. (Antiph. 5. 53–6, trans. Maidment)
This passage shows that the fact of sending a written note, a letter, was not in itself
extraordinary:19 the prosecution must have produced the note with the rest of the
evidence. However, what Antiphon makes of this note is interesting. The fact that he
accuses the prosecution of having forged the letter (an accusation that resonates with
the contemporary debates of tragedians and sophists, as shown by the one developed
around the story of Palamedes) cannot be taken to show that letters were not used
for long-distance communications—the readiness to attempt a forgery might almost
be interpreted in the opposite sense. But the defendant makes two important points
about letters: the first and most important one is that written messages are necessary
only when one does not want the bearer to learn the content of the message. This
reminds one of a number of Herodotean and Thucydidean letters (in particular, the
story of the contacts between Xerxes and Pausanias), and more generally of the
Bellerophontes story, the archetype of the genre. Written notes containing private
information, ‘letters’, are thus as a matter of course likely to be regarded as suspect.20
Would the orator have risked making such a point if writing letters had been
common practice among the members of the jury? The second point is that the
only other reason for writing letters is when the message is excessively long, too
much so to be remembered; and this clearly was not the case with the grammatei-
dion. It may be that Antiphon added this because he felt that his first point alone
might not have been convincing; on the other hand, this too is probably a rhetorical
ploy: the fact is that all the letters we have are extremely short, and that concision in
the writing of letters was always praised by the ancient rhetoricians. His statements
show at any rate that a letter is not envisaged as being anything more than a message
couched in writing, and to be used only if other reasons, such as the length of the
message or secrecy, make oral transmission difficult.
Thus, while letters have a place in the world of long-distance traders, they at the
same time carry a degree of suspicion. This is confirmed by another passage in
Antiphon’s corpus, from his Prosecution of a Stepmother for Poisoning, another
forensic speech, delivered at some point between 420 and 411 bc. Here the
prosecutor at the close of his speech affirms that victims of murder, if they have
the time, summon friends and relatives and tell them who the murderers are,
charging them to take vengeance; or, failing this, they make a statement in writing
(ªæ ÆÆ ªæç ı Ø), call slaves as witnesses, and reveal the murderers to them.
The speaker adds that his father followed the first option, called him, and ‘laid his

19
Stressed by Pébarthe 2006.
20
A point highlighted by Harris 1989; on treacherous letters see also Rosenmeyer 2001: 45–55.
 The Legal and Political Stage 271

charge upon him’, not on his slaves. Interestingly, the verb used is K غ, the
same that would have been used for a letter.21 A chiastic structure is at work here
in which the writing and showing to the slaves corresponds to the showing
and ‘transmitting’ to the son (KØ ººø can be used for oral or written messages);
the second procedure is supposed to be the best one. The defence had certainly
stressed the slaves’ ignorance of the facts, and possibly also the absence of any
written indication on the part of the dead man as to the culpability of the
stepmother. The prosecution counters this by emphasizing the status of the son,
and by using a term that substitutes for the missing written statement. There is
thus here a sense that a written accusation might have some weight; but it is only a
vague intimation, quickly countered by the statement that the writing of a message
is a second-best. But for these two instances, no written documents are part of the
argument in Antiphon’s speeches.

6.2. ANDOCIDES, ISAEUS, LYSIAS

Andocides differs from Antiphon, inasmuch as he quotes a number of documents: two

decrees and three laws (besides numerous mentions of specific decrees), along with
three lists of names, in his forensic speech On the Mysteries (delivered 399 bc); and
another decree in his assembly speech On his Return (dated before 403, probably
delivered around 408 bc). There is, however, no mention of letters in any of his
speeches.22 Similar is the case of Isaeus: among the various pieces of evidence, here too,
there is no mention of letters. Other types of documents, however, are referred to.23
A letter appears in one early speech of the corpus Lysiacum, For Polystratus,
dated to c.410 bc: the defendant asks the jury to ‘consider’ ( ŒłÆ Ł) a letter of
his father—but the letter is not read, the content is simply very sketchily summar-
ized, while the weight of the discussion is carried by witnesses.24 Letters are also
mentioned, but again not quoted, in the speech On the Property of Aristophanes
(387 bc): KØ  ºÆ (messages, and not necessarily all of them written ones) sent
by his father from Cyprus were one of the reasons why Aristophanes decided to go
there and join Conon. Here too, no weight at all is laid on the KØ  ºÆ (whose
status, to repeat, is not necessarily that of written messages), and immediately the

21
Antiph. 1. 30: Ka b  ø ± Ææø Ø, ªæ ÆÆ ªæç ı Ø, ŒÆd NŒÆ  f çæ ı ÆPH
K،ƺ FÆØ æıæÆ, ŒÆd ź F Ø ç’ z IººıÆØ. ˚IŒE  K d ø fi Ø ZØ ÆFÆ Kºø  ŒÆd
K غ, t ¼æ, P  E Æı F  º Ø.
22
Resp. Andoc. 1. 77; 1. 8. 3 (decrees) 1. 85; 1. 87; 1. 96 (laws); and Andoc. 2. 23.
23
A law on adoption in On the Estate of Menecles (2. 16); laws in On the Estate of Pyrrhus (3. 53),
along with a number of written depositions; laws regulating the right of inheritance in On the Estate of
Hagnias (11. 11 and 22); decrees in the speech On the Estate of Philoktemon (6. 50), along with a
number of written depositions.
24
Lys. 20. 27: ‘Consider now the letter (KØ  º) from my father, which he arranged to be
conveyed to me, whether its contents were good or evil to your people. In it he had written about
domestic affairs (  ªaæ NŒEÆ KªªæÆ ), and moreover, that when things were going well in
Sicily I should return. Now surely your interests and those of the people there were the same; so that, if
he had not been loyal to the city and to you, he would never have sent such a letter ( PŒ ¼    ØÆFÆ
K ºº).’
272 Letter Writing and the Polis

speaker calls on a witness, Eunomus, to attest the truth of what he has affirmed.25 The
paucity of letters is to be set against a background of relatively frequent references to
(and quotes in extenso from) decrees, laws, and other documents;26 if letters are
infrequent, no particular connotations are, however, attached to letter writing.
We reach at this point the second quarter of the fourth century, the period in
which, even according to the most conservative estimates, the orators begin to
make an ample use of documents. In this connection, Thomas has pointed to the
‘innovative’ use of written documents, mainly decrees and laws, by Aeschines and,
to a lesser degree, by Demosthenes.27 What of letters?

6.3. DEMOSTHENES

Besides the sixty-one extant speeches by Demosthenes (not all of them authentic),
fifty-six self-standing prologues and six letters survive. Both the authenticity of the
prologues and that of the letters have been disputed, but it is commonly accepted
that the first four letters are probably authentic, while doubts remain concerning
the other two. As already stated, only the references to letters in the speeches will
be discussed here; yet it is important to bear in mind this letter writing activity by
Demosthenes, because it is part of the more general development of epistolarity in
the fourth century.28

6.3.1. Private Speeches

Letters are treated rather differently in the forensic and in the political speeches.
A handful of references to letters occur in some of Demosthenes’ forensic
speeches, alongside references to decrees, laws, and wills. In this context, letters

25
Lys. 19. 23: KØ  ºH ’ ÆPfiH Œ ı H Ææa  F Ææe Åe I æ Ø KŒ ˚æø fi , . . . ‰
 ı ÆF’ K d IºÅŁB, ŒºØ Ø ¯h . <æı>.
26
Lysias quotes a number of decrees in his Against Agoratus (13. 22, 28, 35, 50, 55, 59, 71); another
decree in the speech Against Andocides (6. 24); three laws in On the Murder of Eratosthenes (1. 28, 30, and
31); a law in For the Soldier (9. 8); a high number of laws in Against Theomnestus 1 (10. 9, 14, 16, and 17,
18, 19), two of which are again referred to in Against Theomnestus 2 (11. 3 and 6: in this speech, the
emphasis is on the wording of the laws, hence the number of quotes); and laws in Against Alcibiades 1 (14.
5 and 8; the speech closes with a request to read, one after the other, the laws, the oaths, and the charge,
14. 47). Other documents are also read: thus, in the speech On the Property of Aristophanes, an inventory
of bronze plates (19. 28) and the record of public service of the father of the appellant (19. 58).
27
Thomas 1989: 69–72, 83–93. A complete list would take too much space: one example must
suffice. In Demosthenes’ Against Meidias, the following documents are read: five laws (8, 10, 47, 94,
113); seven witnesses’ statements (22, 82, 93, 107, 121, 168, 174); and an oracle (52–3). The question of
the authenticity of the documents cited is immaterial for our purpose; see MacDowell 1990: 43–7, who
concludes that although documents will not have been present in the original draft of the speech (they
are not included in the stichometric totals), they may have been collected in a separate dossier, as would
have been the case for the trial, and were inserted into the speeches by an editor at a later date. Thus the
authenticity of each document should be judged by itself.
28
The letters of Demosthenes have been thoroughly discussed by Goldstein 1968; see now also
Worthington 2006. All letters except the fifth are addressed to the Athenians, and introduced by the
greeting ˜Å ŁÅ B fi  ıºB fi ŒÆd fiH Å fiH åÆæØ (in the first letter, preceded by a short preamble).
The fifth letter is addressed to an individual, Heracleodorus, and is introduced by the greeting s
æØ. All letters close with the wish PıåE (or the singular PåØ for the fifth letter).
 The Legal and Political Stage 273

are used to impart information, or for recommendations.29 Such is the case of a

letter in the speech Against Polycles (transmitted in the corpus Demosthenicum,
but written by Apollodorus, and delivered at some point between 360 and 358 bc):
the speaker, who was at the time of the events serving as trierarch, mentions
sending a subordinate, a certain Euktemon, with money and letters to friends of
his father ([Dem.] 50. 18). Similarly, the speech Against Nicostratus, delivered
sometime after 368/367 bc, shows that Apollodorus probably used a letter (the
Greek has KØ ººø without further precision) to communicate with Nicostratus
when he had to leave in a hurry ([Dem.] 53. 5); and it is clear that Nicostratus, when
captured and sold as a slave in Aegina, used both letters and messengers to inform
his brother of what had happened and of his current situation ([Dem.] 53. 6).
But letters are also vehicles for intrigue: in the same Against Polycles, the speaker
recalls the arrival of a dispatch boat with a messenger and letters (50. 46) that turn
out to have been sent by Callistratus (the well-known Athenian politician from
Aphidna) to Timomachus (the strategos in charge). This testifies to the use of
written correspondence, although in this case letters might be expected, as Calli-
stratus was at that moment an exile. But the letter of Callistratus is a ‘secretive’ one,
part of a ploy, according to Apollodorus, to have Callistratus illegally transported on
board his trireme—when he learnt of the content of the letter from a sailor who had
overheard slaves talking, Apollodorus refused to continue on the mission.
Something more on the status of the letter as document can be gathered from
Demosthenes’ oration 38, Against Nausimachus and Xenopeithes, probably writ-
ten around 346 bc: there, the speaker stresses the unlikelihood that a written
message (the Greek has grammata, which can mean a letter, but is not specific)
may succeed in enforcing a request for payment:30
 Ø s oø Ø IŁæø ¼  , u Ł’ L  f Œıæ ı ØŒæ  Æ c ŒÆÆŁEÆØ
 F  åæ , ÆFÆ fiH c Œıæø
fi  łÆØ ªæ ÆŁ’ Œg I  FÆØ;
(Dem. 38. 12)
Now is there any human being so singular that, after having evaded payment to the
rightful owners for such a long time, he shall hand it over voluntarily to one not
entitled to it who demanded it by letter?
Finally, a few other speeches illustrate the milieu in which letters were mainly
used: that of the long-distance traders. Thus in the (possibly Demosthenic) speech
Against Phormio, dated 327/326 bc, the speaker mentions having given Phormio
letters for his commercial partners in the Bosporus:
When he went, then, to Bosporus, having letters (KØ  º) from me, which I had
given him to deliver to my slave, who was spending the winter there, and to a partner
of mine—having stated in the letter (ªæłÆ K Bfi KØ  ºBfi ) the sum which I had lent
and the security, and having instructed them, as soon as the goods should be unloaded,
to inspect them and keep an eye on them—this man did not deliver the letters
(KØ  º) which he had received from me, to ensure that they might know nothing
of what he was doing. (Dem. 34. 8)
The letters had been sent to warn the slave and the commercial partner of the
claimant of the need personally to check the actions of Phormio; but they were not

29
Mirhady 2000 discusses Demosthenes’ use of written documents in his private speeches, and
stresses the importance attributed by him to written proofs (be they from witnesses or others).
30
MacDowell 2004: 195.
274 Letter Writing and the Polis

delivered, nor did Phormio trouble himself with summoning witnesses for his
transactions: as the speaker points out, ‘All other men who borrow for a double
voyage, when they are setting off from their ports, take care to have many
witnesses present’ (Dem. 34. 28). Not so Phormio; the lack of witnesses, together
with his failure to deliver the letters, are repeatedly used in the speech to discredit
Phormio’s character:
You did not bring as a witness our slave who was in Bosporus nor our partner, nor did
you deliver to them the letters (a KØ  ºa) which we gave to you, and in which
were written instructions (K Æx  KªªæÆ ) that they should keep close watch on you
in whatever you might do! Well, judges, what crime is there that such a man is not
capable of doing, who, after receiving letters (ªæ ÆÆ ºÆg), did not deliver them
rightly and honestly? (Dem. 34. 28–9)
While the main point in the argument as a whole is the lack of witnesses in the
dealings between Phormio and the ship-owner Lampis, who according to the
speaker has become Phormio’s partner in fraud, it is interesting that the speaker
makes so much of the non-delivery of letters, whose very existence cannot be
proved: evidently, the speaker must have felt that this was a good rhetorical point.31
The use of letters by merchants is also a conspicuous feature of the speech
Against Dionysodorus, delivered sometime after 323/322 bc: Dionysodorus’ deal-
ings are contextualized within a description of the workings of the grain market, in
which retailers manipulate the price of the grain by sending letters to those in
Egypt, informing them of the fluctuations of the price and allowing them to decide
whether to ship the grain to Athens or elsewhere.32 This is what the speaker claims
happened, against the contract that had been agreed: Dionysodorus sent a man
with a letter to Rhodes (a ªæ ÆÆ a Ææa   ı I ƺÆ, 56. 10),
informing his partner that the price of the grain in Athens was low, with the
result that the merchants sold the grain in Rhodes.

6.3.2. Political Speeches

Two types of letter appear fairly often in the political orations of Demosthenes:
royal letters, most of them from Philip II, and the reports that strategoi or
ambassadors would send to the polis while away on a campaign or a mission.

31
Cf. Dem. 34. 32: ‘Yes, he says, for the agreement requested me to repay the money to the
shipowner. But it did not prevent you from summoning witnesses or delivering the letters!’  H ªaæ
ıªªæÆç , çÅ , fiH ÆıŒºæø fi KŒºı I  FÆØ e åæı  . æıæÆ  ª’ PŒ KŒºı
ÆæƌƺE, Pb a KØ  ºa I  FÆØ. Mercantile cases (dikai emporikai), such as this one,
followed a special procedure, whereby the ground for prosecution had to be the existence of a written
contract between the disputants that was claimed not to have been fulfilled: MacDowell 2004: 13. And
indeed, written contracts and other documents are mentioned in the speech: the contract is mentioned
at 34. 6, and the speaker asks for it to be read out at 34. 7; at this point, the entry of the customs officers
and the depositions are also read, as are, later in the speech, laws regulating maritime contracts.
32
[Dem.] 56. 8: ‘Then those at home would send letters to those abroad (   ªæ ÆÆ ƒ
KØÅ F  E I Å F Ø) advising them of the prevailing prices, so that if grain was expensive in
your market, they might bring it here, and if the price should fall, they might put in to some other port.
This is mainly why, judges, the price of grain went up, because of such letters and conspiracies (KΠH
 Ø ø KØ  ºH ŒÆd ıæªØH).’
 The Legal and Political Stage 275

We learn from On the Corrupt Embassy that such reports were discussed collect-
ively, for according to Demosthenes his colleagues in the embassy voted against
sending a dispatch he had written and chose to write a different one, which they
sent to Athens.33 Thus, these are letters, but by no means private, exclusive
communications. Examples of generals’ reports in Demosthenes are to be found
in the speech Against Aristocrates (352 bc): letters from the strategoi Iphicrates
and Timotheus to the Athenians are read as evidence against Charidemus
(23. 151–2), together with a decree and a witness’ statement;34 in the same speech,
a (deceptive) letter of Charidemus himself to Athens—or rather, as Demosthenes
specifies, to his friend Cephisodotus, but meant for the Athenians—is also brought
forward as evidence (23. 153–6), as well as letters from Athenian magistrates of
cities in the Thracian Chersonese (23. 156–62).35 The practice is attested in other
types of literary source (Nicias’ letter in Thucydides, the reports mentioned by
Xenophon in connection with the battle of Arginusae), but also in honorific
inscriptions. Clearly, from the fourth century onwards letter writing was inte-
grated into the practices of the Athenian polis, at least for that which concerned
internal communication (i.e. between Athenians).36
But the bulk of references to letters in the corpus Demosthenicum concerns
royal letters. This is not surprising, as letters are the usual means by which
monarchs communicate their decisions to functionaries and cities within their
kingdom, and also beyond it.37 Philip II of Macedon, Alexander, and after them
the Hellenistic kings made an ample use of letters; but is interesting to find that
the Thracian kings already had been communicating in an official context
through letters. Thus, in his Against Aristocrates, Demosthenes mentions two
letters of Cotys to the Athenians, one written when he was in difficulty (probably
around 362 bc), the other when he felt secure, in late 361–early 360 bc, and
contrasts them sharply: ‘I will read to you the letter that Cotys sent at the time
of Miltocythes’ revolt, and also the one which, when the kingdom was fully in
his power, he sent to Timomachus before seizing your outposts. Letters’

33
Dem. 19. 174: c b ªæÆçE Æ KØ  ºc ’ K F æe  A IłÅç Æ c  Ø, ÆP d
’ P’ ›Ø F ªØb ªæłÆ  łÆ.
34
Dem. 23. 151: ‘read the decree concerning the hostages, the letter of Iphicrates, that of Timotheus,
and lastly this deposition’. Charidemus was a mercenary from Euboea who changed sides a number of
times between 367 and 357, first fighting under Iphicrates for the Athenians, then joining Cotys, then
accepting the proposal of Timotheus and moving to the Athenian side, then from 362 bc joining first
the rebel satraps of Asia Minor, then the Athenians for a short while, and then again Cotys and his
young son Cersobleptes. When Chares in 357 recovered the Chersonese for the Athenians, Charide-
mus’ supporters managed to present this as a result of his efforts, so that he received a crown, against
the opposition of Demosthenes. On the events and specifically on these letters see Heskel 1997: 45–6.
35
Discussion of the chronology of the events and of the letters in Heskel 1997: 54–9.
36
Thuc. 7. 10–15; Xen. Hell. 1. 1. 23 and 3. 3. 8 (Sparta); Xen. 1. 7. 4 and 1. 7. 17 (Athens,
Arginusae); see above, ch. 4. Mentions of letters in decrees: see above, n. 7.
37
See Welles (1934); Bertrand (1985); Lewis (1996); Ceccarelli (2005a). Philip, in the letter preserved
in the corpus Demosthenicum, states more than once that he sent letters: e.g. at the opening of the speech,
in 12. 2, where he also complains of the violation of his privacy: ‘when Nicias, my herald, was kidnapped
from my territory, you not only failed to bring the law-breakers to justice, but you kept the victim a
prisoner for ten months, and the letters from me, of which he was the bearer, you read before your
Assembly (L ' çæ Ææ'  H KØ  º, Iªø' Kd  F  Æ )’ (trans. J. H. Vince). Further
letters are mentioned in 12. 22 (‘For I wrote to you again and again on the subject [Amphipolis]’).
276 Letter Writing and the Polis

(Dem. 23. 115).38 Other letters by Cersobleptes and the Thracian kings are read
out to the assembly to show the unreliability of Charidemus (23. 178). Just as is the
case at times with other documents used by the orators, embedded letters often
carry a meaning that goes beyond the literal import of the letter itself and are used
to give strength to the argument.39
But most of the references to letters in the political speeches of Demosthenes
concern Philip. Sometimes, the text of the letter is read (in a few cases, the text of
short letters, whose authenticity is uncertain, has been incorporated into the
speeches); one self-standing, longer letter of Philip has also been transmitted in
the corpus Demosthenicum alongside the Reply to Philip.40 Of course, the number
of references to letters of Philip is much higher than the number of letters actually
sent, since Demosthenes and Aeschines more than once refer to the same letters:41
as we shall see, these references are exploited by Demosthenes as part of a
rhetorical strategy; they function as weapons of psychological persuasion.
The First Philippic is a case in point: imagery connected with letter and decree
writing plays a central role in the speech (delivered in 352/351 bc). According to
Rowe, Demosthenes here pursues an ‘inversion strategy’, and builds a mundus
perversus out of the opposition between action on the one hand and hearing,
talking, writing, and sitting on the other, an opposition that becomes stronger as
the speech progresses.42 Demosthenes opens the speech with an exhortation to the
Athenians and a first proposal for action; he then continues:
æe b  ø Æ  Ø’, t ¼æ ŁÅÆE Ø, çÅ d æ åØæ Æ ŁÆØ E  A, m
ıåH  º  Ø ŒÆd ŒÆŒH KŒE   Ø Ø.  Ø ıæ ı Åb Ø ıæ ı  ı,
Åb a KØ  ºØ Æ ı ÆÆ ı Ø, Iºº’ m B ºø  ÆØ, Œi  E Æ Œi
º ı Œi e EÆ Œi ›Ø F åØæ   Å æÆŪ,  ø fi  ÆØ ŒÆd
IŒ º ıŁ Ø. (Dem. 4. 19)
But in addition to this, Athenians, I propose that you should keep in hand some force
that will constantly make war and annoy him. None of your ten thousand or twice ten
thousand mercenaries, none of your ‘epistolary forces’, but one that shall be the state’s;
and one which, whether you appoint one or more generals, or this or that man or any
other, shall obey and follow him. (Trans. Rowe modified)
The epistolimaioi dunameis are imaginary forces on paper, such as may have been
often mentioned in letters to and from generals, and seen as typical of the
Athenian democracy; their being couched on ‘letter paper’ (rather than decree
paper) and contrasted with the city’s real forces may add one more element to the

38
The Athenian general Timomachus failed in the war against Cotys in 361/360 bc and chose to go
into exile rather than being condemned. See on the events and for the date of the letters Heskel 1997:
81–3 and 88–91.
39
Carey (1994) explores the role played by written documents (not letters though), and the
possibilities for manipulation they afford.
40
The authenticity of Philip’s letter ([Dem]. Or. 12), whose dramatic date is 339 bc, has been
disputed, with some (among whom Jacoby) attributing it to Anaximenes; contra, most recently Gauger
2000: 73 (‘wohl echt’) and 306 n. 20. The Reply (Or. 11, probably not Demosthenic) answers another
letter of Philip; see above 166–7 and n. 185, and on the historical context, Harris 1995: 75–6.
41
List of the actual letters of Philip in Gauger 2000: 45; see also above.
42
Rowe 1968; see also Mader 2006 for the contrast psephisma/action. I single out those elements of
the ‘topsy-turvy’ Athenian world created by Demosthenes that pertain to letters.
 The Legal and Political Stage 277

inversion picture, as letters in public contexts are mostly the prerogative of kings.
In the next paragraph, Demosthenes pursues his theme by remarking that
’ Kºø  Ç  r ÆØ  F   , ŒÆd a ªØ ’ K  E łÅç Æ Ø
ƃæ   Ø, Kd fiH æØ Pb a ØŒæa  ØE. (Dem. 4. 20)
thinking everything insufficient, you adopt the strongest measures in decrees; but
when it comes to action you fail to do the slightest.
Letters and decrees, and their counterpart, action, are brought together ten para-
graphs later, when Demosthenes exhorts the Athenians to vote for his proposal:
KØa ’ KØåØæ  B a ª Æ, i  E Iæ ŒfiÅ, åØæ   , ¥ Æ c  
K  E łÅç Æ Ø ŒÆd ÆE KØ  ºÆE  º B غø fi , Iººa ŒÆd  E æª Ø.
(Dem. 4. 30)
When you sanction the proposals, if you approve them, you will vote so that you may
fight Philip not only with decrees and dispatches, but with deeds also.
It is worth noting that here letters are on the side of inaction and of the polis
(Athens), rather than on the side of the king; in Demosthenes’ presentation, only
when they are royal do letters correspond to action. Key terms such as ‘decrees’,
‘letters’, ‘hopes’ are repeated throughout the speech, in combination with other
key terms such as ‘Philip, ‘triremes’, ‘forces’, and with the notions of enforcement,
or conversely of emptiness and inaction;43 ‘straight’, tyrannical letters surface as
well, with the mention of hybristic KØ  º (‘letters/orders’) sent by Philip to the
Euboeans (‘Meanwhile Philip has the effrontery to send such letters as these to the
Euboeans’, ›  N  FŁ oæø KººıŁ u  KØ ººØ ¯P F Ø XÅ  ØÆÆ
KØ  º, Dem. 4. 37–8). The letter is read; Demosthenes wryly comments: ‘Most
of what has been read, Athenians, is unfortunately true—possibly, however, not
pleasant to listen to.’44
Similarly, there are numerous references to two letters of Philip in the pseudo-
Demosthenic De Halonneso (342 bc). The speech was written as an answer to a
letter by Philip proposing a settlement through arbitration of the island’s owner-
ship, so references to this letter and to an earlier one, ‘still to be seen in the
bouleuterion’ (l K Ø F K fiH  ıºıÅæø fi ), are not surprising.45 Particularly
interesting is the beginning of the speech, because there, for a moment in the
opening sentence, the ‘letter form’ chosen by Philip is pitted against the free
discussion of the assembly:
Øe ªaæ i YÅ, N c Kd  F  Æ  ÆææÅ Æ ƃ Ææ’ KŒ ı   ÆØ
KØ  ºÆd Iº Ø (Dem. 7. 1)
For it would be monstrous if the letters sent by him were to destroy the freedom of
utterance of this platform.

43
As in 4. 44–5; see Rowe 1968: 365–7. Mader (2006) traces this strategy in all Philippic speeches.
44
The manuscripts indicate the reading of only one letter (singular), against the plural in Demos-
thenes’ reference. Could the discrepancy be interpreted in the sense of a conscious ‘tragic’ choice on the
part of Demosthenes of the plural KØ  º (the term is always found in the plural in tragedy)? This
would fit well in Rowe’s overall argument (1968).
45
Passages in which reference is made to letters are: 7. 1; 7. 18–21; 7. 27 and 7. 29; 7. 32; 7. 33–4 (the
letter in the bouleuterion); 7. 41; 7. 45; and 7. 46.
278 Letter Writing and the Polis

The same picture emerges from the speech On the Corrupt Embassy (343 bc).
A letter of Philip, allegedly composed with the help of Aeschines, is here the centre
of attention: reference to it is made for the first time in 19. 36 (where the letter is
said to be nothing more than an I º ªÆ ªªæÆ Å—an interesting iunctura);
the letter is read out at 19. 38. We find in it a positive evaluation of writing—or at
least, a shrewd appreciation of the respective merits of written and oral reports: for
according to Demosthenes, in the letter Philip took upon himself responsibility
for the failure of the ambassadors to do what they had been charged to do (19. 37),
a gesture for which there was anyway no danger that he would be punished.
But all the matters in which he was trying to deceive and overreach the city were left
for Aeschines to report by word of mouth ( y  Iªªغ), so that you might never
be able to incriminate or blame Philip, as those statements were not in the letter or in
any other direct communication of his ( ’ K KØ  ºB fi ’ ¼ºº ŁØ ÅÆ F H
Ææ’ KŒ ı  ø Kø). (Dem. 19. 38)46
Writing—in this case, written letters—is presented as a powerful means of
associating people with their statements. Demosthenes follows this up with the
reading of excerpts from an earlier letter of Philip (19. 40), from which it appears
that Philip had committed himself to set down in writing what great benefits he
would confer to the polis once a peace was reached and if there should be an
alliance.47 Clearly Demosthenes is here reading more into Philip’s letter than there
ever was, and is manipulating the writing with his own words; clearly also,
whatever the power of written words, Philip could and would back out of anything
he might not have found convenient. Whether Philip was indeed in that instance
deliberately playing against each other the two forms of communication, the oral
report of the ambassadors and the letter, as Demosthenes states (19. 37: ‘and this
was not arranged in this way by accident’, Ie ÆP  ı), or not is a moot
point: Philip, of course, also sent ambassadors to Athens (in Elaphebolion 346 bc,
before the peace of Philocrates was ratified, and again in 343 bc, when Pytho of
Byzantium and other ambassadors from the allies came to announce his willing-
ness to revise the terms of the peace of Philocrates), and he could remain evasive
in his letters as much as through his ambassadors.48 What is important is that
Demosthenes could present as plausible such an interpretation. Moreover, playing
off embassies against letters offered a welcome opportunity for a change in
register. While embassies could only be ‘remembered’ by the orator, the reading
out of a letter is a dramatic moment, a pause in the speech, invested with a sense of
the presence of the sender—if only because the document is read by a person
different from the orator, a clerk.

46
Dem. 19. 37-8: L ’ KŒE  (Philip) KÆÆB ÆØ ŒÆd æ ºÆE B ºø K º , y 
(Aeschines) Iªªغ, ¥ Æ Å’ KªŒÆº ÆØ Åb  łÆ ŁÆØ Åb o æ   E å Ø غø fi ,
’ K KØ  ºBfi ’ ¼ºº ŁØ ÅÆ F H Ææ’ KŒ ı  ø Kø. ºª ’ ÆP E ÆPc c
KØ  º, m ªæÆł b y ,  ł ’ KŒE · ŒÆd Œ EŁ’ ‹Ø  F  åØ e æ , n
غºıŁ’ Kª.
47
19. 41: æd b NæÅ ıåE, N ŒÆd ı ÆåÆ æ ª Ø’ ÆPfiH, ªæłØ ‰ ºªØ ºŒÆ c
ºØ s  Ø Ø· KØc ’ I çæ’ ÆPfiH ªª , PŒ NÆØ çÅ d  i  ØH åÆæ ÆØ . On the vague
character of these promises see Harris 1995: 64–5; on the historical background of the whole affair,
Harris 1995, passim.
48
On the embassy of 343, see Harris 1995: 110–12.
 The Legal and Political Stage 279

Slightly later in the same speech (19. 44–5), the first letter, but also, tellingly, the
words of Philip’s ambassadors (KŒ  F c  f Ææa  F غ ı æ Ø ÆFÆ
ºªØ Åb c KØ  ºc c غ ı) are contrasted with the words of
Aeschines, and the latter is found to be a liar. Two further letters of Philip are
also read out in the Assembly, inviting the Athenians to join him in his campaign
in Phocis: although that was not, Demosthenes says, Philip’s real intention, the
ambassadors played along with him (19. 51–2).49 Letters of Philip are mentioned
or read out and discussed along these lines quite a few other times in On the
Corrupt Embassy (for instance, in 19. 187 they are mentioned, again ironically, in
parallel with the æåØæ  ºª , the ‘easy sentence’ of those who affirm that
Philip is in fact a benefactor).50 It is in a sense the very number of references to
letters of Philip that support Demosthenes’ claim of his heavy-handed interven-
tion in Athenian (and Greek) politics, by showing how forceful his presence is in
Athenian public affairs.
Letters from Philip are also mentioned in On the Chersonese (341 bc) and in the
Third Philippic (341 bc).51 In the latter, Philip’s writings (and actions) are again
contrasted with the inactivity of the Greeks, who ‘see and hear all this, and yet
[we] do not send embassies to one another and express [our] indignation’: as will
become clearer in a moment, the weapons of the tyrant (letters and war) are being
pitted against the weapons of the city-states (embassies, decrees, and assembly-
speeches). This theme is picked up in other speeches: thus, the contrast between
the inactivity of the Athenians and the energy of Philip, including his letter
writing, reappears in the Reply to the Letter of Philip.52
Five letters from Philip are read out (and their texts included) in On the Crown
(a speech held in 330 bc).53 Of these, two are particularly interesting, inasmuch as
through their position they highlight the fact that the decree is the standard
method used by the Greek cities for interstate communication. These letters,
one directed to Athens and the second to Thebes, are Philip’s response to two
decrees of the Athenians (18. 164–7). Later in the speech, other letters of Philip
were read to the public (18. 218 and 221); again, interestingly, Demosthenes
presents this latest outburst of epistolary activity on the part of Philip as a result
of the many decrees moved by himself:
ººa c ¥Æ  IçØ çøa › ºØ  ŒÆd K ¥ÆØ q ÆæÆåÆE Kd   Ø, KŒ
H KØ  ºH H KŒ ı ÆŁ  Ł z N —º Å   . ŒÆ Ø ºª
ÆÆ ºÆ, ¥ ’ NB,  K c ıåØÆ ŒÆd º Ø ŒÆd ƺÆØøæÆØ ŒÆd a  ººa

49
See the comments of MacDowell 2000: 228–9.
50
See also 19. 161, where a letter from Philip is ‘sandwiched’ between three decrees, with the
purpose of showing that the embassy did not accomplish what it had been asked to do; and 19. 316–17,
where, at the close of the oration, Demosthenes returns to the complicity of Aeschines in writing
Philip’s letter.
51
Dem. 8. 16; and Dem. 9. 27.
52
Dem. 11. 17: ‘he always takes a personal share in the hardships and dangers of the campaign,
never neglects a chance, never wastes any season of the year; while we—for the truth must out—sit here
idle; we are always hanging back and passing resolutions and haunting the market-place to learn the
latest news. Yet what more startling news could there be than that a Macedonian should insult
Athenians, daring to send us such a letter as you have heard read a moment ago?’ (trans. J. H. Vince).
53
Dem. 18. 39; 18. 73 (mention of letter) and 77–8 (actual letter); 18. 156–7; 18. 166 and 167 (to the
Thebans).
280 Letter Writing and the Polis
łÅç ÆŁ L F y  Ø ıæ,  IÅæª Æ . . . 221 ¸ª a KØ  ºa a  F
غ ı.
But indeed, what sort of language Philip gave vent to at that time, and what sort of
confusion he was in through these events, you shall learn from letters by him, which he
kept sending to the Peloponnese. Take them and read them, so that you may see what
my perseverance, my travels and hardships, and the many decrees, at which Aeschines
was just now scoffing, what they achieved . . . Read the letters of Philip.
We have come almost full circle from the picture drawn in the First Philippic;
democratic decrees are now the reason for royal discomfort, and for epistolary
royal communication. What comes out clearly from this is the role played by
rhetorical construction in the use of these ‘documents’.

6.4. AESCHINES

Two of the three extant speeches by Aeschines present frequent references to

letters and letter writing—not accidentally, they are the two that relate more
directly to Demosthenes. The orator opens his speech On the Embassy (delivered
in 343 bc as an answer to Demosthenes’ attack), with a summary of what he had
said to Philip at the time of the first embassy (early in 346 bc); this he ends by
stating that he then produced for Philip as proof of what he was saying the letters
of the persons named, the decrees of the demos, and the text of the truce
concluded by Callisthenes with Perdiccas in 363 bc.54 To what letters exactly
Aeschines is referring to in this context is unclear; they must be relatively old
documents, going back some twenty years in time, involving Macedonian kings
such as Amyntas and members of the Macedonian royal house (Eurydice), but
also possibly Athenian generals such as Iphicrates. In this case, Aeschines could
have found the letters, together with the decrees and the text of the truce, in the
Metroon, where the central archive was housed; if these documents had been used
to impress Philip then, their mention now has the purpose of showing to the
Assembly how carefully and comprehensively he has always managed matters.
Aeschines then goes on to recount how, on their return to Athens, the embassy
gave a short oral report to the boule, and handed over a letter sent by Philip (2. 45);
and how, when the embassy gave a report to the Assembly, this same letter was read
at the request of Demosthenes, together with the decree on the basis of which the
embassy had been sent (2. 49–50). Later in the speech, when discussion has moved
to the second embassy (late spring 346), Aeschines has a clerk read a letter sent by
Chares to the Athenians, informing them that Cersobleptes had lost his kingdom
and that Philip had taken Hieron Oros.55 This is again an instance of the use of
letters for internal communication between strategoi on campaign and the polis.

54
Aeschin. 2. 31: ŒÆd ø z Y Ø Ø æıæÆ a KŒø KØ  ºa ÆæØå Å ŒÆd a
łÅç ÆÆ  F  ı ŒÆd a ˚ƺºØ Ł ı I å.
55
Aeschin. 2. 90: IŒ  Æ c B æÅ  KØ  ºB, m K غ  fiH  fi ø, ‹Ø
˚æ ºÅ I ººŒ c Iæåc ŒÆd  Iæe Zæ  ŒÆºÅç ºØ  ¯ºÆçÅ ºØH  Åe  fi Å
çŁ  · ˜Å ŁÅ ’ K fiH  fi ø æ æı   ı  F Å, x  J H æ ø, ŒfiÅ
çŁ  . ¯Ø  º.
 The Legal and Political Stage 281

The letter, however, is used not for its content, but in order to mount a chrono-
logical argument: the fact that the information contained in the letter reached
Athens on 25 Elaphebolion (end of March) serves to exculpate Aeschines from the
accusation of having delayed sending help to Cersobleptes: the boule decreed that
the ambassadors should set out to receive the oaths on 3 Mounichion (April), well
after the fall of Cersobleptes. (Incidentally, we learn that dispatches could travel at
a remarkable speed: the letter announced that Hieron Oros was taken on 24
Elaphebolion; it reached Athens on the next day!)
Letters crop up again a few paragraphs later in this same speech: Aeschines now
defends himself against the accusation of having written for Philip the letter that
the latter sent to Athens (2. 124–31). Again, it is not the content that is—for the
moment—at issue, but the fact that Aeschines supposedly helped Philip in writing
it. This accusation gives Aeschines the opportunity to mention quite a few persons
(Leosthenes, Philip himself, and Pytho of Byzantion) who would have been
perfectly capable of writing such a letter without any need for his being involved
in the affair himself. Interestingly, the ability at letter writing is linked to an ability
in public speaking: so in the case of Leosthenes, who ‘would not have been
competent to write a clever letter—a man whom some do not hesitate to rank
next to Callistratus of Aphidna as an able orator’ (› b ªaæ ¸ø ŁÅ . . . P
ıÆe q KØø KØ  ºc ªæłÆØ, n PŒ OŒ F  Ø I çÆ ŁÆØ a
˚ƺº æÆ  e çØÆE  H ¼ººø ºØ Æ NE Æ ŁÆØ, 2. 124), so in
the case of Philip, ‘against whom Demosthenes was not able to hold his own when
he tried to speak on your behalf ’ (2. 125). The only one of the group whose ability
in writing letters is attested by writing expertise, rather than ability in speaking, is
the otherwise little-known Pytho, ‘a man who takes pride in his ability as a writer’
(¼Łæø  Kd fiH ªæçØ ªÆ çæ H, 2. 125); there may be a slightly negative
connotation in this remark—the man was a student of Isocrates, hence the
connection with writing rather than speech, but he had also gone over to Mace-
don.56 The letter in question is cited a few paragraphs later—and if we are to trust
Aeschines’ summary of it, it must have been a fairly plain and matter-of-fact one.
But the rhetorical question asked of the audience is informative, ‘Does it not seem
to you that Philip would have been able to write that in the daytime, and without
my help?’ (2. 129), as it implies that the writing of a royal letter could be a simple,
banal thing, not requiring any (dangerous) sophistication.57
Moving to a different accusation, Aeschines appeals to composite evidence to
prove his contention that the Phocians had turned towards Philip even before he
was elected ambassador: his evidence comprises the testimony of the two heralds
of the mysteries, a letter of the Athenian general Proxenos, and the testimony of

56
Pytho was at Pella in 346 and may have played a role in the negotiations that led to the Peace of
Philocrates; he certainly came to Athens in 343 as a representative of Philip to renegotiate the peace of
Philocrates (see RE, s.v. ‘Python 4’). On this passage, cf. Harris 1995: 10; Carey 2000: 135 and 139.
57
Aeschin. 2. 128–9: ‘Take this letter (KØ  ºc), which Philip sent. It clearly must be full of
deception for the city, this letter we spent the entire night writing (Mªæı F  ªæç ). Letter. You
hear, gentlemen, that he says “I gave my oath to your ambassadors”, and has written (ªªæÆç) the
names of those of his allies who were present, both the names of the representatives themselves and of
their states; and he says he will send to you (I ºE çÅ Ø) those of his allies who were not there in
time. Does it not seem to you that Philip would have been able to write (ªæłÆØ) that in the daytime
( Ł’  æÆ), and without my help?’
282 Letter Writing and the Polis

the envoys sent by Proxenos to the Phocians.58 The wording of Aeschines’

comments makes it fairly clear that the letter and the witnesses serve two different
purposes:
‘To prove that I am speaking the truth, call the heralds of the truce, and the envoys
sent by our general Proxenus to the Phocians, Callicrates and Metagenes, and listen to
the letter of Proxenus. Testimony. Letter. You hear, men of Athens, as the dates taken
from the public archives are read and compared, and you hear the witnesses, who
further testify that before I was elected ambassador, Phalaecus the Phocian tyrant
distrusted us and the Lacedaemonians as well, but trusted Philip.’
(Aeschin. 2. 134–5)
The letter is referred to only as evidence for the date, while the witnesses attest to
the main point at issue. A further letter of Philip, in which he calls the Athenians
to join against the Thebans, and which is connected with an incident later on
in Skirophorion in 346 bc, is mentioned in 2. 137. In this speech, thus, letters
appear to figure mainly as elements of the case, and not as rhetorical weapons:
Aeschines carefully builds his defence around dates and facts that he extracts from
documents.59
The other speech in which letters figure, alongside other documents, is the
Against Ctesiphon, delivered in 330 bc. But the way they are used here is very
different from On the Corrupt Embassy. Twice, letters are mentioned in connec-
tion with the Persian king: they function as markers of his insolence, and in both
instances they are used to show how much the balance of power has changed. The
first passage in which letters appear defines the power of Xerxes in terms of
digging the channel through Mt Athos, bridging the Hellespont, and daring to
write in his letters ‘that he was lord of all men from the rising of the sun to its
setting’; his power is contrasted with the actual situation of the Great King. The
second reference functions in a similar way: the apparent power of the King (now
Artaxerxes III) is encapsulated in his writing ‘a most insolent and barbarous
letter, in which everything was expressed in the most ill-mannered terms’, and
in which the king refused to give any gold to the Athenians.60 The reversal is made
manifest by the initiative of that same king, when in need of allies, to send Athens
300 talents, which the city wisely did not accept (but part of which, Aeschines
continues, Demosthenes embezzled).
Letters are also used to subtly incriminate, or at least discredit, Demosthenes.
Thus, Aeschines suggests that Demosthenes sent a letter to Alexander (ªæ ÆÆ
 łÆ, 3. 162) through a young, beautiful boy, and thus reconciled himself with
the king of Macedon. The implication is that both instruments of communication

58
On the historical background of the affair see Harris 1995: 54; Carey 2000: 139.
59
See further Thomas 1989: 69–71; Bertrand 1999: 198–200. Demosthenes’ scathing mention of
Aeschines’ activity as  ªæÆ Æ (Dem. 19. 200), or the insult ªæÆ Æ Œçø (‘someone who is
hunched over records’) hurled at him (Dem. 18. 209) show that Aeschines was perceived as being
particularly familiar with written records, but also that such a familiarity could be construed negatively.
60
Aeschin. 3. 132: På › b H —æ H Æ Øº, › e @Łø Ø æÆ, › e  Eºº    ÇÆ, ›
ªB ŒÆd oøæ  f  ‚ººÅÆ ÆNH, ›  º H K ÆE KØ  ºÆE ªæçØ, ‹Ø  Å K d ±ø
IŁæø Iç’ º ı IØ  åæØ ı  ı, F P æd  F ŒæØ  æø r ÆØ ØƪøÇÆØ, Iºº’ XÅ
æd B  F  Æ  øÅæÆ; and 3. 238: ŒÆ ł fiH  fi ø ŒÆd ºÆ æØ ØŒc ŒÆd æÆæ 
KØ  ºc Kfi w   c ¼ººÆ ŒÆd º’ IÆØø غåŁÅ.
 The Legal and Political Stage 283

(the letter and the young, seductive messenger) are disreputable. More import-
antly, letters are symbolic of double-dealing, of hidden contacts: a few paragraphs
later, Aeschines affirms that before the battle of Issus, when the situation was
unclear and looking not too favourable for Alexander, ‘there was not room enough
in the city to contain your odious demonstrations and the letters that you carried
around, dangling them from your fingers’ (c b c IÅÆ  ºØ PŒ KåæØ ŒÆd
a KØ  ºa L KÅæÅ   KŒ H ÆŒºø æØfiØ, 3. 164), although even
then Demosthenes did not dare to take any action. The image is brilliant: Demos-
thenes is showing off his letters (i.e. his contacts) to the city; but these letters do not
have any referent, they are not mentioned in any other part of the speech; the only
purpose of this reference is to show Demosthenes glorying in the secret, private
contact with a number of similarly private, powerful, and secretive persons.
This image must have struck a deep chord: the connection between Demos-
thenes and unsavoury letters will reappear in the speech that Dinarchus wrote
against Demosthenes for the trial of 323 bc; here, however, the letters are
mentioned in connection with events that took place slightly later (the revolt
against Macedonian control coordinated by King Agis III of Sparta, between 333
and 331 bc):
K   Ø  E ŒÆØæ E ˜Å ŁÅ  q, ›  F ı  ıºF ÆØ ŒÆd ªæłÆØ ŒæØ  ŒÆd
ç ø ÆPŒÆ c Ø E a ŒÆŁ HÆ æª ÆÆ; KH ªaæ  f ¼ºº ı ŒØ ı.
ªæÆł Ø æd  ø ŒØø; ı ºı Æ; KæØ Æ åæ ÆÆ; ØŒæ Ø
åæ Ø  Kª ı  E bæ B Œ ØB øÅæÆ æ ı Ø; P’ ›Ø F, Iººa æØfiØ
ŒÆÆ ŒıÇø º ª  Ø , ŒÆd Ææ’ ÆfiH ªæçø KØ  º, ŒÆd ŒÆÆØ åø c B
ºø Æ, KŒ H ÆŒºø IÆł   æØ æ , æıçH K  E B ºø
ŒÆŒ E, ŒÆd Kd ç æ ı ŒÆÆŒ ØÇ   c N —ØæÆØA ›, ŒÆd a H ø
I æÆ OØÇø. (Din. 1. 35–6)
In that hour what was the behaviour of this Demosthenes who has the power to give
advice and make proposals, and who will shortly tell you that he hates our present
circumstances? For I shall pass by the other crises. On these matters, Demosthenes, did
you offer any proposal, any advice? Did you contribute money? Were you of the smallest
value to those working for our safety? Not the least; you went round suborning speech-
writers. Writing a letter at home, and dishonouring the reputation of the city, he walked
about dangling it from his fingers, living in luxury during the city’s misfortunes,
travelling down the road to the Piraeus in a litter and reproaching the needy for their
poverty.
The similarity in the wording makes it clear that Dinarchus here had Aeschines
in mind; again, there is much that is unclear or simply allusive slander in this
passage. Particularly puzzling is the ‘letter at home’ allegedly written by Demos-
thenes: Dinarchus seems to be accusing Demosthenes of moving frequently and in
an unnecessarily luxurious way between his houses in the Piraeus and his house in
the city, and of doing this while holding in a conspicuous way a letter written by
himself, so as to give the impression of being in touch with events and powerful
persons much more than was actually the case.61 It may not be an accident that

61
The connection between Din. 1. 36 and Aeschines’ description has always been noticed
(e.g. Worthington 1999: 153). But no commentator addresses the question of the ‘letter home, defiling
the city’s honour’. The emotional (rather than technical) import of the passage, and for that matter of
the entire speech, is well brought out by Nouhaud 1990: p. xii, who points out that it was a deuterologia
284 Letter Writing and the Polis

the only other (rather cryptic) reference to letters in what we have of Dinarchus—
admittedly not much—also comes from his Against Demosthenes: in 1. 27 the
orator asks a clerk to read a Theban decree, to cite the evidence, and to read
the letters. It is impossible to know what this refers to; but Nouhaud’s suggestion
that some epistolary exchange between Demosthenes and the Thebans is meant is
plausible.62
The image of epistolary exchanges taking place in the last days of Athenian
independence will shape the later tradition. Plutarch, Demosthenes 20. 4–5,
mentions letters sent by the king of Persia to his satraps ordering them to supply
Demosthenes with money, something said to have afterwards come to Alexan-
der’s knowledge ‘by certain letters of Demosthenes which he found at Sardis’, and
by other papers of the Persian officers, which accounted for the large sums that
had been given him.63 Plutarch also claims (Dem. 23. 2) that after the death of
Philip in 335 bc, the orator sent letters to the Persian officers of the King, inviting
them to attack Alexander. More interestingly, when pursued by the Macedonians
to the temple of Poseidon at Calauria, Demosthenes obtained from Archias (his
pursuer) some time to ‘write a letter home’ ( ØŒæe s K å, ‹ø KØ ºø Ø
 E YŒ Ø, Plut. Dem. 29. 3); according to the biographer Satyrus, he would then
have bitten his pen, which had been treated with poison. Now according to the
version of the otherwise unknown Pappus, whose work was used by Hermippus
(as cited in Plut. Dem. 30. 1), when Demosthenes fell by the altar because of the
poison, a scroll was found, with the beginning of a letter, ‘Demosthenes to
Antipater,’ and nothing more—would this have been the letter home?64 Fittingly,
vengeance for Demosthenes overtook Demades in the form of a letter of his
addressed to Perdiccas, in which he had suggested that Perdiccas should deliver
the Greeks from Antipater; when found, the letter excited the anger of Cassander
(Antipater’s son), who had Demades (and his son Demeas) executed (Plut. Dem.
31. 3).65 The epistolary activity ascribed by the ancients to both Demosthenes and
Aeschines fits the picture extremely well (as already mentioned, six letters of
Demosthenes have been preserved; twelve letters have survived under Aeschines’
name, none of them authentic).66

(a second, supporting speech), and as such need not argue a case, but rather excite the anger of the
citizens. For a reconstruction of this period see e.g. Cawkwell 1969.
62
Nouhaud 1990: 58, followed by Worthington 1999: 148.
63
A piece of information known from this passage only.
64
Further references and ample discussion in Bollansée 2002 (FGrH (Continued) 1026 F 51; while
discussing the source of the story, Bollansée could, however, also have mentioned Jacoby’s belief that
Hermippus got the story from a Macedonian archive, the report made by Archias to Antipater: Jacoby
FGrH 3b (Supplement) i: Text, 540–1 (ad Philochorus 328 F 164).
65
In Diodorus 18. 48. 2 ‘letters’ (in the plural) by Demades to Perdiccas are found in the royal archives
by Antipater (not Cassander); Plutarch, Phocion 30. 9–10 has a slightly different version, in which the
addressee of Demades’ letter is Antigonus. Details of the affair, with full bibliography, in Paschidis 2008:
44–6 (the letter would have been found by Hieronymus of Cardia, who is undoubtedly the source from
which all others derive, either in the archive of Antigonus Gonatas, or in that of Perdiccas; Hieronymus
himself might have been the person who informed Antipater of the treasonous letter).
66
On Demosthenes’ letters see above, 272. Nine of the twelve letters ascribed to Aeschines were
known to Photius (Bibl. 61) as ‘the Nine Muses’ (the speeches were ‘the three Graces’), and may have
been composed in Rhodes in the mid-second century bc; the others were added later. See Martin and
Budé 1928: 121–2, with Goldstein 1968: 265–6.
 The Legal and Political Stage 285

To go back to Aeschines: the next mention of letters concerns those allegedly

forged by Demosthenes to gain advantage over Aeschines:
‘I say nothing of forged letters ( ¯Ø  ºa b تH łıE) and the arrest of spies, and
torture based on groundless charges, that I with certain persons was planning a
revolution.’ (3. 225)
Indeed, Aeschines does not say anything else about these forged letters, and we
may suppose that the Assembly did not know more about them than we do; they
figure here simply to complete the picture of a man using all means at his disposal
to reach his goal. Obviously, none of these documents is—or can be—read as
evidence.
Forged letters had appeared at least already once in political Attic oratory. In
discussing the rhetorical figure of parasiopesis, whereby a speaker affirms to pass
over something while in fact stressing it, the rhetorician Rutilius Rufus gives the
following representative example, taken from a speech of the orator Lycurcus:
in praesentia, iudices, iniusso populi quae improbissime gesserit, reticebo, de falsis eius
litteris, quas ad senatum miserit, nihil dicam, quae illi saepe interminati sitis, omittam.
nam et haec vobis nota sunt, et quae novissime multo indigniora commisit, quam
primum cognoscenda. (fr. XV 7 Conomis = fr. 31 Malcovati = Rutil. Lup. 2. 11)
‘For the time being, o judges, I shall not detail the scandalous actions that he committed
without the authorization of the people; I will not say anything of the false letters that he
sent to the council; I will not remember the frequent interrogations you addressed to
him; for these facts are known to you, and you must learn as soon as possible of the new
much greater indignities that he has committed.’
The passage comes from a speech of Lycurgus, possibly that against Lysicles, one
of the Athenian strategoi at Chaeronea; here, as in Aeschines’ speech, the mention
of false letters serves its purpose, to blacken the character of the adversary, without
any need of showing the letters in question. And this applies also to the last
mention of letters in Aeschines’ speech Against Ctesiphon, a passage particularly
important for the richness of its detail. Here is the text:
j P Øe  E r ÆØ  ŒE, N e b  ıºıæØ  ŒÆd › B  Ææ æAÆØ, ƃ ’
KØ  ºÆd ŒÆd ƃ æ EÆØ IçØŒ FÆØ N NØøØŒa NŒÆ, P Ææa H ıåø
IŁæø, Iººa Ææa H æøıø K B fi  Æfi ŒÆd B
fi ¯PæfiÅ; ŒÆd Kç’ x  K Ø
KŒ H  ø ÇÅ Æ ŁÆ , ÆF Ø PŒ KÆæ FÆØ æØ, Iºº’ › º ª F Ø
K fiH  fiø , ŒÆd a KØ  ºa Iººº Ø ÆæÆƪت Œ ı Ø· ÆæÆŒº ÆØ ’ ƒ b
ºØ N a ÆıH æ øÆ ‰ çºÆŒ B Å ŒæÆÆ, æ Ø ’ ÆN F Ø øæa
‰ øBæ B ºø Z. › b B  KŒ B IŁı Æ H ı ÅŒø u æ
ÆæƪªÅæÆŒg j ÆæÆ Æ ƺøŒ, ÆPe    h Æ B Å ŒæÆÆ
æØ ØEÆØ, H ’ æªø æ Ø ÆæÆŒåæÅŒ. Ø’ Iæå Ł KŒ H KŒŒºÅ ØH
P  ıºı   Ø, Iºº’ u æ KŒ H Kæø, a æØÆ Ø   Ø.
(Aeschin. 3. 250–1)
‘Does it not seem to you terrible, if the bouleuterion and the people are being ignored,
while the letters and embassies come to private houses, sent not by ordinary men,
but by the first men of Asia and Europe? And deeds the legal penalty for which
is death, these deeds certain men do not deny, but acknowledge them before the
people; and they read their letters to one another and compare them. And some of
286 Letter Writing and the Polis
them ask you to look into their faces as guardians of the democracy, while others ask
for rewards as being saviours of the city. But the people, discouraged by their experi-
ences, as though senile or declared of unsound mind, preserve only the name of
democracy, and have surrendered the substance to others. And so you go home from
the meetings of your assembly, not having decided on a policy, but as from a picnic,
having divided the leftovers among you.’
The private exchange of letters concerning affairs that should instead be discussed
in the assembly is here presented as destructive of democracy.67
In the two speeches by Aeschines in which letters are mentioned, then, two
different strategies are at work. In his speech On the Corrupt Embassy a few
‘normal’ letters, the dispatches of the Athenian strategoi, are mentioned, and
actually read. But it is not their content that matters; the letters are used for
their date, in order to build a chronological argument. Otherwise, as in Against
Ctesiphon, Aeschines associates letters with foreign rulers (Philip, the Macedo-
nians, or the Persian kings) and with private, unsavoury dealings that are under-
mining the democracy. This time again it is not the content that matters—most of
these letters are probably non-existent; it is the very existence of a correspondence
that is deemed to be sufficient to provoke the anger of the people. In the public
arena, the act of writing (or of receiving) letters can be used as a weapon against
the opponent—all the more easily if the letters in question are not read. It is these
negative connotations of letter writing, present also in some of Demosthenes’
speeches, that had been at issue in Antiphanes’ fragment.

6 . 5 . I S O C R A TE S A N D T H E G E N RE O F E P I S T O L A R Y T R E A T I S E S

On at least two counts, Isocrates’ oeuvre occupies a special place. Firstly, his
orations present themselves as public speeches, incorporating, for instance, ad-
dresses to the audience, and adjusting to specific aspects of the genre; but their
fictional character is often explicit (so for instance the Antidosis), and there is no
desire to imitate spoken language: his speeches were always meant to be read in
front of a public audience or in the presence of an individual.68 This renders his
work rather different from that of other orators. Secondly, he wrote a number of
shorter pieces that have been transmitted as part of the corpus Isocrateum under
the label of ‘letters’;69 their authorship has been questioned, but most are now
commonly accepted as genuine.70 The corpus as it stands comprises twenty-one

67
See already along these lines Lewis 1996: 148.
68
Nicolai 2004: 24–6. The only speeches meant for a court are the forensic ones (16–21), which he
wrote for others, as a logographos.
69
The letters (1. To Dionysius, 2. To Philip (1), 3. To Philip (2), 4. To Antipater, 5. To Alexander, 6.
To the Children of Jason, 7. To Timotheus, 8. To the Rulers of the Mytileneans, 9. To Archidamus) are in
the manuscript tradition placed at the end.
70
All early correspondence (e.g. the letters of Solon, Thales, Phalaris, Anacharsis, Heraclitus, and
the Pythagoreans, and of Themistocles, Artaxerxes, Hippocrates, Euripides, Socrates, and the Socratics)
is certainly a much later fabrication; pseudepigraphic are also the letters of fourth-century figures such
as Xenophon, Diogenes, Crates, Aeschines, Chion, and Dion. The status of the letters of Isocrates,
Plato, and Demosthenes is a matter of scholarly debate; some of the pseudepigraphic might be already
 The Legal and Political Stage 287

speeches (ºª Ø) and nine letters (KØ  ºÆ); however, quite a few of the speeches
present themselves as letters (Busiris, the speech To Philip, but also To Demoni-
cus), while conversely not all the letters correspond to the parameters of length
and privacy established by later rhetoricians.71
Let us first begin with a look at the documents embedded in Isocrates’ writings.
Because of the peculiar character of Isocrates’ oeuvre, it is not surprising to find no
citations of decrees in it; references to the reading out of laws are also exceedingly
scarce (they appear only in the forensic Aegineticus, at 19. 12–14, where the nomoi of
the Aeginetans, the Ceans, and the Siphnians are read); a treaty and oaths are read
out in Against Callimachus (18. 19–20). Letters are mentioned four times in
speeches, twice in Trapeziticus, once in Areopagiticus, and once in Philip.
The first mention in the Trapeziticus, a forensic speech written for a young man
from the Bosporus in the late 390s, concerns a letter that was actually read out in
the court, although its text has not been preserved: the letter was written by the
king Satyrus of Bosporus/Pontus to the city of Athens in favour of the speaker.72
This was thus a commendatory letter, a letter of patronage. The second reference
comes at the very end of the speech, and it is not clear to what letters exactly the
speaker here refers—letters of his father had not been mentioned previously, so
possibly this simply alludes to the already mentioned letter of King Satyrus.
Interestingly, once again the last word is with ‘words’:
It is also worth considering both Satyrus and my father, who have always esteemed you
above all the other Greeks . . . ; also, in the private contracts in which they are arbiters, you
come off not only on even terms but even at an advantage. [58] You would not reasonably,
therefore, consider their letters of little importance (u ’ PŒ i NŒø æd Oºª ı
 Ø ÆØ Ł a KŒø KØ  º). I ask you, therefore, on their behalf and on mine, to
vote according to justice and not to consider the false assertions of Pasion to be more
trustworthy than my own words. (Isocr. Trapez. 57–8)
The other two allusions to letters occur in epideictic speeches. One is in the
Areopagiticus: in a passage in which he is contrasting the ancient grandeur of Athens
with the current sad situation, the orator affirms that in ancient times the Greeks
looked to Athens for protection, while the Persian King feared the city; now, however,
‘As to the hatred of us among the Hellenes, you have heard the generals themselves;
and what the King thinks of us, he has made plain in the letters which he has sent.’73

fourth-century fabrications. The earliest letters we can be sure were kept and circulated seem to have
been those of the philosopher Epicurus (341–270 bc): Trapp 2003: 12.
71
Nicolai 2004: 120 (one of the most accurate and perceptive discussions of the entire issue of
genres and literary forms in Isocrates); Sullivan 2007: 7–8.
72
Isocr. 17. 52: ‘When Satyrus had heard us both, he did not wish to render a decision concerning
contracts made in Athens (æd H KŁ ª ø ı  ºÆø), especially since Pasion was absent
and not likely to comply with his decision; but he believed so strongly that I was being wronged that he
called together the shipowners and asked them to assist me and not suffer me to be wronged, and
having written a letter to the city he gave it to Xenotimus, son of Carcinus, for delivery (ŒÆd æe c
ºØ ıªªæłÆ KØ  ºc øŒ çæØ ¨  fi ø fiH ˚ÆæŒ ı). Please read it to them. Letter (ŒÆ
Ø IªøŁØ ÆP E. ¯Ø  º).’
73
Isocr. 7. 81: ŒÆd æd b  F  ı H  Eººø ÆPH IŒÅŒÆ H æÆŪH· ‰ b Æ Øºf
åØ æe  A, KŒ H KØ  ºH z  ł Kºø . Note here the tradition according to which
Isocrates would have written the reports Timotheus sent to the Assembly, discussed below, n. 75.
288 Letter Writing and the Polis

Reports of the generals (which may have been written or oral, but here are
presented as oral: the impression the text gives is that the generals have just
spoken) are put side by side with letters of the Persian king, in a context in which
Isocrates speaks as though addressing an actual assembly.
There is no remarkable difference between these and the ‘fact-of-matter’ refer-
ences to epistolary matters present in the speeches of the other orators. The last
reference to a letter in a speech, however, is different. It occurs in To Philip:
Isocrates there refers to a statement he has made in his letter to Dionysius (Ep. 1. 9),
quoting part of the passage. Thus, in this speech, an earlier document sent to
Dionysius of Syracuse is treated as a document in the public domain, supposedly
known to the audience, and as an authoritative text, at least as far as concerns
Isocrates and his reasons for approaching the powerful.74
Let us now look at the ‘letters’. Nine attributed to Isocrates have been transmit-
ted; more were known to the ancients.75 Their function is varied: three of them are
letters of recommendation (Epp. 4, 7, and 8); one is a short note to Alexander;
most offer advice on political issues to important public figures (Dionysius I,
Archidamus, Philip of Macedon). But the real nature of these texts is unclear: even
in the case of the letters of recommendation it is uncertain whether they are truly
private documents, or instead documents intended as examples for the school; this
applies a fortiori to the other texts.76 We face here an important issue: that of the
diffusion and publication of Isocrates’ work. His oeuvre (including the letters) is
very closely bound up with his school and his teaching activities. The question
thus arises whether the letters were meant primarily for their addressees, for their
addressees as well as for Isocrates’ pupils and circle, or only for the latter.
This is a question to which an overall answer cannot be given: some letters may
have been simply school exercises; some will have been real letters meant also for
circulation; some may have been purely private and may have been put together
with the others after Isocrates’ death by his students. After a detailed discussion
and a survey letter by letter, Nicolai concludes that we end up with more questions
than answers.77 And it is difficult to see how to break this impasse. For this reason,
my discussion here will be limited to formal aspects, and no evaluation of the
possible repercussions of the choice of the letter format for the discussion of
important political issues will be attempted—although it is evident that this is a
central issue.

74
5. 81: ‘And do not be surprised (as I also said in my letter to Dionysius after he had acquired the
tyranny) that I, who am not a general nor a public orator nor in any other position of authority, have
expressed myself to you more boldly than the others’, ŒÆd c ŁÆı  fi Å, – æ K ØºÆ ŒÆd æe
˜Ø  Ø  c ıæÆÆ ŒÅ   , N  æÆŪe J  Þøæ ’ ¼ººø ı Å
ŁæÆ æ Ø Øºª ÆØ H ¼ººø. Incidentally, the letter to Dionysius is one of the three letters
that are incomplete, presenting only a prooemium; this poses in an acute way the question of whether
the letter circulated in this form.
75
In his Letter to Philip, dated to 343 or 342 bc, Speusippus mentions ‘disgraceful letters’
(KØ  ºa ÆN åæ) slandering Philip and his family, written to the demos by Isocrates in association
with Timotheus (13). This finds an echo in [Plut.] Lives of the ten orators 4. 837c, according to whom
the reports addressed to the Athenian assembly by Timotheus had been written by Isocrates, who
received a talent for his pains. See Mathieu 1962: 163–4; Natoli 2004: 151–3.
76
Sullivan 2007: 9. Cf. the comments of Speusippus, Letter to Philip 13, with Natoli 2004: 153–8.
77
Nicolai 2004: 154–61.
 The Legal and Political Stage 289

The status of the nine texts at issue is made clear by the fact that Isocrates refers
to them by the term KØ  º, which is never used for self-definitions in the
epistolary speeches.78 Moreover, they present the main formal epistolary hall-
marks: the recipient is addressed in the second person; and a prescript was
probably present too.79 Most letters lack closing greetings (the one exception is
the letter To Timotheus, with its final ææø , ‘Fare well’); three letters (1, To
Dionysius, 6, To the Sons of Jason of Pherae, and 9, To Archidamus) are incom-
plete, presenting only a prooemium.80 But for the term KØ  º and for the
prescript—if indeed the letters had a prescript—the other epistolary hallmarks
(mainly the address by an individual to another specific, named recipient) are,
however, present also in some of the texts classified as speeches. And the reference
in the speech To Philip to the letter sent to Dionysius implies that similar
topics can be discussed in a speech and in a letter. So where did Isocrates put
the distinction—if he did mark a distinction—between letters and epistolary
speeches? A comparative analysis of his letters and speeches as well as a consider-
ation of his own normative statements (usually made in a context in which he is
infringing one of the rules he is formulating) offer some answers.
Concerning normative statements, the first and foremost is concision. It is
highlighted in five letters, in general at the moment where the letter might risk
going over the notional limit. The best instance is offered by the letter To Philip I:
 ººa  åø NE Øa c  F æª Æ  ç Ø Æ ÆØ ºªø: . . . æe b   Ø
ç  F ÆØ c IŒÆØæÆ: ŒÆd ªaæ F ŒÆa ØŒæe æ œg ºÆŁ  K Æıe PŒ N
KØ  ºB ı æÆ Iºº N ºª ı BŒ  K ŒºÆ. (Ep. 2. 13)
Although I have much more to say, because of the nature of the subject, I will
cease; . . . Besides, I fear the impropriety; for even now I have unawares gradually
drifted beyond the due proportions of a letter and run into a lengthy discourse.
Here epistle and speech are put side to side, and the distinctive element is length.81
The second letter To Philip (Ep. 3.1), the latest, written after Chaeronea, may also
be taken as implying the same, for it opens with the mention of a conversation
between Isocrates and Philip’s envoy Antipater, which led to Isocrates’ decision to
write to Philip about the course of action to be taken next. ‘This advice is similar to
that written in my speech, but it is expressed much more concisely’ (Ææƺ ØÆ
b  E K fiH ºªøfi ªªæÆ  Ø,  ºf  KŒø ı æÆ); the speech is

78
Nicolai 2004: 125, noting the exception of the letter To Archidamus (Ep. 9), whose authenticity is
doubted; Sullivan 2007: 8. If one includes in the count also the verb KØ ººø, as Sullivan 2007: 8–9
does, then out of twenty-six overall instances, twenty-three refer to letters or to the sending of letters,
while only three retain the older meaning of ‘injunction, command’; all three appear in the Trapezi-
ticus. The letter To Philip is also defined with غ  (ªæ ÆÆ appears for To Timotheus and Rulers
of the Mytilenaeans).
79
See Sullivan 2007: 9 for full discussion. The manuscripts of one family (the less important one)
have the formula › EÆ fiH EØ åÆæØ; manuscripts of the other family give as the heading the name
of the recipient(s) in the dative.
80
The others give a sense of closure, but present no final greetings. See Sullivan 2007: 9–10.
81
The other passages are: Epp. 3. 1; 4. 13; 8. 10; and 7. 10, although there the reason for concision is
not the risk of transgressing the limits of the ‘genre’, but rather the necessity to act quickly. Discussion
in Nicolai 2004: 124; Sullivan 2007: 10.
290 Letter Writing and the Polis

Philip, written in 346 bc. Again, the same topic may be treated in letters and
speeches, with length as the main difference.
Besides concision, there are statements to the effect that letters should be
written in a simple style, not as logoi, but pleasantly; this corresponds to later
epistolary theory, but there is no great stylistic difference between the language
Isocrates uses in the letters and in the speeches. Finally, the value of epistolary
writing can be inferred e contrario from apologetic remarks in which the writer
stresses that he is writing because he cannot have a personal meeting with the
addressee. The letter to Dionysius (Ep. 1) for instance begins by foregrounding
epistolarity: distance is the reason that renders the writing of a letter necessary;
had he been younger, Isocrates says, he would not have sent a letter, but would
have travelled to Sicily to discuss the situation in person (Ep. 1. 1).82 Through all
of the second and third paragraphs, Isocrates enumerates the reasons that render a
face-to-face meeting preferable to sending a speech (his letter): it is easier to
discuss the same matters face to face than to give one’s views by letter;83 men tend
to trust spoken advice more than the written one, because the latter carries the
marks of an artistic composition, while the former is more immediate; and, most
importantly, if there is any misunderstanding, it can be cleared up in conversation,
something that is impossible in the case of a written text or letter, for once the
writer is not present, there is no defender.84 Isocrates, however, manages to turn
all this into a compliment to his addressee, who will be able to disregard difficul-
ties and focus on the matter at hand.85
This type of apologetic remark appears also in the letter To Philip II (Ep. 3. 4: ‘if
I possessed the same strength as before and were not utterly spent, I should not be
speaking with you by letter, but in your presence should myself be spurring and
summoning you to undertake these tasks’), and in a slightly different form
(because generalized to all writing, in a way reminiscent of Plato’s critique of
writing in the Phaedrus) in the speech To Philip (25–9). So again, this motif can fit
both ‘letters’ and ‘speeches’.
The letters and some speeches seem thus to function in the same way. This can
be confirmed by a quick look at the Busiris: among those texts classified as
speeches, it is unquestionably the one closest to a letter. It opens with an address
to Polycrates (a contemporary logographer, based in Cyprus, author of a Defence
of Busiris and of a Prosecution of Socrates), couched in the form of an epistolary

82
N b æ  q, PŒ i KØ  ºc   , Iºº ÆPe ¼ Ø º Æ KÆFŁÆ غåŁÅ. Cf.
Livingstone 2001: 6–7.
83
Ep. 1. 2: Þfi A  ¼ Ø Ææg æe ÆæÆ çæ Ø j Ø KØ  ºB ź Ø; note the contrast
çæÇø / źø (źø had been used for letters in Thucydides).
84
Ep. 1. 3: K b  E KØ ºº  Ø ŒÆd ªªæÆ  Ø X Ø ı B fi  Ø F , PŒ  Ø › Ø æŁ ø:
I  ªaæ  F ªæłÆ  æÅ Æ  F  ÅŁ  K Ø. Here Isocrates is making a more general
point: this touches writings sent (KØ ºº  Ø), but also writing tout court. Inasmuch as this is a
general point concerning writing and not exclusively letters, it is close to the Platonic critique of writing
as presented in the Phaedrus (275 e).
85
Ep. 1. 3: P c Iºº KØc f ººØ ÆPH   ŁÆØ ŒæØ,  ººa KºÆ åø çÆ  ŁÆØ
ºª Æ  A Ø H ø: ª F ÆØ ªaæ ± Æ IçÆ  a ı åæÆ a æ ØæÅ Æ ÆPÆE
ÆE æ Ø æ Ø e  F. Note that writing is here forgotten: Isocrates now speaks through his
letter.
 The Legal and Political Stage 291

prologue.86 Among the elements of epistolary writing present in the opening are
the use of the verb KØ ººø ( d b KØ EºÆØ, 2), the insistence on the fact that
this is only a second-best solution for a direct, personal encounter, and the further
statement that the choice of Isocrates has been dictated by his desire for sending
his remarks to Polycrates privately, without sharing them with others (Isocr. Bus.
2). Here Isocrates is using the characteristics of privacy typical of the letter to
present the fairly critical content of his speech as if it were disinterested advice;
meanwhile, the audience (the readers of the speech) enter into a shared relation-
ship of complicity with the author, through being allowed access to a private
document.87 After the scene has thus been set, the alternative (Isocratean) enco-
mium of Busiris follows; but at 28 again a topos of epistolary literature appears,
with the remark that ‘If one were not determined to make haste . . . ’ (å Ø  ¼ Ø
c Ø ‰æ Å   . . . ), a topos mentioned again towards the close of the
speech (Bus. 44).
To the question of why Busiris was not classified with the epistles, three answers
have been given: that it contains an encomium, and is thus better classified
among Isocrates’ encomia; that the other letters are addressed to powerful foreign
kings or magistrates, and display Isocrates as the advocate of a Panhellenic
idea: Busiris would look out of place in such a collection; and that it is too long
for a letter (more than twelve pages, against some six apiece for the longest letters
of the corpus, the second and the ninth).88 For these reasons, then, the Busiris
finds a place among the longer speeches addressed to an individual (To Nicocles,
To Philip, To Demonicus, Euagoras), which also present epistolary features,
although not as marked. Ideally, a longer treatment would be needed here; but
what we have seen is sufficient to conclude that the main formal distinction is one
of length—that is, there are no formal differences besides the length between an
‘epistle’ and a ‘speech’ in epistolary form; but also, that the form of the epistle was
felt to have certain specific advantages (a faint notion of privacy and friendship)
that could be exploited to colour one’s speech, without the speech becoming, for
that reason, an epistle.
This phenomenon has been traditionally discussed as part of a more general—
and rather remarkable—contamination of genres and forms in Isocrates.89 How-
ever, as pointed out by Nicolai, this is our way of looking at the situation: for
Isocrates, and more generally for the ancients, the contamination of aspects
typical of different genres is not felt as a transgression or an innovation.90
Isocrates is simply experimenting, and using those rhetorical frames that fit the

86
Livingstone 2001: 5–7.
87
Nicolai 2004: 121; see also Sullivan 2007: 12, who points out that Polycrates is addressed in the
second person singular no less than twenty-five times in this relatively short text.
88
Livingstone 2001: 5–6.
89
The Antidosis is possibly the more striking instance: it is presented as a trial, as forensic (Isocrates
even asks for an indictment to be read—an obviously non-existent indictment (15. 29), besides having
extracts from other speeches read), but the opening gives away the fiction: ‘If the speech which is now
about to be read had been like those produced either for the law-courts or for oratorical display,
I should not, I suppose, have prefaced it with any explanation. Since, however, it is novel and different
in character, it is necessary to begin by expounding the reasons why I chose to write a speech so unlike
any other (15.1).’ See on this speech Ober 2004; Too 1995.
90
Nicolai 2004: 34, and more generally, 32–40.
292 Letter Writing and the Polis

situation best: a letter format for a short text; a mode of address close to the
epistolary one, together with a careful avoidance of the word KØ  º, which is
felt to be out of place in an extended context, for longer texts, such as the Busiris,
in which Isocrates wants to present what are very hard criticisms indeed (the
conclusion shows it) as friendly, personal, and private advice while playing on the
complicity of the public, who are made part of a private exchange; and a letter
format for long texts, when he wishes to foreground physical distance, as in the
letters to Philip and Dionysius. Above all, the epistolary format, in its strict
meaning or in the extended one, allows Isocrates, under cover of apologies for
the distance, the possibility of sending a type of writing that is closest to speech,
including direct apostrophe of the addressee, a writing that presents itself as a
personal and private exchange on a friendly level, but at the same time succeeds in
avoiding face-to-face speech, and that can be circulated among a wider public. But
such a strategy, I submit, may have worked also because Isocrates never directly
entered the arena of the Athenian assembly (and there is remarkable disagreement
as to the effective extent of his influence on Athenian, and more generally, Greek
politics in the fourth century bc).

6 . 6 . C ON C L U S I O N

The evidence on the use of letters from the speeches of the orators has to be
considered in both qualitative and quantitative terms; but it is also important to
distinguish between public and private letters (and public and private speeches).
To begin with the quantitative aspect: it is absolutely striking how very few private
letters are mentioned (let alone read), when compared to other documents such as
decrees, laws, contracts, and also witnesses’ statements.91 However, if we move to
the qualitative level, it transpires from forensic speeches that letters were indeed
relatively frequent in private dealings, especially among merchants engaged in
maritime trade. It is also to be expected that in a context of litigation quite a few of
these letters should be ‘deceptive’. In general, letters appear to have been used on
all sorts of private occasions. So why was evidence from letters not used more?
The reason must surely be that all other documents (obviously decrees and laws;
but also testaments, contracts, and statements by witnesses) could be accepted as
technically correct and reliable because written in the presence of witnesses; but
the same did not apply to private letters, whose authenticity might always be
doubted.92 The issue of preservation probably plays a role too: letters simply
transmitting a message, a tip, some instructions, or a request, would not be

91
After 380 bc, witnesses make a written deposition, to be read out in the law court, without any
possibility of modifying it or ‘cross-examining’ the witness. For a discussion of the role of the witness,
see Harrison 1971, 133–54; Bonner 1905; important discussion in Todd 1990; Thür 2005. See Ar. Rhet.
1355b35 for the distinction in artistic and non-artistic ‘proofs’, and Ar. Rhet. 1375a24 f. for the division
of ‘non-artistic proofs’ into  Ø, æıæ, ıŁBŒÆØ,  Æ Ø, ‹æŒ Ø (laws, witnesses’ statements,
agreements, statements produced under torture, oaths). Laws here must include decrees, while the
agreements include all other written documents.
92
See along these lines Scafuro 2004: 4–5 and esp. n. 14. Written testaments too could be
disregarded: see above, 82 and n. 81.
 The Legal and Political Stage 293

preserved. And if the letter did not exist anymore, if its evidentiary value was
moreover doubtful, mention of it might have been made en passant, but it would
not have made much sense to spend time on it.
Letters are quoted more frequently in deliberative speeches, because official
letters would be kept in an archive— it would thus be possible to make use of them
on a later occasion, while their authenticity would not be open to doubt; more-
over, even if the letter did not exist anymore, capital might be made out of a
mention of it, in a way that was impossible in forensic speeches. A qualitative
analysis shows that letters were used for public service, in the form of reports sent
by generals or embassies to their own polis; but at the level of interstate diplomacy
(as well as in their relation with the cities under their control) the greatest letter
writers by far are Philip of Macedon and the king of Persia. References to their
letters are concentrated in a few important speeches. The letters are used as
documents of a policy (when their content is at the centre of the discussion, they
are read in their entirety), as archival documents permitting an exact dating of
events, or simply to show, by the sheer number of references to them, the heavy-
handed intrusion of a foreign power into the city’s affairs. In this context, the act of
writing letters (and at times the act of writing tout court) can be used as a weapon
against the opponent;93 this is particularly true when letters having implications for
the entire community are privately written by common citizens—or when letters
having implications for the entire community are privately sent to common citizens.
We see here the other side of the discourse of comedy. For comedy had linked
letter writing to femininity, small talk, and private affairs, pairing it at the same
time with the corrupt speech of the rhetors. Letter writing is something that
belongs to the private domain; when it moves into the public space, it has to be
kept under control.
Isocrates is one of the first instances of the use of the letter form for non-
personal, and yet non-official, communication; with him, we are at the beginning
of what will become a tradition of prose treatises clothed in epistolary form.94
The comparison between his letters and his epistolary speeches shows that there
is by now a clear sense of what exactly a letter is (or should be); but also, that
the principal element distinguishing a letter from a written speech addressed to
someone is length, nothing more.
From now onwards, letters will enter the public domain, and will be part of
everyday life. Thus, they are cited among the instances of peculiar behaviour
displayed by two of Theophrastus’ characters, the ‘boastful’ or ‘fraudulent’ man
(IºÆÇ), and the ‘arrogant’ man (æçÆ ). According to Theophrastus, the
fraudulent man boasts of non-existent possessions as much as of non-existent

93
See above, n. 59, for Demosthenes’ sneers at Aeschines’ activity as public clerk. To the instances
cited there (Dem. 19. 200 and 18. 209) add Dem. 19. 237, where the irony involves also Aeschines’
brothers; 19. 314, which presents Aeschines as himself denying his past activity of undersecretary; 18.
127 (Aeschines as a ‘poor devil of a secretary’, ZºŁæ  ªæÆ Æ), and 18. 261, where being a
secretary is ironically characterized as ‘the best job possible’ e ŒººØ   H æªø. See also Scodel
2011: 81–2 for the famous passages from Dem. 18. 259 and 19. 199, in which Demosthenes recalls
insultingly that Aeschines used to read the books for his mother when she did initiations.
94
For a sweeping analysis of the use of the letter form in philosophy, based on the letters of Plato,
Aristotle, and Epicurus, see e.g. Gigon 1980; Wohl 1998 offers a detailed, fascinating discussion of the
Platonic letters. There would be much to be gained from a contrastive reading of the two epistolary
corpuses (Isocrates’ and Plato’s), for they present some startling similarities.
294 Letter Writing and the Polis

connections: he pretends to have been campaigning with Alexander; he ‘will say

that he has three letters from Antipater asking him to come to Macedon’, an
invitation joined moreover to an offer to export timber without paying tax on it,
which he refuses because of his alleged fear of sycophants.95 As for the arrogant
man, his fault lies in the disrespect for the forms of epistolary courtesy:
ŒÆd KØ ººø c ªæçØ, ‹Ø ‘åÆæÇ Ø ¼ Ø’, Iºº ‘‹Ø  º ÆØ ª ŁÆØ’ ŒÆd
‘I ƺŒÆ æe b ºÅł  ’96 ŒÆd ‘‹ø ¼ººø c  ÆØ’ ŒÆd ‘c Æå Å’.
(Theophr. Char. 24. 13)
And in sending instruction by letter, he will not write ‘I should be much obliged,’ but ‘I
wish it to be thus and thus’, and ‘I have sent to you for’ this or that, and ‘that it may not
be differently’, and ‘without a moment’s delay’.
The fact that Theophrastus adds written communication to the forms of behav-
iour that may express arrogance, besides direct speech or actions, shows how
much letter writing has become part of everyday life. A similar joke on the
omission of all forms of epistolary courtesy appears in the vignette of the character
of the inconsiderate man (ÆPŁÅ) composed by Ariston, who followed Lycon,
Theophrastus’ own successor, as head of the Lyceum in c.225 bc: the inconsiderate
man, ‘in writing a letter, does not prefix it with greetings (åÆæØ), nor does he
close it with a farewell’.97 Clearly, there was a sense of what ought to be the content
of a letter, and of how this content should be presented—the apologetic remarks in
Isocrates’ letters prove the existence of generic expectations—as well as of ways to
get round them. At the same time, the fact that Theophrastus mocks the boastful
man for his pretensions at having direct access to the king of Macedon shows that
there were tensions linked to the public exhibition of direct, private contacts
between Athenian citizens and foreign kings: the boastful man reminds one
forcefully in his epistolary behaviour of the sketch of Demosthenes ‘dangling
letters from his fingers’ drawn by Aeschines and then Dinarchus; it also fits the
context sketched in the analysis of the riddle proposed by Sappho in Antiphanes’
homonymous play.
Notwithstanding these tensions, from this point onwards, and throughout
the Greek world, the elites will assume the role of intermediaries between the city

95
Theophr. Char. 23.4: ŒÆd ªæ ÆÆ b NE ‰ æ Ø Ææ Øæ ı æØ b ºª Æ
Ææƪª ŁÆØ ÆPe N ÆŒ Æ. See on the passage Diggle 2004: 27–9 (for the historical context
of this sketch and a dramatic date for it of c.319 bc), and 436; and the notes of Rusten 2002, ad loc. An
IºÆÇ is a fraud, an impostor, as much as a boastful person; he is a typical character of new comedy.
96
See Diggle 2004: 451–2 for close discussion of the text. I ƺŒÆ normally requires an object,
and the corrections ºÅł   or, better, ºÅł  ı have been proposed (the latter finds a parallel e.g.
in Xenophon, Cyr. 3. 1. 2; other parallels in Diggle, 2004: 451–2); but as Diggle points out, the lack of
object may also be attributed to the shorthand style and the self-centredness of the arrogant man.
Interestingly, some of the messages embedded in Herodotus presented such an ellipsis.
97
ªæçø KØ  ºc e åÆæØ c æ ªæłÆØ Å KææH ŁÆØ ºıÆE , Ariston fr. 14 Wehrli =
Rusten 2002, col. 17. Ariston of Ceus composed a work On Relief from Arrogance, which contained
sketches of characters; the work is lost, but fragments are preserved in Philodemus. See on him, besides
Wehrli and Rusten, Diggle 2004: 9–10. Diggle compares with Plautus, Bacch. 1000 (‘non priu’ salutem
scripsit?’), a good comparison, as these character sketches have to do with comedy; but also with Plut.
Mor. 1035 b–c ‘those who propose decrees to the cities preface them with the words “To Good
Fortune” ’.
 The Legal and Political Stage 295

and the king—through embassies, but also through epistolary communication.98

A story narrated in Plutarch’s Life of Demetrius (thus concerning a slightly later
period than that which has been tackled until now) well illustrates the difficult
situation of the Athenians, trapped between the desire to refuse acceptance of
letters emanating from a king (which at this time means orders), and the necessity
of accepting them. According to Plutarch, a certain Cleainetus, one of the ‘young
friends’ of Demetrius Poliorcetes, asked from the king a letter requesting that the
Athenians cancel a debt of fifty talents owed by his father Cleomedon, and
presented that letter in the assembly; the Athenians agreed, but then, exasperated,
they also passed a decree that no citizen should bring a letter from Demetrius
before the assembly. When Demetrius made his anger at this clear, the Athenians,
frightened, rescinded the decree, condemned to death or exile its proponents, and
passed another decision, that ‘it seemed good to the Athenian people that what-
ever King Demetrius should ordain in future, this should be held righteous
towards the gods and just towards men’.99 Plutarch’s narrative may have been
embellished, and the exact chronology of the affair is uncertain; but surely the
issue was not the refusal of any letters from a king—that seems an impossible
notion, whatever the exact date of the story—but rather the refusal that any letters
from a king concerning a particular citizen be brought by that particular citizen to
the assembly.100

98
Epistolary contacts in the context of embassies: cf. the decree for Hegesias (I. Lampsakos 4)
discussed above, 169. For the role of the elites see Paschidis 2008: 469–70. Contact with the king could
be initiated via letters: cf. Plut. Arat. 15. 4; Diog. Laert. 4. 39. The seal imprints found at Kallipolis show
that the strategos of the Aetolian league kept archived in his house letters from (among others) the
Ptolemies, the Seleucids, and the Attalids. See Paschidis 2008: 470–71 and n. 4, and specifically on the
Kallipolis archive, Pantos 1996; Invernizzi 2004: 317–19.
99
Plut. Dem. 24. 6-10: . . . ˚ºÆ  › ˚º   , n TçºÅŒØ fiH Ææd ŒÅ Œ Æ
ƺø IçŁBÆØ ØÆæÆ  , ªæ ÆÆ Ææa ˜Å Åæ ı Œ  Æ æe e B , P  
Æıe ŒÆfi åı, Iººa ŒÆd c ºØ ıæÆ. e b ªaæ ˚º  Æ B ŒÅ IçBŒÆ, KªæçÅ
b łçØ Æ ÅÆ H  ºØH KØ  ºc Ææa ˜Å Åæ ı Œ ÇØ. [ . . . ] Ø b æ łÅç Æ ,
åŁÆØ fiH  fi ø H ŁÅÆø, A ‹ Ø i › Æ Øºf ˜Å æØ  Œº fi Å,  F ŒÆd æe Ł f ‹ Ø 
ŒÆd æe IŁæ ı r ÆØ ŒÆØ . On the affair see Paschidis 2008: 96–7.
100
Paschidis 2008: 96–7 accepts the chronology implicit in Plutarch’s narrative, and places the affair
in c.303 bc, just before Demetrius’ initiation into the Mysteries; De Sanctis 1893 [1966] 273 considered
the Athenian reaction impossible before the battle of Ipsus, and thus dated the anecdote to after 301,
when Demetrius’ control over Athens was looser; a date c.294 bc was proposed by Manni; see Santi
Amantini 1995: 347–8.
 7
Poleis and Kings, Letters and Decrees
Official Communication in the Hellenistic Period

The analysis of the references to letters in the speeches of the orators has shown
how complex and nuanced the connotations of epistolary writing are in fourth-
century Athens: an everyday thing, when used for private exchanges between
individuals or for official communications between the generals and the polis; a
way of circulating ideas publicly, while presenting them as addressed to a specific
individual; a treasonous enterprise, when information that should go directly to
the assembly is conveyed by letters to individuals; and the typical means through
which kings such as Philip of Macedon, and after him, Alexander, conduct
diplomatic affairs with the Greek poleis. A certain diffidence towards letter writing
is evident in the treatment of epistolary communication in tragedy, and similarly,
the possibility of a misuse of the letter has emerged from the analysis of the
fragment from Antiphanes’ Sappho; but Antiphanes had left open the possibility
of a correct use of letter writing (and of the rhetors’ speech). The picture offered by
the speeches of the Attic orators is similarly nuanced.
What of the following period? And what of the other Greek poleis? Is it possible
to find in other types of documents a confirmation of what has been suggested?
How did official letter writing actually function, and more generally, what was the
formal shape of the documents through which interstate diplomacy was con-
ducted? We have already seen some instances in the works of the ancient
historians of the use of letters (by Greek tyrants and generals, oriental kings,
Hellenistic kings, Roman consuls, and Carthaginian mercenaries). But rather than
once again using literary sources, I propose to turn now to the epigraphic material.
The advantages will be twofold. Firstly, inscriptions are possibly the only strongly
polis-oriented type of evidence available, once we leave the Athenian drama and
law-courts; thus they are our only possibility of getting an idea of how letter
writing for official purposes was perceived in the various Greek communities.
Secondly, inscribing a document on stone (as opposed to simply keeping it in the
archive) means endorsing it in a strong way. Documents are not simply (or
fundamentally) inscribed for the purpose of information, although they may be
used for this purpose: an oral proclamation, and/or temporary inscription on a
whitened tablet, together with an archive copy, would have taken care of that
aspect. Rather, the inscription on stone inserts a decision in a very physical way
into the landscape of the city; whoever sees the stone, without any need of going
through the document, will be reminded of the process through which this
298 Letter Writing and the Polis

document—and all similar ones—are validated.1 Not all documents were in-
scribed on stone, and not all official letters will have been either. Mapping the
localities and the situations in which it was decided to publish an official letter on
stone allows us to better understand an aspect of the ‘epigraphic habit’ of Greek
cities, the reasons why some texts were published and others not, some on certain
types of support and others on different ones, some in certain areas of the city and
others in other areas, etc.2 After all, the decision of the recipient(s) to have the
letter inscribed, the reaction of the polis or the body responsible for its publica-
tion, are, for the purposes of understanding the connotations of official letter
writing, almost as important and significant as the initial decision of the sender.
For these reasons, in what follows we shall look at letters on stone—or more
precisely, following in the footsteps of a seminal article by Jean-Marie Bertrand,
we shall look at the contrast between (and the distribution of ) two specific types of
official writing: the letter and the decree. For the type of language chosen by a state
is not innocent: ‘une structure étatique révèle sa nature par la façon dont elle
s’exprime’, the choice of the mode of communication reflects the constitutional
context from which the document itself originates.3 The comparison between the
functioning of the two forms will help illuminate their specificities.

7 . 1. TH E L A N G U A G E O F R O Y A L L E T T E R S
AND CIVIC DECREES

Although the two types of document appear prima facie to be used in much the
same way in interstate diplomacy, they present markedly different formal charac-
teristics, and are couched in a relatively different language.
A decree is basically a long sentence including the date, the reasons, the
approval of decisions taken, and provisions for their realization. There are of
course minor variations from case to case, and even more so from polis to polis,
but the basic features do not change.4 A typical Athenian decree of the fourth or

1
For the localization of inscriptions at Athens, see Liddel 2003. The best discussion of the complex
relationship between oral discussion, decision, publicity through a public proclamation (ŒæıªÆ) or
exposition on a whitened tablet (ºŒøÆ), and inscription on stone is Faraguna 2006, who offers also
an excellent discussion of the archival practices of the Greek poleis. See also Rhodes with Lewis 1997:
525–7.
2
The necessity of paying attention to the ‘epigraphic habit’ of the various poleis (where ‘epigraphic
habit’ comprises both the type of document being written, and the choice of place of publication) has
become clearer in recent years: see Hedrick 1999; 2000; Osborne 2009; Liddel 2009. An early instance is
Boffo 1995; but official letters are absent from the typology she sketches of the various documents
emanating from Greek poleis.
3
Bertrand 1985; Bertrand 2006: 91 for the quote on language and the state.
4
Plat. Gorg. 451 c, where the formula a b ¼ººÆ ŒÆŁ æ, etc. is treated as typical of those who
compose proposals in the assembly; ƒ K fiH ø fi ıªªæÆç  Ø shows that the decree was perceived
as a specific form, with a peculiar language (see Dover 1981: 3). The same emerges from Plato’s
Phaedrus, 257 e 1–258 a 9. Here, Socrates considers that the way the name of the proposer (in the
formula ‘x said’) follows on the enactment formula (Resolved by the council and/or the people) is a
mark of arrogance (‘ ’  çÅ Ø ‘B fi  ıºBfi ’ j ‘fiH ø
fi ’ j Iç æ Ø, ŒÆd ‘n r ’—e Æe c
ºªø ºÆ H ŒÆd KªŒøØÇø › ıªªæÆç— ØÆ ºªØ c a  F , K Ø ØŒ   E
 Greek Cities and Epistolary Writing 299

third century would for instance have featured the following elements: an invoca-
tion to the gods (Ł ), and possibly also a heading; a prescript, which would have
included dating elements (name of officials, nature of the meeting, etc.), the
enactment formula (  followed by a dative, e.g. B  ıºB fi ŒÆd fiH ø fi , ‘It
seemed good to (or: Resolved by) the council and/or the people . . . ’), the name of
the proposer(s) of the motion (› EÆ r , ‘X said’); the preamble or motivation
clause, giving the background and the reasons for the motion (usually introduced
with K Ø , ‘whereas’, ‘inasmuch’ or K , ‘since’); there might at this point also
have been a hortative formula (‹ ø i s . . . , ¥ Æ s . . . , ‘in order that . . . ’, ‘so
that . . . ’); then, the resolution or motion formula (  åŁÆØ or KłÅç ŁÆØ with a
dative, e.g. B  ıºB fi ŒÆd fiH øfi , ‘let it seem good to (or: let it be resolved by) the
council and the people . . . ’, followed by the decisions, expressed with infinitive
clauses depending on  åŁÆØ, or on the verb of the proposal; and finally, closing
formulae, concerned mainly with the public inscription of the decree and/or
sanctions for disobeying its provisions, as well as rider-formulae (eventual other
stipulations and documents).5 The main characteristic of decrees is their imper-
sonality. Through being written in the third person, the decree eliminates con-
frontation and brings forward the polis as a whole: the entire city, not a specific
individual, stands behind the decree.6 There is no named addressee: the text is
simply there. The proposal, which constitutes the one opening to individuality in
the decrees, is, with its r , the recital of a story long past;7 the use of the aorist
  (and of the perfect  åŁÆØ) in the enactment and in the resolution
formulae renders the decision narrated an established fact endorsed by the entire
civic community. Dover acutely pointed out that three persons are needed to
produce the text that will be inscribed on the stele: the proposer of the decree, the
secretary of the council, and the under-secretary ( ªæÆÆ), a professional
clerk; the three get together to write down the text, as is shown for instance by a
passage of Aristophanes’ Thesmophoriazusae (ÆFÆ b çÆæH ºªø. | a
 ¼ººÆ a B ªæÆÆø ıªªæł ÆØ, ‘This I say openly. The rest I shall
take down with the secretary’, 431–2). But the verb used for that writing is
always in the middle voice, ıªªæç ŁÆØ, a verb used of copying down some-
one else’s utterance, and never the active ıªªæçØ, which takes as its object

K ÆØÆØ c Æı F çÆ, K  ı ÆŒæe ØÅ   ªªæÆÆ). It is important to note that
while Socrates’ reference to the peculiar format of the decree is immediately evident to Phaedrus, his
point about the arrogance is not; as Dover 1981: 1–2 states, this implies that the Athenians’ perception
of the decree would probably not have been, as Socrates submits, that of an individual praising himself.
5
See Guarducci 1969: 11–19; Rhodes with Lewis 1997: 3–7; McLean 2002: 215–27; Rhodes and
Osborne 2003: pp. xix f.
6
Dover 1981: 1–2.
7
For the dialogue instituted between the proposer of a law and the assembly, between the r  and
the answering  , a feature that found in certain moments a visual expression in the separation of
preamble and text (the ‘perfect pattern’), see Bertrand 1985: 104 and n. 21, with further bibliography.
As Liddel 2009: 423 points out, most Peloponnesian cities actually omit from their decrees the name of
the proposer, even if the language is clearly modelled on the Athenian decree language. This anonymity
may be interpreted in various ways; Liddel 2009: 423 suggests that it implies a less democratic mode of
publication, because it takes away the emphasis from the free speech and deliberation taking place in
the assembly.
300 Letter Writing and the Polis

something ‘which is wholly composed by the person who is the subject’.8 Thus,
the decree presents itself as the transcription of an act of the past, endorsed
by the entire community.9 As pointed out by Robin Osborne, the gap created by
the temporal structure of the ‘genre’ decree between the story of the decree itself
and its actual performance is in fact a depoliticizing gap: the decree does not
record (or script) a performance, it is an independent inscribed performance, a
peculiarly democratic performance inasmuch as it showcases the democratic
process of decision-making.10
As for official letters, a typical product of Hellenistic royal chanceries, they tend
to be structured more loosely, depending on the specific issue at hand. In its most
advanced form, a royal letter would have comprised the following elements: a
prescript, in the form of a greeting, expressed by the name of the sender, that of
the addressee in the dative, and the infinitive åÆæØ (formal titles being usually
absent); sometimes, depending on the situation, a formula valetudinis (very rarely,
mainly in letters internal to the royal administration, or in those of uncertain
authenticity transmitted through literary texts), or a reminder of earlier relation-
ships; then, the subject-matter of the letter; and a concluding clause (ææø ,
ææø Ł, ‘be well’, PåØ, PıåE, ‘best wishes’, or something similar).11 The
defining characteristics of letters are thus an opening with greetings in the third
person, where both sender and addressee are named, followed by a text written in
the first person, singular or plural, and addressing someone else (an individual or
a group, with a second person singular or plural) directly and explicitly; letters
often include a (brief ) narrative of past events, but may otherwise be composed in
the present or even future tense (barring the use of the ‘epistolary past’).12 That
kings should have chosen to write letters to communicate is not surprising:
epistolography is a strongly personal mode of discourse; the choice of this mode
corresponds to the fact that kings (or tyrants) conceive of their territory as their
æªÆÆ, their private property.
The contrast between the two formats is evident. There are, it is true, in the
fourth century and afterwards, some instances of mixed formats: they are mainly
the result of non-Greek kings having to deal with Greeks. As far as I can see,
the only instance of the welding of letter format and city decree is the text of
the King’s peace of 387/386 bc, as Xenophon reports it, which begins as a rescript
in the third person and with no direct addressee, but then shifts to first person
discourse.13 There are, however, other instances of the adoption of the decree-
format by groups or individuals external to the polis. An inscription from
Labraunda dated to the mid-350s bc, with which Mausolus and Artemisia grant
proxeny to the Cretan city of Cnossus, offers an excellent example of such an
adoption:

8
Dover 1981: 1–2.
9
Bertrand 1985, esp. 105–6. For a detailed study of this form of official language across the Greek
world see Rhodes with Lewis 1997.
10
Osborne 1999: 357.
11
See Guarducci 1969: 105–20; Welles 1934, pp. xli–xlvii; Rhodes with Lewis 1997: 7.
12
On the tenses of Greek official letters see Welles 1934: p. lxx; Virgilio 2010 and Virgilio 2011: 37–69
for a survey of the style of Hellenistic royal chanceries.
13
Xen. Hell. 151–3, see above, 151–3. I am grateful to Peter Rhodes for calling my attention to this
passage.
 Greek Cities and Epistolary Writing 301
[]  Æı ººøØ ŒÆd [æ]Ø ÅØ. K Ø [c] | ˚  Ø Ø ŒÆd N ÅØ ŒÆd Å Å[Ø Øa]
[º ı] | ¼ æ IªÆŁ  N Ø æd Æ [ ]øºº _ _ [] vvv | ŒÆd a _ Æı ºº ı
_ _æªÆÆ, r Æ[Ø] vv | 5 ÆP f [æ]  ı ŒÆd PæªÆ K[ ]e[] | Id åæ  .
__
r ÆØ b ŒÆd IºØÆ  ÆP[ E,] | › Å Æ øºº  ¼æåØ, _ _ ŒÆd _ [ ]º [ı] | ŒÆd
Œ º ı I ıºd ŒÆd _I_ _  · Ka[  Ø] | I ØŒBØ ˚ø  ı, _K غ _ _ _ŁÆØ v | 10
Æ øºº  ŒÆd æØ Å, ‹ ø vvv | c I ØŒ ÆØ, ŒÆa ÆØ_ v | c
_ _
ÆPH. (Rhodes and Osborne 2003, _ _ no. 55 = I. Labraunda ii, 40) _ _

Resolved by Mausolus and Artemisia. Since the Cnossians both privately and publicly
have consistently been good men with regard to Mausolus and the affairs of Mausolus,
they shall be proxenoi and benefactors for all time; they shall also have immunity, in as
much territory as Mausolus rules, and the right to sail in and out inviolably and
without a treaty. If anyone wrongs the Cnossians, Mausolus and Artemisia will take
care that they are not wronged, as far as it is in their power.
The text opens directly with the enactment clause and two personal names,
followed by the motivation clause, and then by the decisions in the infinitive. The
decisions are verbalized in what appears to be a blend of the autocrat’s language
and the language of the Greek polis; yet, the display of power is here overt, there is
no suggestion at all of a democratic process.14 The important point is that civic
privileges and the civic language typical of the Greek poleis are here adopted by a
Carian satrap, and blended with the forceful exercise of an individual, personal
power, such as finds expression in official letters (note a æªÆÆ, to define the
territory under Mausolus’ control: this is the term that will be used to denote the
Hellenistic kingdoms; or the use of ¼æåØ for the land controlled, a term which on
the one hand points to absolute power and on the other is used for civic
magistratures).15 The reason for the adoption of Greek civic privileges and
language is certainly to be found, with Gauthier, in the fact that Mausolus is
addressing Greeks, and wants to be recognized as a member of the club: the
honours attributed to the Cnossians serve the purpose of highlighting Mausolus’
own Hellenism. A few years earlier the Erythraeans had awarded Mausolus,
defined in their document by the ethnic ‘Mylaseus’, exactly the same honours:
the titles of euergetes and proxenos, and the citizenship (I. Erythrae 8 = Rhodes
and Osborne 2003, no. 56); they would do the same for ‘Idrieus, son of Hecatomnus,
Mylaseus’ (SEG 31, 969). In designating Mausolus through his origin in a civic
community, Mylasa, that is, in treating him as a Hellene, the Erythraeans certainly
followed the satrap’s own wish; in dealing with a Greek community, Mausolus
adopted its formal language.16 A certain awkwardness shows that this text is
something of an experiment (note especially the final sentence, beginning like a
threat to punish wrongdoers, but turning midway into an attempt to prevent

14
Hornblower 1982: 75, stressing the brutal frankness; Gauthier 1985: 163; Bertrand 1992: 108
(who sees in this text an example of how in Asia Minor in the fourth century concepts of power are
forged and elaborated that will form the basis of the Hellenistic kingdoms; note also his commentary on
the difference between the Aramaic and the Greek text in the trilingual from Xanthos, 121–2).
15
Hornblower 1982: 75, who compares with Leukon of Bosporus, speaking of himself as ¼æåø, and
cites in n. 157 Tod’s observation that ‘with reference to his Greek subjects Leucon used the title of
¼æåø’; Gauthier 1985: 164 also compares with the Bosporan kings.
16
Rhodes with Lewis 1997: 354. Note that the fourth-century tragic poet Theodectas wrote a
Mausolus, which supported the (dubious) claim to Greekness of the Hekatomnids. Gauthier 1985:
164 compares the grants of proxeny by Paerisades king of Bosporus (below), made over all the territory.
302 Letter Writing and the Polis

wrong being inflicted). As for the award of proxenia, it presents itself here as the
opposite of the normal situation: instead of a community appointing a member of
another state to act as their local representative or proxenos, here two individuals
award the proxenia to an entire community. What is extraordinary here, however,
is not so much the fact that proxenia is granted to a city (there are instances of the
grant to entire communities), but rather that two individuals, one of whom a
woman, feel themselves in a position to grant such an award (as well as exemption
from tribute and security over all the territory they control), and choose to do so
in the language typical of the Greek poleis.17
There are some other instances of kings or groups adopting formats and
procedures typical of the Greek cities: Rhodes and Lewis list the case of the
kings of the Bosporus, who in documents dating to the mid-fourth century bc
give proxenia to individuals (CIRB 1= Syll.3 217, cf. CIRB 2 and 3). Cassander in
the last quarter of the fourth and Lysimachus in the third century make grants of
land through decrees (and not through letters).18 At the very end of the fourth
century, Berenice and her sons at Seuthopolis in Thrace blend together elements
of a decree and of an oath in a text meant to guarantee the safe passage of a certain
Epimenes, inclusive of his possessions, to Spartokos, in accordance with a promise
made by Seuthes.19 Strikingly, after the invocation to Agathe Tyche and the
statement that this is an oath, followed by the names of the parties, an K Ø c
(line 2) introduces a motivation clause, followed by the resolution formula
 åŁÆØ BæŒfiÅ ŒÆd  E ıƒ E ÆPB, ‘let it be resolved by Berenice and her
children’, lines 7–8. The editor adduces as the reason for this liberal use (‘freier
Umgang’) of the language of Greek decrees the desire on the part of the person
who styled it to give an official patina to the document, and at the same time to
confer on Berenice and her sons the same authority as that of the decision-making
organs of a Greek polis; in this context, he compares the appropriation of
the language of the Greek polis by Mausolus and Artemisia some fifty years
earlier.20 Another striking example (and on this I will close) is that of the priest
of Labraunda Korris, for it shows very well what could be done—and what was not
done—by the Hellenistic royal chanceries. At Labraunda, in the mid-third century
(250–225 bc), Korris and his syngeneis (here ‘associates’) drafted and published a
honorary decree for someone whose name is lost:
[․․․․․․․․․c.34․․․․․․․]Ø e [․c.6․․] | [․․․․․․․․․c.23․․․․․․․․․․]Æ N  ÆØ  ŒÆd
Œ[ ]ØBØ [․c.5․] | [․c.5․] _ _ åŁÆØ HØ ƒ[æ]E ˚ ææØ Ø ŒÆd  E [ ı]ªª
_ _ _ _[Ø  F]_ |
_[˚ ææ]Ø  K ÆØ ÆØ  ÆP , ‹Ø ŒÆºe ŒÆd IªÆŁe [ªª]- | 5 []Å
_ ÆØ N ÆP
_ f ήd
_ _

17
On the awkwardness of the final sentence see Rhodes and Osborne 2003: 264, who also affirm
that the award of proxenia to a community makes nonsense of the institution; but on this see Gauthier
1985: 162–4, with list of proxenia grants to entire communities; the objection should be rephrased as
above.
18
Respectively Rhodes with Lewis 1997: 544; Syll.3 332; and SEG 38, 619. There are, however,
interesting nuances: the second text (Cassander) uses the present  ø Ø throughout (three times) for
the present donation, while Lysimachus uses the past (  øŒ); both adopt the typical formula for the
exemption from taxes.
19
Text, German translation, and commentary: Elvers 1994 (SEG 42, 661).
20
Elvers 1994: 253–4 (detailed analysis of the various parts) and n. 44, with earlier literature, and
the remark that associations (e.g. the Dionysiac artists) imitated in their decisions the style of the
decrees of the Greek poleis.
 Greek Cities and Epistolary Writing 303
IƪæłÆØ ÆP[e] æ   [Œ]Æd [P]- | æªÅ  F  ƒæ[ø] ˚ ææØ  ŒÆd [H
ıªªH H] _ _ _ | [ _ F] _˚ ææØ_ ,__ çÆ[H]
_ _ ÆØ ’ ÆPe K [HØ I]ª[H]Ø [åæı _ HØ] |
_ _
çøØ I[æ]B Œ Œ[Æ]d P Æ, [e] b ŒæıŒÆ - | [ ] Iƪ æF ÆØ· _ _
 ŁÆØ b_ ÆPHØ _ _ ŒÆd_ _ [_ _]º_ØÆ ŒÆd ª- |_10 [Œ]Å
_ _Ø_ _ŒÆd
_ _ æ _  æÆ K  E IªH
_ Ø·
_ _ ÆØ
r _ _ _ _ ’ ÆPe
_ çıºB | [_Æ] _ ººØ  å Æ ƒæH ŒÆd H º Ø H ø vv? | ŒÆa
ÆPa b  ŁÆØ ŒÆd  E Kªª _  Ø ÆP F· IÆ|ªæłÆØ_ b e {ł} łçØ Æ K HØ ƒæHØ
 F ˜Øe  F ¸ÆæÆ [ ı.] (I. Labraunda 11)
Privately and publicly . . . Let it be resolved by the priest Korris and Korris’ syngeneia
to praise him for having [been] honourable and good towards them and to inscribe
him as a proxenos and benefactor of the priest Korris and Korris’ [syngeneia], and to
crown him in [the game] with a [golden] crown for his virtue and benevolence. Let the
herald publicly proclaim this. Let also citizenship and the right of holding land and the
privilege of a front seat at the games be given to him. Let him belong to the tribe of
Ibanollis (?) taking part in sacred occasions and all other things. Let this also be
awarded on the same terms to his descendants. And that the decree be engraved in the
shrine of Zeus Labraundos.
Korris and his syngeneia (a group that had been formed by Korris himself )
give themselves here the status and the authority of a city; a number of other
associations will set themselves up in the same way. Yet, it is important to
distinguish between categories, and between the reasons for the appropriation
of the language of the polis. For the documents emanating from relatively
peripheral kingdoms or from small independent associations, the assimilation
of the language of a Greek polis not only was unproblematic, but could actually
constitute a positive feature, inasmuch as it could add to the prestige and sense
of finality of the decision. But the text of the King’s peace, as given by Xenophon,
or in a lesser measure the grants of land by Cassander and Lysimachus, cannot
be read as the results of a desire to appropriate the authority and cultural status
that went with the decrees emanating from Greek poleis. Cassander and Lysi-
machus were perfectly familiar with the Greek city-decree, but had to deal with a
new situation;21 as for the King, one may wonder whether he would have been
interested in dynamics of appropriation—the Achaemenids had a long tradition
of dealing with very different subjects on the latter’s terms, at least in part;
but his authority was the main point, and the mixed format is possibly simply
to be explained by the interference on the part of the Greeks in the redaction of
the document (to repeat, this is the only document showing traces of interfer-
ence between the letter and the decree—for good reason, since the letter is the
typical oriental format for official communication, and the decree the typical
Greek one).
Strikingly, notwithstanding these experiments and whatever the explanation for
these new attempts, the language of the Hellenistic Greek royal letters consciously
steers clear of some of the key terms used in the decrees of the poleis; and
especially when considered against the background of the possibilities for

21
Rhodes with Lewis 1997: 544 further note that the same could happen with Rome: they refer to an
edict of Caesar from Sardis, framed in the style of a decree of a Greek city (SEG 39, 1290. iii). Whatever
the reason behind this choice, authority will not have been it. For the relations between kings and cities
in Macedonia see the careful analysis of Mari 2006; Hatzopoulos 2006, with bibliography.
304 Letter Writing and the Polis

adjustments that we have just seen, this means that the difference between the two
formats must have been keenly felt.22
The best instance of the differentiation in the language of the two formats is
offered by the (non) use in royal letters of the verb ο, a staple of Greek
decrees: it had been used by Mausolus and by Berenice, and there might have been
scope for the use of some of its forms to introduce the decisions taken by kings.
Yet, this verb is exceedingly rare: out of seventy-five royal inscriptions recorded by
Welles, only five documents use a form of ο.23 Of these five, the first is a long
letter sent by Antigonus to all Greeks (we only have the copy sent to Scepsis),
announcing the peace concluded with Cassander, Lysimachus, and Ptolemy in
311 bc (Welles RC 1). Here, forms of ο appear three times, at the very end of
the document (ll. 65 and 70), after a long narrative in which Antigonus recounted
the events. They frame the actual request for all Greeks to swear an oath that they
would help each other to preserve a common freedom, an oath that, Antigonus
says, ‘we see as neither discreditable nor disadvantageous for the Greeks; therefore
it seems to me best for you to take the oath which we have sent’ ( PŒ ¼ - |  
P b I çæ   E  ‚ººÅ Ø øæH | Z. ŒÆºH   Ø ŒE åØ O ÆØ A
| e ‹æŒ  n K  ºŒÆ, 63–6). After a further remark in the first person
plural ( ØæÆ ŁÆ b ŒÆd N e º Ø e . . . , ‘In the future also we shall try to
provide both for you and the other Greeks whatever advantage we have in our
power’), Antigonus concludes (ll. 69–72):
 bæ c  ø ŒÆd ªæłÆØ  Ø | K ŒØ ŒÆd I EºÆØ @ŒØ  Øƺ - |  · çæØ b
E ŒÆd B › º ªÆ w - | ŁÆ ŒÆd  F ‹æŒ ı IªæÆçÆ. ææø Ł.
It seemed best to me then to write you also about these matters and to send to you
Acius to speak further on the subject. He brings you copies of the treaty which we have
made and of the oath. Farewell.
This is not worded as an order—the letter was a proclamation addressed to the
entire Greek world, a clever propagandistic stroke.24 The order was in fact part of
the treaty agreed upon; Antigonus here appears to express a wish (note that ο
does not directly govern the infinitive expressing the actual request, but a ŒÆºH
åØ that in turn governs the request to take the oath). This also explains the sudden
switch to the first person singular ( Ø twice, both times in connection with Œø):
while in the rest of the document Antigonus speaks in the first person plural, in part
also to include the other kings, here he expresses his personal opinion.25

22
This is not to deny that the two groups interacted, the letters of the one being the answer to the
embassies and decrees of the other, nor to deny that the two ‘formats’ shared in a common ‘language of
euergetism’, made of accepted key terms recognized as typical of each party: illuminating discussion in
Ma 1999: 182–94; Mari 2009.
23
Since Welles’s publication (1934) numerous other royal letters have been found—their number
has probably doubled. The lack of an updated corpus makes it difficult to extend the analysis to all royal
letters; but there is no reason to expect a different distribution. Meanwhile, for a more detailed analysis
of some distinctive features of royal epistolary language see Bencivenni 2011; Ceccarelli forthcoming a
and b.
24
See Welles 1934: 6–8.
25
Welles 1934: 10. In the entire document, Antigonus uses the first person singular three times
only: at l. 25, an incidental remark: ‘what zeal we have shown (  ŁÆ) in these matters will I think
be evident (çÆÅæe r ÆØ   ŁÆØ) to you and to all others from the settlement itself ’), and the two
 Greek Cities and Epistolary Writing 305

The second instance of ο for a royal decision is in the second (RC 4) of two
letters of Antigonus to Teos, both concerning the proposed synoecism of Teos
with Lebedos, dated to between 306 and 302 bc. Discussing the style of these
letters, Welles highlights the affinity in details to the previously discussed one,
concluding that it is very likely that Antigonus himself had a hand in the
composition of all of them; the two letters from Teos show occasional shifts to a
more personal and direct form of expression, just like the one sent to Scepsis.26
Although the content of both RC 3 and 4 definitely consists in orders, ‘the royal
will is expressed as if it were opinion (TØ ŁÆ E usually;  ºÆ  in ll.
30 and 93), or as if the king were an arbitrator (cf. K ØŒŒæŒÆ, l. 60)’.27 In RC 4
in particular, the order was probably expressed with [ . . . E s ŒÆºH Œ]E
åØ (ll. 7–8), that is, through a use of the verb Œø that does not strictly
correspond to the language of the decrees.
After these two examples, there is a huge chronological gap: evidently the
Hellenistic royal chanceries developed an independent language appropriate to
the new situation—or the instances just discussed were a consequence of Anti-
gonus’ own personal style.28 Œø next occurs in two letters of Eumenes of
Pergamum—and it is not actually used for royal decisions. The first letter (RC
52) is addressed to the Ionian league, and is dated to 167/166 bc; there the term is
used for the decision of the Ionian league to honour the king ( Ø æ  
E . . . çÆH ÆØ b A . . . , RC 52. 23). This decision Eumenes recounts
in great detail (unsurprisingly), before reacting to it and accepting the honours, a
resolution expressed with   ØÆ çغ [çæ ø I ]- | å ÆØ . . . (‘The
honours I accept kindly . . . ’, RC 52. 41–2). This letter constitutes thus one instance
of the king’s inserting into his own document the text approved by the League,
appropriating it for his own greater glory. The second letter by Eumenes where a
form of ο appears is a long document dated to the mid-second century
addressed to the guild of Dionysiac artists and aiming at resolving their dispute
with Teos. ο is used in column 2B, not for any decision, but simply to state the
impression that the king has formed of the situation and on the basis of which he
will decide:
ŒÆd a b ›º åæB æe c | IçØ Å Ø m K ŒØ  Ø Ø ØÆ åØ ŒÆd | Iç w
ÆNÆ ŒÆ Æ ı ÆŁBÆØ ÆF K . (RC 53, IIB ll. 7–9, Le Guen 2001 no. 47)
This then it seems to me is the meaning of the general issues which have led to the
dispute and the cause from which each arose.
The last instance allows us a glimpse of a royal ‘private’ correspondence: it is a
letter sent around 156 bc by Attalus II to Attis, the priest of the temple of Cybele at

occurrences of  Ø here. This is not to say that the present text does not represent a final decision; but it
is not couched as one.
26
Welles 1934: 32.
27
Welles 1934: 26, on the first letter (RC 3). It is interesting to notice that Antigonus, who expresses
his commands in this way, also uses it for the opinion (indeed, this time) of the envoys from Teos in RC
3 l. 56, while for the ambassadors of the Lebedians IØ ø is used. Œø is also used in RC 3 l. 33
(omitted from Welles’s index), not for command, but to mean simply ‘it does not seem that’ (and there
is no dative).
28
This should be integrated with a discussion of the Macedonian تæÆÆ, and an analysis of the
language of the decrees/ ØƪæÆÆ of Alexander. See meanwhile Mari 2006: 210–13; Hatzopoulos
2006: 75–96; Mari forthcoming.
306 Letter Writing and the Polis

Pessinus (RC 61), and is part of a correspondence that was certainly not meant to
be inscribed, since it deals with vital information; as Welles notes in discussing
this dossier, ‘in most cases the real messages between king and priest were carried
orally by messengers. The written word might incriminate, and was liable to
interception, but the spoken word was a secret as the faith of the messenger.’29
Those letters that were sent were kept, however, in an archive, and inscribed with
other documents of the same group more than a century after the events to recall
the earlier glories of the temple. Here, too, the æd z   E (‘concerning
what we resolved’, l. 6) concerns not an official decision, but a private one taken at
an earlier meeting between the king and the priest; this private decision the king
reported to his friends, who discussed it with him and in the end convinced him to
change his mind. The final decision is thus announced at line 20: ‘I thus decided’,
ŒæØ  s . . . If out of the seventy-five royal letters collected by Welles only the
two letters sent by Antigonus use a form of ο in order to express an official
decision, this implies, I suggest, a conscious avoidance of the term on the part of
the chanceries.30
A look at the attestations in this corpus of the other term that is a staple of the
idiom of Greek decrees, namely the conjunction K Ø  (‘whereas’, ‘since’) that
typically introduces the motivation clause, leads to the same conclusion. There are
all in all only four instances of K Ø ; K , the other term that may introduce
such a clause, is slightly more frequent (ten instances); but not all of these are used
for motivation—in fact, very few.31 In this case, the avoidance by the chanceries of
a formal motivation clause is probably due to the ideological desire to present the
decision of the king as free, or due to the king’s own personal considerations,
rather than resulting from any one event. I take only one instance, RC 44. 21, a
letter of Antiochus III to a governor concerning the appointment of a chief-priest
at Daphne, written in 189 bc. This is the context in which K Ø  appears:
We shall take care that in the future he will receive all that pertains to honour and
glory, for (K Ø ) as the chief-priesthood of Apollo and Artemis Daittae and of other
sanctuaries whose precincts are in Daphne requires a man devoted to us and compe-
tent to fill it suitably in view of the interest both our ancestors and ourselves have had
in the place and of our own reverence towards the gods, we have appointed him chief-
priest of these sanctuaries in the conviction that their administration will be carried
out properly by him. (RC 44, 18–31, Welles’s fairly literal translation)

29
Welles 1934: 246. On this letter as ‘private’ correspondence see Savalli 2003; Ceccarelli forthcom-
ing a. Contra, Virgilio 2011: 49–50 and 57.
30
Outside the corpus of Welles, a royal decision taken with   is recorded in the letter of
Antiochus III to Zeuxis quoted by Josephus, AJ 12. 148–9—but this is a problematic document, and
embedded in Josephus’ text (see Savalli 2003; Ceccarelli forthcoming a). The language of the ordin-
ances of the Ptolemies, collected in Lenger 19802, confirms the analysis proposed: N ŒE (or Ka
ŒBØ) appears three times in the Corpus des ordonnances des Ptolemées, always in requests addressed to
the kings, with the meaning ‘if you please’ (in 58. 17, request of the priests of Chnum, dated to 115 bc,
and repeated twice in 62. 16 and 19, and 63. 11 and 15, a request of the great embalmer at the Sarapeion
of Memphis, dated to 99 bc); no form of d ο ever appears in documents emanating from the king.
31
K Ø  occurs in RC 3. 79 and 44. 21 not in the context of a motivation clause; and at RC 7. 9 and
15. 21 for something that may look like a motivation clause. K  occurs in RC 1. 13 and 16; 3. 31 and
86; 21. 6; 36. 17; 54. 8; 60. 9; 65. 11 and 15; 67. 1; 73. 3; 75. 1. In the letter of Eumenes to Toriaion (SEG
47, 1745, 39–43) K Ø  introduces as motivation an earlier decision of the king; see Ceccarelli
forthcoming b.
 Greek Cities and Epistolary Writing 307

K Ø  is followed by a long clause in the genitive absolute, stating the specific

needs to be covered through the appointment; but formally the motivation is the
king’s action of having nominated a priest (K Ø  . . . I  åÆ), and not any
independent merits on the part of the priest concerned!32
The two forms of diplomatic language could have lived side by side without any
difficulties as long as the poleis preserved their independence; but, as pointed out
by Bertrand, the situation changed radically when the letter became the normal
means of communication of a king situated outside the polis, but expecting that
his orders would be followed and obeyed by the polis, and for this reason
addressing the polis directly in his writing.33 Now, the polis had to find stratagies
to integrate into its landscape this deeply extraneous form of discourse. Various
approaches could be taken: a polis could refrain altogether from inscribing a royal
letter—unless the king explicitly required that it be inscribed, or unless the letter
granted some privileges, for in the latter case, it might have been in the interests of
the city to inscribe the document.34 Apart from silence, various ways of dealing
with the royal text are attested.
The polis could publish the royal letter before, or after, a related city decree (the
one sent to the king, to which the letter was the response, or one proclaimed by the
polis in response to the letter). Thus the inhabitants of Seleucia in Pieria and their
epistates Theophilus reacted in 186 bc to the request expressed in a letter from
Seleucus IV that the bronze statue with which he was honouring his ‘honoured
friend’ Aristolochus be erected in their city, by inscribing a long decree of twenty-
seven lines, in which they expressed their acquiescence in the royal request and
granted citizenship and a place for the statue to Aristolochus; they then inscribed
below their own decree also the letter of the king.35 Sometimes the royal letter is
lost, but the text of the decree shows that the letter had been inscribed below, or
above, as in the case of a fragmentary decree of Telmessos (but found in Cos)
dated to c.220 bc, where the first four preserved lines state that
[K]ŒŒºÅ Æ ŒıæÆ ª  [Å] | ŒÆd ªæłÆ  ¸ı Øå [ı c] | æ ªªæÆ<>Å
_ _ _ _ ø BØ º[Ø]· (Segre I. Cos ED 56)
K Ø  º, |   º

32
That the nominee had been a valuable friend of the royal house had been mentioned in the
opening of the letter. Welles 1934: 182–5 offers a detailed analysis of the overall structure of the letter.
Again, the ordinances of the Ptolemies confirm the tendency to avoid direct causality. (There is only
one instance of K Ø , in a letter of Ptolemy II Philadelphus to Apollonius dated to 259 bc, Lenger
19802, no. 23; there are four instance of K , in Lenger 19802, no. 45. 7, where the meaning is not
causal; in no. 52. 22, a letter addressed by priests to the Ptolemy Euergetes, his sister Cleopatra, and his
wife Cleopatra; and in nos. 53. 85 and 54. 8, two collections of ordinances and amnesties by Ptolemy
Euergetes II, Cleopatra II, and Cleopatra III).
33
Bertrand 1985: 110: ‘Ce qui est rare [the letter] dans le monde des cités et ne pose guère de
problèmes véritables, ne l’est plus quand la lettre est le mode de communication ordinaire de la part
d’un souverain situé par nature hors de la polis, qui compte pourtant bien que ses ordres y seront reçus
et suivis.’ For the situation in Macedonia, I once again refer to Mari 2006; see also Mari forthcoming.
34
We touch here on one of the difficulties raised by the documentation: any dossier will inevitably
be skewed, and not just because of the accidents of preservation, but because various groups, states, and
individuals had an interest in whatever was inscribed, and would have tried to highlight some aspects
and to hide others. Bencivenni 2010 is a fascinating recent discussion of the position of the cities of Asia
Minor concerning the inscription of letters from the King.
35
RC, no. 45; cf. Wörrle 1978: 203 n. 1.
308 Letter Writing and the Polis
at the meeting of the main assembly, and since Lysimachus had sent the letter above
inscribed, it seemed good to the polis of the Telmessians. . . .
Alternatively, the polis could simply summarize the royal letter in the course of
its own decree, or give the gist of the letter in the decree, and publish the text as
well, preceding or following the decree. A remarkable instance is offered by a
decision taken by the city of Miletus in answer to a letter of Ptolemy II, dated to
c.262 bc: first on the stele came the royal letter (fifteen lines: Æ Øºf — ºÆE 
غŠø BØ  ıºBØ ŒÆd HØ øØ åÆæØ· . . . ææø Ł, ‘King Ptolemy to the
council and the demos of the Milesians, greetings . . . Farewell’); this was followed
by a short text (ll. 16–21) in which the assembly decided that the letter should be
read, and then a decision taken, after consultation, if required, with the person
who had brought it;36 finally came the decision of the city concerning the letter,
comprising thirty-seven lines (from ll. 22 to 59), of which nineteen were dedicated
to a detailed summary of the royal letter and of what the ambassador had said
(from ll. 23 to 42). The inscription thus begins with the letter—which in a sense is
correct, since it is the arrival of the letter that had caused all subsequent events; it
then relates what happened when the letter arrived, and closes with a long city
narrative, which includes the decision to inscribe it (the king had not asked for
this), and to send ambassadors to the king, to bring him copy of the decree.37
The third possibility is basically a variant of the previous one, consisting in the
insertion of the royal letter in the city decree, by literally quoting it. The best
example of this kind of response is offered by an inscription from Telmessos dated
to 279 bc: the text begins as expected, with the date and the mention of an
assembly (KŒŒºÅ Æ ŒıæÆ ª Å); it seamlessly goes on, with another
participle in the genitive absolute coordinated with the preceding one, to mention
the reading out of the royal letter (ŒÆd B Ææa  F Æ Øºø K Ø  ºB
Iƪø Ł Å), and then branches out into the royal message, with ‘in which it
was written’ (K wØ KªªæÆ  ) followed directly by the text of the letter: BÆ Øºf
— ºÆE  º ø BØ ºØ ŒÆd  E ¼æå ı Ø åÆæØ . . . O ø s N
øæa c ŒÆÆåøæÇÅ Ł,  E  Ææ H Kø º ªÅ ŁÆ ŒÆd æe  f
æd !غ ŒºB ŒÆd æØ  ºÅ ªªæÆ ÆØ.  0Eææø Ł (‘King Ptolemy to the city
and the archons of Telmessos greetings . . . That you are not to be assigned as gift
we have agreed with your representants, and it is stated in writing to Philokles and
Aristoteles. Farewell’). The decree resumes with the expected motivation clause,
K Ø c Æ Øºf — ºÆE  — ºÆ ı K åæÅ  . . . , ‘Whereas king Ptolemy
son of Ptolemy has granted . . . ’38 In 217 bc the people of Larisa dealt in exactly

36
Milet i, 3, 139B (cf. RC, no. 14): []  BØ  ıºBØ ŒÆd HØ øØ· ªÅ K Ø ÆH·  ¯ Æø
 E ØÆ ı r · | []a bª ªæÆÆ, L KŒ Ø   Hª æÆ  Ææa  F Æ Øºø_ — _ ºÆ [ı] |
_ _
[K] ªŒE eª ªæÆÆÆ B  ıºB N c _KŒŒºÅ Æ BØ æ Ø I[ ]- | [] ı  F K H 
Åe[ Œ]Æd IƪHÆØ HØ øØ, K ƪƪE b ŒÆd | 20 [ H]ª æÆ _  f K Ø []Æ[] K d c
KŒŒºÅ Æ, e b B  IŒ  Æ[Æ] | [] ıºF ÆØ a _ Œ FÆ ı  Ø. Cf. Wörrle_ 1978:
_ 203 n. 1,
with other instances.
37
As Welles 1934: 73–4 points out, it is clear that there had been difficulties in Miletus at this time,
and the ‘overdoing’ here may be connected to the situation.
38
SEG 28, 1224, see Wörrle 1978: 201–3, who comments: ‘Das Schema ist so consequent durchge-
halten, daß der ganze Text sich zu einer einzigen Periode zusammenschließt, die der Redaktor freilich
nicht ohne Mühe gemeistert zu haben scheint, wie man an dem Nebeneinander des einleitenden
absoluten Genetivs und der resümierenden K Ø -Konstruktion sieht, das wegen des Ausfalls der auch
 Greek Cities and Epistolary Writing 309

the same way with two letters of Philip V, possibly the most significant instance of
an attempt at integrating a royal letter into a city decree and of the difficulties
posed by such an integration. In Larisa too, the text of the royal letters is cited
directly after the prescript with the dating elements.39
Finally, a city could play on the location. This is the case of a fascinating decree
from Delphi concerning the request of asylia (inviolability) made by Smyrna in
244 or 242 bc. In the motivation formula, the Delphians acknowledged the
reception of a letter from Seleucus son of Antiochus in which the king asked
that the temple of Aphrodite Stratonikis and the city of Smyrna be sacred and
inviolable, ‘he himself (the king) having previously obeyed the oracle of the god’
and having already done what he was now asking the Delphians to do; the request
was reinforced orally by ambassadors from Smyrna, who ‘also ask, as does the
king as well, that what has been granted to them be inscribed in the temple’
(IƪæÆçB K HØ ƒæHØ, u [æ ŒÆd] | › Æ Øºf IØ E, a K ØŒåøæÅÆ
ÆP E·). The decision of the Delphians followed:
 åŁÆØ AØ ºØ H ˜º- | çH   ƒæe e A çæ Æ A "æÆ ØŒ  ŒÆd
a ºØ H ["ıæ]- | Æø ƒæa ŒÆd ¼ ıº  r , ŒÆŁ æ ‹  Æ Øºf
K  ºŒ [ŒÆd] | ± H "ıæÆø ºØ IØ E, KºÆ ÆØ b ŒÆd  E Łøæ E
 E[ a] | 15 —ŁØÆ K ƪªºº  Ø K ÆØ ÆØ e Æ ØºÆ "ºıŒ  K    [ Ø]
| ŒÆd AØ P ÆØ ŒÆd HØ K ÆŒ º ıŁÅŒÆØ HØ  F Ł F åæÅ HØ, | ŒÆd ŁF ÆØ AØ
çæ ÆØ· IƪæłÆØ b e b łçØ Æ   a ºØ | K HØ ƒæHØ  F Ł F, a ’
K Ø  ºa K HØ IæåøØ K HØ  åøØ. (Fouilles de Delphes III, 4. 153, cf. Rigsby
1996, no. 7)
Let it be resolved by the city of the Delphians that the sanctuary of Aphrodite Stratonikis
and the city of the Smyrnaeans be sacred and inviolable, just as the king has sent and the
city of the Smyrnaeans asks; that it be enjoined to the theoroi who will announce the
Pythia to praise the king Seleucus for these actions and for his piety and for having
followed the oracle of the god, and that they sacrifice to Aphrodite; that the city inscribe
this decree in the sanctuary of the god, but the letter in the archive on the wall.
As Rigsby notes, the civic decree is inscribed ‘in the temple’ as the king and
Smyrna had asked (or to be precise on the back of the temple); this is the text we
have. The Delphians, however, decided to publish the lost royal letter in the city
proper, on the wall of the archives building, thus dividing what was otherwise a
unitary dossier.40 Size was also an option. When Magnesia on the Maeander
decided to inscribe the letters sent by those poleis and kings who had accepted
its proclamation of the festival of Artemis Leukophryene and granted the city
asylia, the letters of the kings were inscribed in larger letters than the decrees of
most of the poleis, and in a special location.41

in Telmessos sonst üblichen Sanktionsformel (  º ø BØ ºØ etc.) unorganisch wirkt’;
Bertrand 1985: 111.
39
Syll.3 543 (translation in Bagnall and Derow 2004, no. 32); see Wörrle 1978: 203; Bertrand 1985:
111–12, with discussion of the influence of the king against the civic decision; Mari 2009: 91–2.
40
Rigsby 1996: 105. It is unclear which oracle Seleucus had followed: Apollo certainly, but not
necessarily the Delphian one: he could have consulted Didyma or Claros (Rigsby 1996: 104). But the
vagueness would have suited the Delphians: the default assumption for any reader would have been
that the king had followed the oracle of Apollo of Delphi.
41
See Kern, I. Magnesia 1900, p. xxx: ‘Die Schrifthöhe ist verschieden, am größten bei den
Königsbriefen (im Durchschnitt 0.012) und bei dem gefälschten Kreterdekret, kleiner (etwa 0.01) bei
310 Letter Writing and the Polis

Until now I have limited myself to describing the possibilities open to a city
receiving a royal letter. That the royal language was formally extraneous to that of
the polis, although the kings at the same time shared with the polis in the language
of euergetism, can, I believe, be accepted; but it remains to show why a royal letter
(or any letter, for that matter) should prove disturbing. The theoretical reason for
this is grounded in the very concept of epistolarity: for epistolary discourse is
marked by the presence of two entities, an ‘I’ and a ‘You’, and by the foreground-
ing of a relationship between them. Sending a letter is moreover always an attempt
at pushing the addressee into accepting the relationship in the first place, at
eliciting from the recipient a response.42 How this worked in practice in the
power relationship of the Hellenistic kingdoms and the Greek poleis is shown
by an example discussed by John Ma in his seminal analysis of the interactions of
empire. In a letter sent to the city of Iasus in c.196 bc, the Seleucid queen Laodice
informed the Iasians of her decision to donate to the city a large quantity of wheat.
Unsurprisingly, the queen offered in her letter a very positive depiction of herself
and of Seleucid power, represented at the time by Antiochus III. What of the
image of the Iasians? Laodice states that she has often heard from her brother
Antiochus how ‘after having recovered your city, as it had fallen into unexpected
calamities, he gave back to you your liberty and your laws, and in the other
matters he strives to increase the citizen body and bring it to a better condition’.
For this reason, Laodice has decided to follow in his footsteps and to donate 1000
Attic medimni of wheat a year for ten years—and to promise other donations, if
the behaviour of the Iasians towards her brother and the royal house will conform
to expectations, and if they will gratefully remember her donations.43 As pointed
out by Ma, in accepting the presents, the Iasians must also accept the royal version
of their history, of a city reacquired (and not conquered), of a freedom that
depends on the royal benevolence. The queen in her letter had sketched the
image of a Iasus such as the Seleucids wanted it to be, and the Iasians not only
found it expedient to accept this new version of their history, they also inscribed it
on stone; beside it, however, they inscribed a decree presenting a slightly different
version of the events, ‘rephrasing’ the Seleucid viewpoint.44 A royal letter, and
moreover one offering a donation over ten years, with the promise of further
largesse on top, can take some liberties in drawing the image of its interlocutor.
And a letter is the perfect diplomatic instrument for such purposes, since a letter is

den übrigen Inschriften’. The approval of the kings carried the acceptance of the festival also by the
cities of the kingdoms, hence the privileged position given to their letters. On the dossier from
Magnesia see Ceccarelli forthcoming b.
42
See in general Altman 1982, esp. 117–21. Personae of course are adopted by or attributed to
others also in the most banal of face-to-face conversation, but they are not fixed, inscribed on stone for
everyone to see. In a decree, the persona of the polis speaks, alone and to no one. But with letters, an
individual reaches out, and a polis who writes or accepts letters enters into a relationship with others—
something that could not have been easy for the fundamentally autarkic Greek poleis. A further issue,
referentiality, is discussed in Ceccarelli forthcoming a and b.
43
I. Iasos 4, see Ma 1999: no. 26A, ll. 25–9: ªØ  Ø Ł’ E Y  e | I ºçe ŒÆd ŒÆŁ º ı e
r Œ  H ¥ ı ŒÆŁŒØ | [ŒÆ]d H I Æø<>ø Pæª ØH Å Ø | [P]åÆæ ø
Øæ ÆØ ŒÆd ¼ººÆ L i K Ø H ı- | [ Ææ]Æ ŒıÇØ.
44
Ma 1999: 180–2, 196–8, and 216–17 (217 for the citation). Iasus may in fact never have been
under Seleucid control before 201 bc: Ma 1999: 199.
 Greek Cities and Epistolary Writing 311

directly addressed by an ‘I’ who can be presented as the author chooses, to a ‘you’,
singular or plural, that the letter recognizes as such, to whom it gives a seeming
autonomy, a notionally independent existence, but that it also shapes as the
author pleases. By answering in a different language, through decrees, the Greek
poleis kept in touch with the royal discourse, but avoided being fully drawn into it.

7. 2. L E T T E R S B Y G R E E K P OL E I S

And yet, notwithstanding the existing and accepted distinction between the letter
as the means for royal communication at all levels (within the administration, and
in dealing with cities) and the civic decree, and notwithstanding the suspicions
surrounding the appearance of letters in the official, public sphere, especially
when relating to interstate exchanges, quite a few poleis wrote letters. The
phenomenon of letters from poleis has not received much scholarly attention;
yet it is quite important.
Bertrand, the scholar who has dedicated most attention to the problem of the
formal relationship between royal letters and civic decrees, ignores this aspect. He
correctly states that ‘Les grammata que transmettent les cités à d’autres cités pour
demander que soit publié dans leur territoire un document, pour étayer leur
position (dans le cas d’un arbitrage) ne sont pas nécessairement des lettres’
(true: in most cases, they seem to have been decrees); but he then goes on to
write that ‘Quand le roi Ptolémée dit avoir reçu une lettre des citoyens de
Telmessos, c’est par contamination avec sa propre pratique; ce que les ambassa-
deurs lui ont transmis est sans doute la couronne et le décret qui la décernait,
accompagné de considérations annexes’.45 In this specific instance, this may—or
may not—have been the case; but the list of letters sent by poleis and published on
stone given below in appendix 3, as well as references to letters by poleis in other
inscriptions or in literary texts, show that some poleis made use of letters for
official communications.46 From at least c.242 bc, and probably from at least

45
Bertrand 1985: 109 n. 53. The king (Ptolemy II) speaks explicitly of a ‘letter’: I  øŒÆ E ƒ
I ƺ | Ææ H æ ıÆd e çÆ- |   ŒÆd c Ææ H K Ø  º, SEG 28, 1224 ll.
9–11 (cf. Wörrle 1978; see above, 308). The decree within which the letter of the king is quoted is dated
to 279 bc; this is thus the earliest reference on stone to a letter from a polis. Ma 1999: 179–242 (‘Empire
as Interaction’) focuses on the interaction between royal letter and civic discourse, but is not interested
in the uses of letters per se.
46
Literary texts: e.g. the letter of the magistrates of Laodicea to the proconsul Gaius Rabirius, dated
to 47–46 bc, Jos. AJ 14. 241–4. Other inscriptions: the Ptolemaic strategos Thraseas, in the letter he sent
some time after 238 bc to the city and archons of the Arsinoeis (SEG 39, 1426), stated that he had
received a letter from them (l. 2, [ ¯ Œ] Ø ŁÆ c Ææ H K Ø  ºc). The reference to ªæÆÆ
in the letter of Ptolemy III Euergetes_ to Xanthos of 243/242 (  ªæÆÆ I  øŒÆ, ‘and they
handed in documents’, SEG 36, 1218, l. 10) need not necessarily imply a letter; similarly for the
reference to ªæÆÆ in the regulation from Euromos (SEG 43, 707, ll. 12–15), discussed below, 327.
Rubinstein 2013 offers a detailed analysis of the diplomatic conventions of inter-polis relations in the
Hellenistic period; she focuses especially on the role of the envoys and on the impact of oral
performance, but emphasizes the wide range of possibilities open to each polis.
312 Letter Writing and the Polis

279 bc, the date when Ptolemy says that he received a letter from the Telmessians
(if we accept that Ptolemy meant what he was saying), to the beginning of our
era—the moment where I have, somewhat arbitrarily, decided to draw a line—
more than forty-five poleis used, at one time or another, the letter form to
communicate with other poleis or groups; and many of these letters were con-
sidered important enough to be inscribed.47 Rome is of course part of the group—
it is after all a Hellenistic polis. A very high number of letters was sent by the
Romans, and within the same time-frame, a minimum of seventy-one instances of
such letters were inscribed in the East; to avoid skewing the data, I have listed the
Roman letters separately, at the end of appendix 3.48 Without taking into account
the letters sent by Rome, we are still left with a total of some seventy-four inscribed
letters from Greek poleis or koina for the period before our era.
A quick analysis shows that the documents fall into two basic categories: self-
sufficient official letters that on their own convey a decision, specific information,
or a request; and the so-called ‘covering letters’, sent to accompany a decree. This
type is rather frequent (leaving out the excessively fragmentary instances, under-
lined in appendix 3, thirty-five certain occurrences). Of course, since the covering
letters do not accomplish any specific function besides that of accompanying the
decree, the polis or the group at the receiving end could decide (unless there was a
specific request by the sender, and unless the sender had the power to enforce it)
whether to inscribe everything, whether to put everything into their archive, or
whether to put the covering letter in the archive and inscribe the decree only. This
means that the overall number of covering letters effectively sent might have been
far higher than what we in fact have; it still remains the case that we begin to find
letters from cities (or mention of letters from cities in epigraphical documents)
only from the second quarter of the third century onwards.
A close analysis of the text of these letters shows that the cities have no qualms
about drafting letters to convey other documents: the covering letters are felt to be
unproblematic.49 The situation is different for the ‘self-sufficient letters’, which

47
The number is based on the list in appendix 3; but it is here only to give an idea (in some cases the
name of the sender is lost; the cities that decided to engrave letters from poleis could for instance also be
added to the tally, inasmuch as they would have been familiar with that form of official communi-
cation). The practice of course continued, as shown by the letter of recommendation for one of their
own citizens sent at the end of the second century ad by the city of Tlos to the prytaneis of nearby
Sidyma, TAM II 174, ll. 1–31 (cf. SEG 50, 1356), or by the letter sent by the ephors and the city of
Gythion to the ephors and the city of Pyrrhichion, announcing the death of one of their citizens (IG V
1, 1524, of imperial period).
48
On Roman official letter writing see the collections of Viereck 1888 and Sherk 1969, as well as
Ferrary 2009, Hofmann 2013 and Osborne forthcoming, for the Republican period; Hurlet 2010, for
the imperial period. I shall not discuss Roman letters here.
49
Thus in the decree honouring Hermias of Stratonicea published in Delphi c.85 bc, the city of
Daulis stipulates that they shall send a letter to Stratonicea, appending to it the decree ( łÆØ . . . d
"æÆ ØŒE . . . [ Ø]ªæłÆÆ fi A K Ø  ºfi A e ªª e | [ Ææ ÆH Ø   Eæ]Æ fi . . . , Fouilles de
Delphes III 4: 69, ll. 24–5 = Canali De Rossi 2002, no. 141, ll. 24–5; similarly the Delphians in the first
century bc decide to inscribe the honours awarded to Artemidoros of Mazaca in the most visible
location at Delphi (this is the document we have), and add that ‘a copy of this decree be sent to the city
of Mazaca through a letter sealed with the public seal, in which the archons of the city state these
honours’ (Fouilles de Delphes III 4: 59, ll. 16–19 = Canali De Rossi 2002, no. 142, ll. 16–19: Iƪæ[łÆØ
b   a] Ø[a K fiH K ]ØçÆ ø fi [ ]- | [ fi ø  F ƒæ F  F  ººø , ŒÆd I EºÆØ e
IªæÆç   F łÅç Æ   F]- | [  æe ]c ºØ H [ÆÇÆŒÅ]H Ø’ K Ø[ ] ºB[ ] e
[ çæ]ƪE[ Ø Bfi Å ]Æ fi , | [Kªª]æłÆÆ  f ¼æå Æ[ ]B ºø   a Ø. Examples
could be multiplied (see above n. 46).
 Greek Cities and Epistolary Writing 313

present enunciative shifts that betray unease, not to say a difficulty, and that are
absent from both contemporary private correspondence and royal correspond-
ence. This second type of official letter from a polis can be further subdivided
into letters conveying information, or a request (relatively unproblematic); letters
conveying a decision; and documents that seem to offer a mixed format: they are
basically decrees, but at their conclusion present the epistolary closing formula
‘farewell’, ææø Ł. It is worth noting that all the examples of this typology are
from Cretan cities; in general, Cretan poleis seem to have made a relatively
frequent use of the epistolary format in interstate communication.
Let us first look at how covering letters function. A stele from Cos (no. 29 of the
list given in appendix 3), preserving a letter and the beginning of a decree of the
Eretrians, offers a representative instance of the type of dossier introduced by a
covering letter:
[æ  ı(?)]  F æ  ı [ F ˜Æ] ŒºF !غ çæ    F —Ææ Œ ı Øæå ı
 F Ø Æ (three _ crowns) _ __ _
 ¯ æ[æØø] ƒ ¼æå  ŒÆd  ºØ ˚Øø BØ |  ı[ºBØ ŒÆd] HØ øØ åÆæØ· H
KłÅçØ - | ø[ ç ]H ØH HØ  øØ H ŒÆd | 5  E[ I ]ƺE Ø
ØŒÆ ÆE I  ºŒÆ- |  [E I]ªæÆç  çæÆªØ  Ø BØ Å - | ÆØ
[ çæƪE] Ø ¥ Æ ÆæÆŒ º ıŁB. ææø Ł. | vacat. |
[K Ø c ]łø H łçØ Æ ŒÆd ØŒ<Æ> - | 10 [ƪøª f Œ]Æd æ ıa æe
e B  e | [˚Øø ˜Å]æØ  ¯PŒ  , ¸ı Æ ˚º - | [․․․․․
ÆæÆŒ]ƺ Æ ˚Ø ı FÆØ E_ | [ ØŒÆ a ¼] æÆ æE ¥ Ø Ææƪ -
| [ Ø N_  ¯ ææØ]Æ Ø ı Ø ŒÆºH ŒÆd Ø- | 15 [ŒÆø a Œ]Æ ŒÆd ıçæ ø
[ ¯ ææØF]- | [ Ø ŒÆd  E K Ø]Œ F _Ø K [BØ ºØ, ․․c.7․․] | [․․․․․ __
c.14․․․․․]ˇ$¯ [․․․․․c.18․․․․] (IG _ xii 4, 169)
_ _
(Three crowns, and above each, the name of each honorand)
[Aristos?] son of Aristos of Damokleus; Philophron son of Parmeniskos; Timarchos
son of Timidas.
The archons and the city [of the Eretrians] to the [council and] the people of the
Coans, greetings. We have sent to you a copy of the honours voted by us to your
people and to the judges you sent, authenticating it with the public seal, so that you
may know. Farewell. Vacat
Whereas having sent a decree and dikastagogoi and ambassadors to the people of the
Coans, Demetrius the son of Euctemon, Lysias the son of Cleo-[ – ], to [call] on the
Coans to provide for us [as judges] three men who on their arrival to Eretria would
settle well and justly the lawsuits and in the interest of the Eretrians and those living in
the city, the . . .
Here we find, after the three crowns highlighting the names of the honorands, a
letter presenting, after the usual epistolary prescript, the expected shift to first
person of the sender and second of the addressee, plural in both cases since the
magistrates and city of Eretria are collectively addressing the council and people of
Cos; these deictic personae are maintained until the end of the letter. The decree
then follows. Uncommonly, it begins directly with the motivation clause: there are
no dating elements. Even though the prescript is omitted, the text as a whole flows
smoothly: because of the preceding letter, the actors are already known; but
interestingly (and logically, since the motivation follows directly from a letter)
the motivation clause refers to the Eretrians with first person pronouns (H
and E, lines 9 and 12). This is usually not the case in Athenian decrees: the
314 Letter Writing and the Polis

motivation clause is throughout in the third person. And (more pertinent to the
case at hand, since this is a letter from Eretria), this is not the case in Eretrian
decrees: the honorands are said to have typically shown their goodwill to the polis
and the demos, not to ‘us’.50 Of course, whoever had spoken in the assembly to
promote the honorands’ cause would have talked of ‘our city’; but at the moment
of the redaction of the decree, the deictic pronouns would be eliminated. Here,
possibly because of the ‘letter effect’, two first person plural deictics have re-
entered the motivation (the decree would have had ‘the polis’, or the Eretrians, as
we indeed find later in the text, lines 14–16). After this part, there would probably
have been a hortative formula, and then the decree would have moved towards the
resolution. This part is lost, but in it we would probably have found a third person
of the body decreeing (the Eretrians) and a third person of the addressee of the
decree. Three elements in this documentary dossier are, however, surprising: the
first one, already mentioned, is the omission of the prescript with the usual dating
elements at the beginning of the decree. This may have been already so in the text
the Eretrians sent the Coans, or it may reflect a decision by the Coans, by the
stone-cutter, or by some magistrate.51 One can only advance guesses as to the
reason why. The second is the presence of the first person plural pronouns in
the motivation clause, which may depend on the presence and influence of
the covering letter. As for the third surprising element, it may explain why the
covering letter was inscribed. Crowther noticed, in his discussion of the text, an
anomaly: the letter mentions the award of honours to the people and the judges;
but the three crowns each bear the name of an individual judge, so that there is no
place for a separate crown acknowledging the honours to the demos in the
heading of the stela.52 At this point one may suspect that the Coans decided to
engrave the letter, because it mentioned the honours for their own demos, absent
from the text of the decree.53
Another instance in which a covering letter adds information is the letter sent
by the archons of Elatea to Delphi, to accompany a manumission (no. 32 in
appendix 3). The important document is of course the manumission (ll. 8–28); but
the covering letter contains the request to the Delphians to inscribe the other
document, and the information that the manumission had also been inscribed on
the sanctuary of Asklepios at Elatea. The archons of Elatea thus sent two official

50
One Athenian instance of the early Hellenistic period, IG ii², 495, 303–302 bc: K Ø c ºŒÆE 
 HæÆ- | [ ]ı `YØ  ØÆæø Ææa HØ Æ Øº- | []E ˜ÅÅæøØ ØƺE æø IªÆ- | [Ł]e ‹Ø
ÆÆØ ŒÆd º ªøØ ŒÆd æªø[Ø] | [ ]æ   f IçØŒ ı ı N ÆØ []- | 15 [ŁÅ]Æø æe e
Æ ØºØÆ ŒÆd Œ Ø[]- | [BØ] æd e B  e ŁÅÆø. One Eretrian instance, of approximately the
same date as the letter, IG xii 9, 222: ˝ØŒø & ØæBŁ ˚Œ ı r []· | K Ø c !غ   Icæ IªÆ- | Ł 
K Ø æd c ºØ c[] |  ¯ æ[æØ]ø,   E  ıºE | 5 ŒÆd  E  Ø Œº.
51
The same omission is visible in another honorific _ decree for Coan judges, found in Cos but
emanating from Chalcis: Crowther 1999: 284–93, no. 8 (IG xii 4, 168). Possibly in the papyrus version
that arrived in Cos this one too had been provided with an accompanying letter. But note Savalli 2010:
129 on the fact that inscribed documents do not always and systematically bear an indication of the date.
52
Crowther 1999: 299. Note that in the decree sent by Chalcis mentioned in the note above, the
decree’s heading has three crowns, but only two judges were honoured: the third crown was left for the
demos of the Coans (and no covering letter was inscribed).
53
There may of course have been a separate decree honouring the demos. Note that from the
preserved motivation clause of the decree it appears that the affair was initiated in the standard way,
through sending a decree and ambassadors.
 Greek Cities and Epistolary Writing 315

documents, a manumission and an official request, couched in the form of a letter,

to have the latter inscribed; their request also contained information as to the
publication of the manumission in Elatea itself. The Delphians might still have
chosen to inscribe the manumission only. They did not, and we get a glimpse of a
type of letter that must have been relatively frequent, containing not a decision,
but a (polite) request, often phrased, as here, with variations on the formula ‘you
will do well’, s s Ø . Within the list in appendix 3, a similar request for
inscription is also made in the covering letter accompanying the decree with
which the Cnossians had voted honours to Magnesia on the Maeander
(I. Magnesia 67, no. 8 in appendix 3).
This example is very close to those self-sufficient official letters that, on their
own, without referring to other documents, convey information or a request.
Their number is relatively limited (I count eight, leaving out the cases in which the
fragmentary status of the stone makes it impossible to decide: nos. 9, 10, 16, 37, 38,
43, 62, and 63 in appendix 3)—but then, there might not have been strong reasons
for inscribing them. The two letters sent by the kosmoi and the polis of Gortyna
and Hierapytna to the kosmoi and polis of Itanos, to warn the latter of the
movements of the Praisians (nos. 37 and 38), are actually only mentioned as
archival documents, retrieved (and then inscribed) in the context of a later
arbitration between Hierapytna and Itanos. The other instances comprise the
request sent by the Kytenians to help them reconstruct the walls of their city
(no. 10, where the request is expressed through a form of IØÇø = IØ ø, ‘think
right, expect, resolve’); the letter sent by the Cretan city of Axos (no. 16),
informing the Aetolians of the situation of their citizen Epikles, and requesting
that any harm to him and his family may be prevented at both the public and the
private level (a request expressed with ŒÆºH t ØÅ E çæ  - | 
‹ ÆØ . . . , ‘you will do well in taking care that’); finally, notwithstanding its
fragmentary status, the letter sent by Tyre to Delphi with information on some
event (no. 43) may also be counted this group, as well as the exchange between the
Gephyraioi and Delphi (nos. 62–3).
More interesting are the letters that express a decision (some eighteen docu-
ments, nos. 1, 2, 4, 5, 6, 11, 12, 13, 14, 24, 26, 36, 45, and 70–4; in a few instance,
the fragmentary status does not allow precise examination of their wording, but
the overall context makes the meaning plain). One instance will do here, the letter
sent by the polis of Polyrrhenia to Teos for the asylia in 201 bc (no. 12):
— ºıææÅø | — ºıææÅø ƒ Œ  Ø ŒÆd ± ºØ Åø HØ øØ | ŒÆd AØ øºAØ
åÆæØ· Œ Ø  Ø e łçØ Æ e Ææ | H Iªø ŒÆd H æ ıA
 ºº - | 5 ø{Ø} ŒÆd ˚øºÆ{Ø} ØÆŒ  Æ Ææƌƺ ø a |  Æ
ı A ŒÆd çغ ØÆ IŒ º Łø  E K HØ | łÆç ÆØ ŒÆÆŒåøæØ  Ø· æd
c  ø  - | åŁÆØ — ºıææÅø  E Œ  Ø ŒÆd AØ ºØ I ŒæÆ - | ŁÆØ
Å Ø Ø Ø e ˜Ø ı  ŒÆd ÆP d  ŁÆ ŒÆd  - | 10 ÆØ HØ ŁHØ [ Ææ ±]H
  ºØ ŒÆd a åæÆ IØ | ƒæa ŒÆd ¼ ıº  F  ŒÆd N e – ÆÆ åæ  ·
q b | ŒÆd [  Ø ] E K ø
fi I çºØÆ ŒÆd ŒÆa ªA ŒÆd ŒÆa | [ŁºÆ ]Æ K[] e
– ÆÆ åæ  . ææø Ł. (IC 2 xxiii 3)
Of the Polyrrhenians. The kosmoi and the city of the Polyrrhenians to the people and
the council of the Teans, greetings. Having received your decree we read it and we
heard the ambassadors Apollodotus and Kolotas who invited us with great eagerness
and zeal in conformity with what was detailed in the decree; concerning these matters,
316 Letter Writing and the Polis
let it seem good to the kosmoi and the city of the Polyrrhenians to answer to the Teans
that we do ourselves respect Dionysos and that recognition of the city and the country
as sacred and inviolable, now and for all time, is granted to the god from us; and may
all Teans be safe both on land and sea for all time. Farewell.
The letter begins with an epistolary prescript in the third person of sender and
addressee. This is followed by the expected enunciative shift to the first person
plural for the sender, and to the second person plural for the addressees (the
citizens of Teos). But with the resolution formula (  åŁÆØ) a further shift
intervenes: for a moment, when relating the decision, the Polyrrhenians refer to
themselves in the third person; more importantly, they now answer the Teans,
referring to them in the third person, instead of the second-person ‘you’ as before.
The Polyrrhenians go immediately back to a ‘we’, but maintain a perspective in
which they speak of the Teans in the third person. It is only in the final greeting,
ææø Ł, that the second person plural, referring to the addressees, the Teans,
reappears. What appeared as two distinct moments in the ensemble formed by
covering letter and joined decree is here brought under one heading.
The same happens, less visibly, in a small group of five documents from Crete
(nos. 70–4) that are from all points of view decrees with an appended greeting,
ææø Ł; for here, until the end, readers have no way of knowing that what they
have in front of them is actually a letter; only retrospectively can the text be
construed as such.
The use of a resolution formula with  åŁÆØ in these letter-decrees is worth
stressing: this is exactly what the royal chanceries had tried to avoid. In general,
letters from cities presents formulations that are relatively close to the language of
the decrees.54 Out of the eighteen letters that convey a decision (the only relevant
here), three (nos. 5, 11, and 14) are too fragmentary to be of use; seven (nos. 1, 4, 6,
12, 24, 26, and 36) use a form of Œø (mostly  åŁÆØ) for the resolution; to
these seven must be added the five Cretan documents (nos. 70–4) that appear as
decrees with a greeting at the end, and which also use forms of ο for the
resolution. Thus, only three official letters that also express a decision, the letter
sent by the Cretan cities of Phaistos and Kydonia to Cos and Teos respectively
(nos. 2 and 13) and the one sent by Delphi to Athens (no. 45) deviate from this
pattern. The letter from Phaistos has simply a string of infinitives; the letter from
Kydonia too omits the resolution formula altogether, and moves directly to the
decision itself; in the third letter, KŒæÆ is used to express the decision, but
the Delphians are here simply summarizing and confirming an earlier decision for
the benefit of the Athenians.
The pattern of use with respect to the motivation formula is only slightly
different. Out of the eighteen letters conveying a decision, three letters (nos. 5,
6, and 13), to which must be added the five Cretan letter-decrees (70–4), introduce
the motivation of the decision with K Ø  (moreover Phaistos, n . 2, has K );
interestingly, Kydonia, which had avoided a form of Œø, uses K Ø . As for

54
Moreover, in the letters nos. 1 and 2, the document itself is referred to as łçØ Æ (łçØ Æ) (e.g.
when stipulating the inscription on stone), and similarly in no. 11 as e ªæÆçb ªÆ. Conversely, the
prytaneis of the Milesians (no. 39) name the covering letter they are sending (which gives actually quite
a few details) an K Ø  º.
 Greek Cities and Epistolary Writing 317

those self-standing official letters that introduced a request or transmitted infor-

mation, the forms used to ask for the favour (or to give a warning) are variations
on s s Ø  (‘you will do well’); the reason for the request may sometimes
be given through the use of K Ø , but genitive absolutes or independent state-
ments are also often used. The letters from the cities thus behave on the whole
more freely than the royal letters: they may make use of the polite language
employed in the royal letters (ŒÆºH or s s Ø  is a typical way for a
king to give an order), but they often rely on the language of the decrees.
In order to appreciate the importance of the use of letters within ancient
diplomacy and to evaluate its place in the context of the administrative practices
of the Greek poleis, it remains to determine which poleis sent letters, on what
occasions, and why. The discussion that follows will rely on the collection,
documented in appendix 3, of all letters by Greek poleis or groups preserved on
stone. This material will occasionally be integrated with instances of official letters
attested in the literary tradition; but, as already stated, it is the very fact of their
inscription on stone which makes these letters particularly significant, because it
put them on display, marking their significance for the collectivity. As a result,
equal attention will be given in analysing this material to the sender of the official
letter and to the locality where the text was engraved: for the place where it was
decided to inscribe the letter and the polis or the group responsible for its
inscription are as important and significant as the sender. A specific location
had to be designated for the erection of the stele, money had to be paid, and
publicity would result: some thought would have to be given to what exactly to
post and where.55 If letter writing for official reasons was part of the adminis-
trative practice of the recipient (polis or group, this is not important), this might
have made the inscription of a letter (and even just a covering letter) more likely.
There must have been a significant volume of official epistolary exchanges, both
between a polis and strategoi or ambassadors on a mission, and between poleis;
but in the majority of cases these letters will have been destroyed after the message
had been conveyed, or will have been kept in an archive (and in this case they may
then resurface in the literary tradition, or in later documentary dossiers). Only in
very rare instances would these epistolary exchanges have been deemed worth
recording on stone, and it is worth seeing why.56
Let us begin with the geographical distribution of the inscribed non-royal
official letters.

7.2.1. The Cretan Poleis

A first, relatively consistent group is formed by the Cretan poleis: altogether,

fifteen letters (nos. 1, 2, 4, 5, 6, 8, 11, 12, 13, 14, 16, 24, 26, and 37–8), sent by
twelve cities, Istron, Phaistos, Lappa, Cnossus, Gortyna, Sybrita, Polyrrhenia,
Kydonia, Hierapytna, Axos, Allaria, and Aptera, to Tenos, to Cos (four letters),
to Teos for the request of asylia (five letters), to the Aetolians, to Magnesia on the

55
Liddel 2003; Ma 2009.
56
See above, 298, for the importance of paying attention to the epigraphic habits of the various
poleis.
318 Letter Writing and the Polis

Maeander, to Paros, and to Itanos in Crete (two letters). If we count also the five
‘mixed’ cases from Crete (from Axos, Arcades, Hyrtacina, and Priansos, all sent to
Teos in answer to its request of asylia, two in 201 bc, three in 170 bc), the Cretan
group begins to have a remarkable weight (twenty instances altogether, out of
which only one is a covering letter, accompanying a decree; if Aptera sent a letter
rather than a decree to Teos in 201 bc, the total rises to twenty-one). The ‘mixed
cases’ are definitely more meaningful than a simple covering letter, even if they
present only an epistolary greeting and do not have an epistolary prescript, since
they imply an interpenetration of letter and decree. Finally, we may want to
consider whether the letter sent by the polemarchs and the synhedroi of the
Thebans to the kosmoi and the polis of Polyrrhenia (no. 15) should fall within
this group, since the decision to have the letter inscribed may be counted as an
instance of the epigraphic habit typical of the latter city.
However, there are some problems: not only did almost all of these poleis also
send decrees to other poleis; in most cases the writing of decrees is attested within
a close chronological range, so that it is impossible to postulate an evolution or a
constitutional change.57 A closer scrutiny shows that of the twenty instances, ten
concern the Tean request for asylia: five are ‘real’ letters and five others take the
form I defined as ‘mixed’ (among these, three concern a later request for renewal).
It may seem reasonable to suppose that in this specific case either the request of
the Teans, as presented by the group of theoroi that visited the cities in question,
or the format in which one of the first Cretan cities visited by the Tean theoroi
answered, shaped the format in which other Cretan cities answered.58 It should be
added that the first Tean request had been endorsed by both Philip V of Macedon
and Antiochus III: ambassadors from both accompanied the theoroi, and the
desire to please the ambassadors and through them the kings is at times men-
tioned in the answers of the Cretan cities. This at any rate means that around 200
bc the Cretan cities did not have strong views on the formal shape they wanted to
give their documents (a hypothesis actually supported by the variations in the
order and formulations of the decrees), and in particular, that they did not have
problems with letters.59

57
Lappa, for instance, sends in the second half of the third century bc a letter to the Tenians (no. 4,
IC 2 xvi 2); but in 201 bc the same polis sends to Teos, in answer to the Teans’ request of asylia, what
seems to be a decree, IC 2 xvi 3 (both documents are however fragmentary). Allaria at the beginning of
the second century bc writes to the Parians a letter (no. 24, IC 2 i 2 B); but a decree had been sent to
Teos in 201 (IC 2 i 1).
58
A textbook case of Empfängerformular, although I think it was rather an answer that shaped most
other answers (and, for the renewal of the asylia in 170 bc, the earlier answer may have provided the
basis: for instance, Aptera might have sent a letter to Teos already in 201 bc, and another letter in 170 bc,
cf. appendix 3, no. 26); otherwise, there would be no Cretan decrees for the request of asylia at all, and
especially not by poleis who otherwise also could send letters (see n. 57 above). On the bureaucratic
workings of the Empfängerformular see Chaniotis 1999, who stresses that requests of asylia are one of the
elements that led to a unification and standardization of the various cities’ formularies (he takes the
example of Magnesia, on which see now Ceccarelli forthcoming b, with bibliography): in such situations,
the same group of theoroi embarks on a tour, presenting to a number of cities, one after the other, the
same documents, and showing the answers received. Each time the city addressed would use as a basis the
last answer given, or at any rate model its answer on the text of the request, using it as a primer.
59
To give an idea of the proportion: in 201 bc, seventeen Cretan cities answered the request of
asylia; six answered with letters; a seventh, Aptera, may have used a letter (cf. discussion at appendix 3,
 Greek Cities and Epistolary Writing 319

What of the remaining instances? The decision of the Polyrrhenians, to inscribe

the covering letter that the polemarchs and synhedroi of the Thebans had sent
them, is not surprising; the Polyrrhenians, who had themselves sent a letter to
Teos in answer to the request of asylia, had no problems with this type of format,
and inscribed it along with the decree. The other instances of letters from Cretan
cities show that the use of the epistolary format in Crete (or at any rate in some of
the Cretan cities) is to be explained by more than simply the occasional use of
someone else’s formulation (Empfängerformular). Two documents are letter-
decrees by Cnossus and Gortyna for the same Coan doctor, Hermias son of
Emmenidas, who had distinguished himself for his services in a war in which
both cities participated on the same side; both texts were inscribed in Cos, in c.219
bc (nos. 5 and 6 = IG xii 4, 1 247 and 248). As these were honorific decrees, their
inscription would gratify the city that had given birth to the honorand, so the fact
that Cos monumentalized these letters is not particularly surprising; and as the
services for which Hermias was honoured were rendered to both cities on the
same occasion, we could suppose that one city followed the lead of the other.
A third decree for Hermias is preserved, by the city of Halicarnassus, dated to the
end of the third century and also inscribed in Cos: it praises Hermias for having
taken care of citizens of Halicarnassus in Cos, and is shaped like a formal decree
(both the decree of the Halicarnassians, and that of the Coans accepting the
honours voted by the Halicarnassians, were inscribed on the same stele).60 Thus,
it was not because of any influence of Hermias that Cretan cities produced letters
rather than decrees—otherwise, Halicarnassus too might have written a letter.
Of the remaining documents, one is a very fragmentary text from Lappa (no. 4),
sent to Tenos in answer to the Tenian request for asylia. The beginning is lost, but
there is an epistolary greeting at the end of the letter. The Tenians had grouped the
answers geographically, but the other four Cretan texts are all so badly preserved
that it is impossible to be sure of their format; it is thus impossible to know
whether Lappa’s reaction of sending a letter to the Tenians was unique, or shared
with other Cretan poleis.61 Then, there is a letter (the beginning is lost) of Cnossus
to Magnesia on the Maeander, accompanying an honorific decree for two citizens
of the latter (no. 8), dated to c.208 bc; the important item was here the decree, and
unfortunately its end is lost; but in the letter the Cnossians mentioned that they
had themselves inscribed the decree in the temple of Apollo Delphidios, and this
may well have been the reason that pushed the Magnetes to inscribe the letter as
well.62 Also unrelated to any request for asylia is the letter sent around c.200–170 bc
by the city of Axos in Crete to the Aetolians, and inscribed in Delphi, together with a
corresponding decree of the Aetolians (no. 16). The use of the letter format here can
be explained by the fact that Axos was not communicating to the Aetolians a

no. 26 = Rigsby 1996, no. 145; for the request of renewal, around 170 bc, Aptera will send a letter); ten
cities, Apollonia, Biannos, Cnossus, Istron, Lato, Lato pros Komara, Rhaukos, Allaria, Eleutherna, and
Lappa, sent decrees.
60
I. Cos, ED 132 = Samama 2003, no. 128 = IG xii 4 142.
61
See Rigsby 1996: 159–162, nos. 56–60; the text from Lappa is his no. 59 (= IC 2 xvi 2). Note also
the separately presented decree found in Phaestus, and referring to the same request, Rigsby 1996, no.
55, also extremely fragmentary.
62
Behind such decisions were, of course, also the interests of individual honorands.
320 Letter Writing and the Polis

decision: the family of a citizen of Axos had, after long vicissitudes detailed in the
stele, established itself in Aetolia, and the Axians are simply stating (probably
following a request of the interested party, Epikles) that Epikles and his family are
citizens of Axos, and thus should be covered by the existing agreements. It is up to
the Aetolians to take a decision, and this they do in the decree that follows on the
stone. Here one sees why both documents would have been needed, but one also
sees why Axos sent a letter, rather than a decree: there were no grounds for one. The
last two letters from Crete are part of a long and complex dossier, the arbitration
conducted by Magnesia between Itanos and Hierapytna (nos. 37 and 38); the letters,
sent by the kosmoi and polis of Gortyna and Hierapytna to the kosmoi and polis of
Itanos, were written around 145 bc and evidently kept in an archive; the dossier was
inscribed later, in 112/111 bc, and in this context, the letters were cited as proofs.63

7.2.2. Sparta

A second smaller, but still consistent cluster concerns Sparta. Against the rather
low number of official inscribed documents found in Sparta (and this is taking
into account both the decrees issued by Spartans and those issued by other poleis
and then inscribed in Sparta), the proportion of official letters appears all in all
rather high.64 Four, or possibly five, letters of Greek poleis addressed to Sparta
have survived. Of these, only one was almost certainly a decree in form of a letter
(no. 50 on the list); unfortunately, it is so fragmentary that it is impossible to
identify its origin, but what remains makes it clear that it was not just a matter of
conveying another piece of writing. The other letters seem to have been covering
letters, sent by the prytanis and demos of the Bylliones in Illyria, by the gramma-
teus of a Macedonian or Achaean polis, by an unidentified polis, and by either the
secretary of a polis or a king or a magistrate (respectively nos. 23, 44, 59, and 69 in
the list given in appendix 3). The frequent inscription of covering letters over a
relatively ample span of time (from the beginning of the second century to the end
of the first) cannot be brought back to the policy of one (or a very small) group of

63
See Ager 1996: 431–46, no. 158, for discussion of this case. References to letters appear in other
instances of arbitration: beside no. 39 on my list, a sealed letter by one Hippocrates of Oloosson is
mentioned (but the addressees are lost) in the dispute between Herakleion and Gonnoi, around 220–
200 bc: I. Gonnoi, no. 93, ll. 23–7. Hippocrates’ status and role in the affair are not known: the
traditional hypothesis saw in him one of the epistatai in charge of regulating the conflict; but according
to Hatzopoulos, he might have been a member of the royal administration, or else an arbiter designated
ad hoc (see Magnetto 1997: 305, who concludes that he was an epistates who would have collected,
written down in a letter, and sealed the testimony of those witnesses who could not be personally
present at the debate; see also Ager 1996: 147–50). The dossier comprised also a letter sent to Gonnoi
by Petraios, an official of Philip V, conveying to both cities the king’s decision.
64
Boring 1979: 6–7 states that out of twenty-seven decrees voted by other poleis and found in
Sparta nine are epistolary, dated to the period from the end of the third century bc to the end of the
second century bc; to these documents he adds the fragment of a treaty of the third century bc. As for
Spartan official documents, they comprise a fragment of a treaty with the Aetolians Erxadieis, dated to
the end of the fifth century; two lists of contributions for the Peloponnesian war; six honorary decrees,
of the third or second century bc; a financial document possibly to be dated to the first century bc; and
the regulations of the Leonidaea, first century bc. But there are no references, and it is not clear to me
how he counted his documents.
 Greek Cities and Epistolary Writing 321

individuals. To these documents, we may add three instances of official letters sent
by Spartan magistrates to other poleis (no. 28, 49, and 64 in the list). The first is a
letter of the grammatophylax (secretary) of Sparta to Amphissa, dating to the mid-
second century bc, accompanying a decree. The second is a letter sent by the
Spartan ephors and polis to the tagoi (chief magistrates) and the people of Larisa,
also to be dated to the second century bc, and found in Larisa. This is not a letter
of the grammateus, but a letter sent by the authorities: it may have functioned as a
covering letter for a decree appended below (presbeis are mentioned), but it might
also have been a self-standing document; regrettably, the text breaks off after four
lines. The third one is a later document, a letter of the ephors and polis of Sparta to
the Delphians accompanying a decree concerning honours for Diodorus, and part
of a larger dossier dated to the years 30–25 bc.
The relatively high number of letters connected with Sparta ties in well with
what can be gathered from the literary sources: in a world that still made slight use
of writing for long-distance communication, Herodotus and Thucydides had
reported a fairly large number of letters from Spartans; the same trend had been
apparent in Xenophon. We may set aside the problem of whether or not these
letters are authentic: what matters for our purposes is that for Herodotus, Thu-
cydides, and Xenophon exchanging letters was part of normal Spartan practice,
and that an unknown comic poet of the archaia could happily compare the size of
what must have been a very small field to that of a Laconian letter.65
The ŒıºÅ, the wooden stick carried by official Spartan messengers and used
to transmit orders or (later) secret messages, is also part of the picture: whatever
the details of its functioning, it is clear that it served to transmit written orders.66
Similarly, the importance of the (written?) transmission of messages is evident
from the title of K Ø  º, attested in Xenophon five times for the second-in-
command of the fleet (Xen. Hell. 1. 1. 23; 2. 1. 7; 4. 8. 11; 5. 1. 5; 5. 1. 6, and note
Plut. Lys. 7. 3). It is true that Pollux (1. 96) explains the title with › K d  f  º ı
Ø å   F Æıæå ı, ‘the successor of the navarch as head of the expedition’;
but the equivalence with K Ø  ºØÆç æ  in Xen. Hell. 6. 2. 25 makes it evident that
at least some Greeks heard K Ø  º in the title, and understood it as stressing the
close contact to be kept—via letters?—with the Spartan authorities.67

65
Herodotus, Thucydides, and Xenophon: see above, ch. 4. Comic comparison: see Comica
adespota, fr. fr. 456 K.–A.: ªæe  å’ Kºø ªB å ’ K Ø  ºB | ¸ÆŒøØŒB. Harris 1989: 112–
14 is sceptical about the level of literacy in Sparta; but see Cartledge 1978; Richer 1998: 489–90 and
519–20; Millender 2001 (on letter writing, esp. 141–2). Overview on uses of writing at Sparta, Thomas
1992: 136–7.
66
See Pritchett 1974: 45–6 (slightly outdated); Boring 1979: 39–41; Lewis 1996: 143–4 and n. 91; for
the problem posed by the extended (but anachronistic) description of the functioning of the skytale in
Plut. Lys. 19. 7–12 see Piccirilli 1997: 267 with further bibliography; detailed discussion of the skytale
and more generally of the modalities of communication in Sparta in Richer 1998: 482–90, following on
the main points in Kelly 1985. Skytale and letter are put on the same level in schol. Pind. Ol. 6. 190–1:
KåæH b ºÆÆØ ŒıºÆØ ƒ ¸Œø, Kªªæç  ÆıÆE a K Ø  º.
67
See Pritchett 1974: 45–6. Later, the secretaries of Hellenistic kings will bear the title of
K Ø  ºÆªæç  (the form present in Polybius): cf. the dedications from Delos honouring Menochares,
H æø çºø and K Ø  ºÆªæç  of Demetrius I of Syria (162–150 bc); honouring Bithys,
syngenes and K Ø  ºÆªæç  of Antiochus VIII Epiphanes Grypos (125–113 bc); and honouring
Stolos of Athens, officer of Ptolemy X Soter II (116–81 bc), syngenes of the king, IæåŁÆæ  ŒÆd
ÆÆæå  ŒÆd K Ø  º ªæç , ƪ  b ŒÆd æe ÆE ÆØ, I. Délos 1543, 1549, and 1534, with
322 Letter Writing and the Polis

Moreover, the later literary tradition offers further instances of letters sent by
Sparta. The epistolary dossier quoted in the first book of the Maccabees is a well-
known example. The first letter was supposedly sent by the high priest Jonathan to
the ‘Spartan brothers’; to it is joined an earlier letter, sent by King Areus (309–264)
to the high priest Onias; a third letter of the Spartiatai closes the dossier,
addressed, after Jonathan’s death, to his brother Simon; this one would have
been engraved on a bronze tablet.68 These letters are also mentioned by Josephus,
who gives a slightly different prescript, more complex and detailed, closer to the
standard epistolary one, while at the same time preserving the appellation ‘broth-
ers’: IæåØæf  øŁÅ  F Ł ı H  ı Æø ŒÆd  ªæ ı Æ ŒÆd e Œ Øe H
ƒæø ¸ÆŒ ÆØ ø Kç æ Ø ŒÆd ªæ ı Æ fi ŒÆd ø fi  E I ºç E åÆæØ,
‘Jonathan, high priest of the Jewish nation and the elders and the council of
priests to their brothers the ephors, the elders, and the people of the Lacedaemo-
nians, greetings’.69 And the collection of letters attributed to Apollonius of Tyana
comprises both a letter sent by the Spartans to Apollonius (a covering letter for the
decree it accompanies) and the latter’s answer.70 The fiction is evident; but here as
in the case of the correspondence between the high priests and the Spartan kings,
what is interesting is the choice of Sparta for this kind of fiction. Similarly, the
exchange between Aristocrates, the Arcadian king who betrayed Aristomenes, and
the Spartan king Anaxandros, at the time of the Second Messenian war, is clearly
an instance of ‘invention of tradition’ that may have its origin in the rewriting of
Messenian history in the years following the liberation of Messenia (from the mid-
fourth century onwards); here the letters, instruments of betrayal, are exchanged
between the two villains, the Spartan king and the Arcadian one.71
The Spartans’ laconic style seems to have become one of the models of
epistolary style: Plutarch remembers a number of letters by Spartans, some
concerning public and official communications, others exchanges that may be
better defined as ‘private’ or personal.72 This Spartan epistolary tradition, as

Virgilio 2011: 62–9. Note also the very fragmentary I. Délos 1535, and SEG 26, 1123 (Callisthenes
epistolagraphos of Alexander).
68
Respectively, 1 Macc. 12: 6–18 (letter of Jonathan): øÆŁÆ IæåØæf ŒÆd  ªæ ı Æ  F Ł ı
ŒÆd ƒ ƒæE ŒÆd › º Ø e B  H  ı Æø " ÆæØÆØ  E I ºç E åÆæØ. Ø æ æ 
I  ºÅ Æ K Ø  ºÆd æe ˇØÆ e IæåØæÆ Ææa æ ı  F Æ Øº   K E ‹Ø K b I ºç d
H, ‰ e IªæÆç   ŒØÆØ . . . ; 12: 19–23 (annex letter of Areus): @æÅ Æ Øºf " ÆæØÆH
ˇØÆ ƒæE ªºø fi åÆæØ. æŁÅ K ªæÆçBfi æ  H " ÆæØÆH ŒÆd  ı Æø ‹Ø N d I ºç d . . . ;
and 14: 16–23 (letter of the Spartans to Simon): " ÆæØÆH ¼æå  ŒÆd  ºØ "øØ ƒæE ªºø fi
ŒÆd  E æ ıæ Ø ŒÆd  E ƒæF Ø ŒÆd fiH º Ø fiH ø fi H  ı Æø I ºç E åÆæØ . . .
69
Fl. Jos. Ant. 12. 166–70; and Ant. 12. 225–8 (letter of Areus to Onias, beginning: BÆ Øºf
¸ÆŒ ÆØ ø @æØ   OÆ fi åÆæØ. Kıå  ªæÆçB
fi ØØ oæ  ‰ K Ke r  ª ı  ı ÆE Ø ŒÆd
¸ÆŒ ÆØ Ø Ø).
70
Apoll. Tyan., Epp. 62. 1–2: ¸ÆŒ ÆØ Ø Ø  ººøø fi . A  A ØA Ø I  ºŒÆ
  IªæÆç  ÆÆ Ø fi A Æ Æ fi çæƪE Ø, ¥ Æ Y fi Å. 'çØ Æ ¸ÆŒ ÆØ ø, ŒÆŁg ƒ
ªæ  . . . (‘The Lacedaemonians to Apollonius: We have sent this copy of the honours we are giving
you, sealing it with the public seal, for your information’), and 63:  ººØ  Kç æ Ø ŒÆd
¸ÆŒ ÆØ  Ø . . . (‘Apollonius to the ephors and the Lacedaemonians . . . ’); no closing greetings
(I follow here the Loeb text and translation of Jones 2006).
71
According to Pausanias (4. 22. 5–6), the letter was carried by a slave, who was intercepted as he
was coming back. See Musti and Torelli 1991: 234.
72
Discussion in Celentano 1991.
 Greek Cities and Epistolary Writing 323

preserved in Plutarch, is construed as going back to Lycurgus: two of his apoph-

thegms were reputedly handed down in writing:
There are also handed down similar answers which he made by letter to his fellow-
citizens (çæ ÆØ b ÆP F ŒÆd Ø’ K Ø  ºH I Œæ Ø  ØÆFÆØ æe  f ºÆ).
When they asked how they could ward off an invasion of enemies, he answered: ‘By
remaining poor, and by not desiring to be greater than the other.’ And when they
asked about fortifying their city, he answered: ‘A city will be well fortified which is
surrounded by brave men and not by bricks.’ (Plut. Lyc. 19. 4, trans. Perrin)
At least the second one is a topos that had been circulating since the time of
Alcaeus; and as Plutarch himself wisely remarks, æd b s  ø ŒÆd H
 Ø ø K Ø  ºH h I Ø B ÆØ Þfi  Ø  h Ø F ÆØ, ‘regarding these and
similar letters, neither scepticism nor belief are easy’ (note that in Apophth. Lac.
228 d, nos. 27 and 28, Lycurgus gives the same answers orally). A number of other
Spartan kings and generals have among the stories attached to their lives some
anecdote connected to writing. The famous story of Demaratus’ tablet and the
messages that according to Thucydides were circulating at the time of the Ionian
war have been discussed above; but an epistolary exchange is mentioned also in
connection with Brasidas: the text ( ¯ ºŁg ’ K d º  ªæÆł  E Kç æ Ø
‘– Æ º ÆØ æø ŒÆa º  j ŁÆ FÆØ’, ‘When going to war he wrote
to the ephors: “What I want to do I’ll do, concerning the war, or I’ll die” ’, Plut.
Apophth. Lac. 219 d, no. 3) is extraordinarily banal, and this makes one suppose
that the dictum was made up because it was felt that Brasidas too should have
some kind of anecdote, and if possible in writing, attached to his name.
Lysander is also at the centre of quite a few exchanges: Plutarch reports the
common tradition concerning his message to the ephors at the moment of Athens’
surrender, together with the reaction of the ephors (famously, Lysander wrote
‘Athens is taken’, to which the ephors objected that ‘taken’ would have been enough;
if Plutarch is to be trusted, the Lacedaemonians themselves were behind the diffusion
of this sort of anecdote).73 The theme of the deceptive letter that brings death, which
had played such an important role in the story of the regent Pausanias, is given a new
twist in the story of the dissent opposing Lysander and Pharnabazus: as a result of
Lysander’s excesses, Pharnabazus sent men to Sparta, denouncing him, and when
the ephors found that men of Lysander’s entourage had money in their possession,
they sent a skytale to recall Lysander to Sparta (Plut. Lys. 19.4);
But Lysander, when the skytale reached him at the Hellespont, was much disturbed,
and since he feared the denunciations of Pharnabazus above all others, he hastened to
hold a conference with him, hoping to compose their quarrel. At this conference he
begged Pharnabazus to write another letter about him to the magistrates (K E
ªæłÆØ æd ÆP F æe  f ¼æå Æ æÆ K Ø  º), stating that he had not
been wronged at all, and had no complaints to make. But in thus ‘playing the Cretan

73
Plut. Lys. 14. 4: ŒÆ Ø ¸ÆŒ ÆØ ø K d IŒ F ÆØ ºª ø ‰ ¸ Æ æ  b ªæÆł  E
Kç æ Ø  : ‘±ºŒÆØ Æd ŁAÆØ,’ ¸ı  æø fi ’ IªæÆłÆ ƒ ç æ Ø: ‘IæŒE  ª ƺŒØ.’ Iºº’
P æ Æ åæØ y  › º ª   ºÆ ÆØ (and cf. Apophth. Lac. 229 b, 5). In the same context
Plutarch also quotes the actual decree ( ªÆ) sent back by the ephors, adding that the Athenians
accepted this ŒıºÅ, in a neat equation of message and support: cf. Plut. Lys. 14. 5; for the message of
the ephors cf. Andoc. 3. 11–12 and 31, Lysias 13. 14, Xen. Hell. 2. 2. 20.
324 Letter Writing and the Polis
against a Cretan,’ as the saying is, he misjudged him. For Pharnabazus, after promising
to do all that he desired, openly wrote such a letter as Lysander demanded, but secretly
kept another by him ready written (çÆæH b ªæÆł ¥ Æ › ¸ Æ æ  Mø 
K Ø  º, ŒæçÆ b r å æÆ ÆP ŁØ ªªæÆÅ). And when it came to putting
on the seals, he exchanged the documents, which looked exactly alike, and gave him
the letter which had been secretly written (K b fiH a çæƪE Æ K غºØ
KƺºÆ a ØºÆ Å b ØÆçæ Æ B fi ZłØ,  ø Ø KŒÅ ÆPfiH c ŒæçÆ
ªªæÆÅ). Accordingly, when Lysander arrived at Sparta and went into the
meeting place of the ephors, as the custom is, he gave them the letter (a ªæÆÆ)
of Pharnabazus, convinced that the greatest of the complaints against him was thus
removed; for Pharnabazus was in high favour with the Lacedaemonians, because he
had been, of all the King’s generals, most ready to help them in the war. But when the
ephors, after reading the letter, showed it to him (K d b Iƪ  ƒ ç æ Ø c
K Ø  ºc  ØÆ ÆPfiH), and he understood that ‘Odysseus, then, is not the only man
of guile,’ for the time being he was mightily confounded and went away. (Plut. Lys. 20.
1–4, trans. Perrin slightly modified)
This did not make Lysander lose his faith in epistolary communication: we find
him slightly later (Plut. Lys. 23) sending letters to his friends in Asia, asking them
to request from Agesilaus a general for the war in Asia—the ploy initially worked,
and Agesilaus took Lysander with himself on campaign. Later, Lysander, already
out of favour, is presented as committing to memory a speech prepared for him by
Cleon of Halicarnassus (º ª  KºÆ æe c  Ł Ø ªªæÆ ). This
was meant to convince the Spartans that they should elect as king the best of them,
regardless of the family. Linked to this endeavour are his beginning a collection of
oracles and his attempts at bribing the Pythia, the priestess of Dodona, and the
oracle of Ammon in Libya. All this culminated in an extraordinary stratagem in
which a pretended son of Apollo should have reclaimed some very ancient tablets
held in Delphi, which contained oracular responses and prophecies stating that ‘it
was more for the honour and interest of the Spartans to choose their kings from
the best citizens’.74 But this endeavour too failed, and the story became public after
Lysander’s death. The last epistolary theme appearing in connection with Lysan-
der is the topical one of the intercepted letter: it also signifies his death. Plutarch
narrates that before the battle of Haliartus (395 bc), Lysander sent a letter to
Pausanias ( ł b fiH —Æı ÆÆ fi ªæÆÆ), asking him to leave Plataea and
join forces with him at Haliartus. The letter (ÆFÆ a ªæÆÆ) was, however,
intercepted and read by the Thebans, who managed to reach Haliartus before
Lysander; when the latter attacked alone (Pausanias was marching from Plataea to
Thespiae), he was killed, and the army routed (Plut. Lys. 28. 2–5).
A familiarity with letter writing is also attributed by Plutarch to Agesilaus and
his son Archidamus. A longish letter (clearly fictive) is quoted in its entirety, with
which Agesilaus in writing back to the ephors showed his natural superiority;75

74
Plut. Lys. 25–6; see especially, for the ancient oracles on tablets, to be handed over only to a son of
Apollo who could prove his descent, 26. 2: ‰ K ªæÆ Ø I ææ Ø  e H ƒæø çıº Ø
Æ ºÆØ Ø  Ø åæÅ  , ŒÆd ºÆE PŒ  Ø   ı P  KıåE ŁØ , N  Ø ¼æÆ ªª g
K  ººø  IçŒ Ø fiH ººfiH åæ ø fi ŒÆd ŁÅÆ  E çıº ı Ø B ª ø ªæØ 
ÆæÆ åg Œ  ÆØ a º ı K Æx  q Æ ƒ åæÅ  . The story was already in Ephorus.
75
Apophth. Lac. 211 bc: ˚Æd IªæÆł  E Kç æ Ø K Ø  ºc  · “ªÅ ºÆ   E Kç æ Ø
åÆæØ· a ººa A  Æ ŒÆ æłŁÆ ŒÆd g Æææø Kº Æ ŒÆd K fi A  øÆ fi ‹ ºÆ
K Ø Æ ºº· K d b Œº Ł  ŒÆa a æ Ł Æ Ææƪ ŁÆØ,  ÆØ fi A K Ø  ºfi A, å e ’
 Greek Cities and Epistolary Writing 325

another anecdote concerns his refusal to write letters of recommendation on the

grounds that his friends would naturally act justly, even without a recommenda-
tion: this betrays a certain contempt towards the genre (Apophth. Lac. 212 e).
Especially informative for the question of the traditional character of these
anecdotes is Ages. 23. 6 = Apophth. Lac. 213 de (no. 69), retelling a story already
narrated in Xenophon’s Agesilaus (8. 3, discussed above, 154–5), concerning the
Spartan king’s refusal to accept letters from the Great King.
As for Archidamus, who had received a harsh letter from Philip II after
Chaeronea, he reportedly wrote back that ‘were he to measure his shadow now,
he would not find it any longer than before his victory’ (Apophth. Lac. 218 ef, but
elsewhere this is a riposte made orally); a remarkably laconic letter of his to the
Eleans is also known (Apophth. Lac. 219 a). Another very typical Spartan answer
to a letter of Philip II is preserved in Apophth. Lac. 235 ab (cf. 513 a): ‘Concerning
what you wrote to us, no’ ( æd z ¼Ø ªæÆłÆ, h).
The Apophthegmata further preserve the letter of an Hippocratidas to a Carian
governor (222 ab), while in 225 cd Plutarch transcribes an epistolary exchange
between Leonidas and Xerxes(!) Finally, a few letters are attributed to Spartan
mothers.76 Among these, particularly interesting is the letter quoted in the Sayings
of Spartan Women and attributed to Pedaritus’ mother Teleutia: after having
listened to the complaints of some Chian exiles in Sparta, she wrote severely to
her son, saying:
± Åæ —ÆØ Ææø fi · j º Æ æA  j ÆsŁØ , I ª f a K " æÆ
øÅæÆ. ([Plut.] Apophth. Lacaen. 241 e, 11)
Your mother to Pedaritus: either behave better, or remain where you are, and don’t
expect a safe return to Sparta.
In his eighth book, Thucydides had mentioned a letter of Pedaritus to Astyochus
(Thuc. 8. 33. 3), as well as another letter accusing Astyochus and sent to Sparta
(Thuc. 8. 38. 4). Pedaritus’ words must have carried weight in Sparta if his message
could provoke such a quick reaction as the one described in Thucydides;77 at
any rate, the letter of the Apophthegmata is predicated on the assumption of a
powerful and literate family.
These letters are clearly part of a tradition that grew in time; the initial corpus
was enlarged through the addition of variants. Plutarch is the main collector, but
versions are preserved in other authors too. The exchange between Menecrates
and Agesilaus for instance, narrated three times in the same way, with the same
terms and the same protagonists in Plutarch (in Ages. 21. 5, in Apophth. Regum et
Imperatorum 191 a no. 5, and in Apophth. Lac. 213 a), is adjusted in Athenaeus
(7. 289 de) and Aelianus (VH 12. 51) to an exchange between Menecrates and
Philip II of Macedon. While these mainly Plutarchan stories of Spartan letters are
evident fictions, and while their number is minimal when compared to the

ÆPa ŒÆd çŁÆ H· ¼æåø ªaæ PŒ KÆıfiH Iæå, Iººa fi A ºØ ŒÆd  E ıå Ø· ŒÆd   ¼æåø ¼æåØ
IºÆŁø ŒÆa ŒÆ, ‹Æ ŒÆd ¼æåÅÆØ    ø ŒÆd Kç æø j x Ø i ¼ºº Ø K ºØ ¼æå  t Ø.”
76
e.g. Apophth. Lacaen. 241 e 10; note Gorgo’s ability to understand Demaratus’ stratagem, above,
113.
77
Thuc. 8. 39. 2, cf. Dover 1981: 69; Hornblower 2008: 844, 861–3; and above, 146–7.
326 Letter Writing and the Polis

number of sayings of these same Spartans, these stories do support the general
hypothesis of a relatively frequent use of epistolary writing for official—and
unofficial—communication in Sparta.

7.2.3. Athens

After having looked at some letter-writing poleis on the list, it is important to

examine a relative absence: Athens—all the more striking, since it is the polis that
always comes to the fore when discussing issues of literacy. In the list given in
appendix 3, Athens appears eight times, in nos. 27, 36, 42, 45, 46, 54, and 62–3.78
The first document is a letter written by the Athenian strategoi to the epimeletes of
Delos Charmides and accompanying a Roman senatus consultum; it was inscribed
in Delos and is dated post 164 bc. The second document is a letter dated to shortly
after the mid-second century bc, sent by the archons and the city of Delphi to the
Athenian boule and demos, acknowledging receipt of a psephisma and honouring
the Athenian demos; we have this letter (which moves into a decree halfway
through) because it was inscribed in Delphi. The other documents have all been
inscribed in Athens—but none of them was written by Athens as a polis. Six letters
publicly inscribed in Athens, out of a list numbering all in all seventy-four official
letters, might not seem so few. But a closer look shows that the second to fourth
documents are letters from the Amphictions concerning the Dionysiac artists:
these are not documents concerning the polis, but a community within the polis,
and the initiative for their inscription will have rested with the Dionysiac artists.
The fifth letter comes from Delphi as well, and concerns honours for a priestess
(but ll. 44 and 52 are heavily restored—the text is almost a matter of guesswork).
And the final two documents are a letter sent by the genos of the Gephyraioi to
Delphi in order to renew the old friendship and to consult the oracle, and the
answer of the Delphians accompanying the text of the oracle received. Here too,
the Gephyraioi will have taken care of the inscription. In short, all six letters found
in Athens have to do with Delphi—none of them originates from Athens itself,
and they tend to concern specific ‘communities within the community’; neither in
Athens nor elsewhere is it possible to find any inscribed letters written by the
Athenian boule and demos (although note two instances in which the sender is
unknown but has been suspected to be Athens: nos. 55 and 56). As for the Delian
letter, it is sent by the Athenian strategoi (note the heading ƒ æÆŪ d &Ææ Ø
K غÅ- | E ˜º ı åÆæØ) to accompany a decision taken by the Roman senate:
it thus fits into the pattern of communication internal to a polis, and its inscrip-
tion was most certainly due to the initiative of the priest, Demetrius of Rheneia,
who benefited from the decision of the Senate and, consequently, of the Athen-
ians, and who had an interest in having this put on stone.79 One further point
worth noting is that the series begins relatively late—in 164 bc at the earliest

78
One may want to keep in mind also the letter written by the Delphians to the Dionysiac technitai
of Athens, no. 61; the letter was inscribed in Delphi, but on the wall of the Treasury of the Athenians.
79
See on this document Ferrary 2009: 126–7; particularly interesting is Ferrary’s interpretation of
the lack of a Roman letter, which he sees as consistent with the concern of the Senate not to appear to be
giving orders to the Athenians.
 Greek Cities and Epistolary Writing 327

(in Delos) and in 134 bc at Athens. Finally, eight letters are simply as nothing
when compared to the thousands of extant epigraphical documents from Athens.

7 . 3 . I N T E R P R E T A TI ON

How to interpret the data collected in the list given in appendix 3? For most cities,
the data are so scanty that it is better to refrain from proposing interpretations; for
others, variations can be traced, that show that the same format was not system-
atically employed.80 Yet some particular clusters seem too pronounced to be
explained away as accidents or idiosyncrasies. Those poleis that made official (if
only occasional) use of epistolary communication or were happy to display a letter
from another polis in the midst of their own city share (I submit) a common trait:
a weakly developed sense of the distinction between public and private.81 Sparta
was notoriously not a democracy, it was ruled by a tight-knit oligarchy; the
interference of the polis in all spheres of life was very strong. Similarly, in their
analysis of the decrees of the Greek poleis Rhodes and Lewis remark apropos of
Crete that a number of features imply that in the Hellenistic period the Cretan
poleis were not among the most democratic of the Greek world.82 If we are to see
‘letters’ rather than generic ‘writings of all sorts’ in the ªæÆÆ mentioned in the
constitutional regulation from Euromos concerning the introduction of two new
magistracies (SEG 43, 707, ll. 12–15: ªæÆÆ b ¼ ı ÅØ   ŁÆØ  bæ B
ºø j  bæ ¼ºº ı | Øe Øa H Iæåø  ø KÆ ºº Łø ªæÆç Æ
K- | [  ] Hª Œ ø ŒÆd H æ ÆH, c _ K ı Æ b  ø Å b ›- | 15
_ _
[ æøØ] H Iæåø  øª ŒÆŁ’ N ƪ ªæÆÆ  Ø, ‘and if letters must
be sent by _these magistrates concerning the city or any other matter, let there be
sent a letter written in the presence of both the kosmoi and the prostatai, and let it
not be allowed for [either] of these magistrates to send a letter on his own’), then it
is interesting to note the presence at Euromos of the Cretan institution of the
Œ  Ø, which may have resulted from an influx of Cretans into the citizen body;
the clause on letters (if letters they are) might have to do with Cretan habit.83
Another place where letter writing on a relatively important scale seems to have

80
The importance of letter writing for each polis could possibly be evaluated by comparing, case by
case, the number of extant decrees for internal matters, the number of extant decrees on external
matters, and the number of letters. But very high numbers would be needed to counter the accidents of
chance survival.
81
The date at which these letters begin to appear fits the change visible in the composition of the
council at the ‘basse époque hellénistique’: see Hamon 2005.
82
Rhodes with Lewis 1997: 312: ‘Although Crete has produced some of the earliest public docu-
ments in the Greek world, such considerations as the prominence of the kosmoi, the failure in most
cities to identify the authors of decrees, and the failure in decrees which provide for an appointment
after the enactment of the decree to stipulate that the appointment is to be made from all citizens,
suggest that in the Hellenistic period the many small cities of Crete were not among the most
democratic in Greece.’
83
Decree from Euromos: cf. Ma 1999: 339–40 (no. 30), whose translation I quote; the ample
discussion of Savalli-Lestrade 2010: 138–42 (who insists on ‘letters’, and develops the possibility of a
massive insertion of Cretans into the civic body when Euromos was refounded as Philippi) and 142–8;
Virgilio 2011: 39. Boffo 2003: 50 translates simply ‘scritti’.
328 Letter Writing and the Polis

been the norm is Sicyon (see nos. 33–5, two letters of Sicyon to Argos, and one of
Kerynia to Sicyon, all dated to after 146 bc): remarkably, in this period there is no
record of an assembly in Sicyon.84 If we accept the thesis that the distinction
between public and private is more marked in democratic systems, while it is often
blurred in oligarchies, and non-existent, at least in theory, for kings, then we could
posit a correlation between the use of letters (or rather: a greater facility in the use
of letters) and the latter types of regime.
Before drawing conclusions, an issue should be addressed, namely that of the
chronology. The inscription on stone of letters written by poleis seems to be a
relatively late feature: the oldest documents so far discovered are the two letters of
Istron and Phaistos to Cos, for the asylia of 242 bc; next is a covering letter of the
inhabitants of Thessalonike to the Delians, dated to c.230 bc, accompanying a
decree of the same Thessalonikeis (it is fitting that one of the earliest epigraphic-
ally attested letter from a polis should come from one founded c.315 bc by the
Macedonian king Cassander on the site of Thermae, in an area that was under
Macedonian control!) But if we accept that Ptolemy, when writing to Telmessos
that he had received their letter (K Ø  º), indeed meant a letter, then we must
accept that at least by 279 bc a polis could write an official letter in the context of
interstate relations.85 This is actually not surprising. The letter sent by Adeiman-
tos of Lampsakos to Demetrios shows that not just Adeimantos himself, but all the
king’s ‘friends’ (who will have been members of the city elites coming from all
over the Greek world) had written letters on the occasion of the re-establishment
of the Corinthian league, and had had them inscribed:
 Æ  vacat Æ ØºE ˜ÅÅæøØ åÆæ[Ø]·   łçØ Æ n K  Å ƒ
çØŒ   [æı]- | Ø æ Ł b K  Ł Ø, K ØŒıæ Æ[] b K ˜ºç E,
I [ ]ƺŒÆ ŒÆŁ _ æ þØ_ ı E· [› ]- | [ø] b ŒÆd a Ææa H çºø K Ø  º,
¥ Æ ÆæÆŒ º ıŁB _
fi  Œ  Ø [  ª] , IƪæłÆÆ N [ºÅ] | [ŒÆ]a e
ª   ªÆ IÆŁEÆØ n Œ[Æd K ] ØŒØ K HØ [ı æøØ _ Œ]Æd a ªª  Æ _
Å[ - c.4 - ] | [çغŁæø Æ ]Ææa H çØŒØ [ø]_ _[ - c.7 - ]ˇ[ - - - c.25 - - -
]ˇ˝Bˇ$- ¸O[ - c.8 - ] (SEG 45, 479 = CID 4 no. 11, _ll. 1–5)
_
‘Adeimantos to King Demetrios, greetings! I have sent you, as you wished, the decree
which the Amphictions made last year, having proposed it at the Isthmia and ratified
at Delphi; and equally also the letters from the friends (philoi), so that you may know
that to each it has seemed good, having inscribed (the letters) on a stele, to dedicate it,
according to the decision (dogma) which we took in the Synedrion . . . and the honours
voted by the Amphictions . . .
This kind of collective letter writing (and letter inscribing) by members of the
elite must have made a huge impression: if the decision had been taken at the
Synedrion to inscribe all these letters—one would like to know where: in Delphi,
as a dossier? in the city of origin of each friend?—then it had a value that went far
beyond the individual approval of the various philoi.86

84
Rhodes with Lewis 1997: 76. Argos too wrote letters: cf. I. Magnesia 40, app. 3 no. 7.
85
Above, nn. 45 and 46.
86
The text we have was inscribed in the early second century bc, with other documents, on a huge
pillar (at least 3 m high) probably bearing a statue of Perseus; Laroche, CID 4, no. 11 argues
persuasively that Perseus republished old documents, in order to emphasize, in the years preceding
the third Macedonian war, the good relations existing from the beginning between the Antigonids, the
 Greek Cities and Epistolary Writing 329

In discussing the origins of the format assumed by the Hellenistic royal letter
Welles argued that such letters derived from city-decrees, which the chancery
personnel took as their models. Or, to put it more precisely, he argued that ‘the
official letter to an individual was originally in form a private letter; the official
letter to a community was based on the prevailing form of communication
between communities, the city decree.’87 Welles did not discuss the letters sent
by poleis, but the logical step from his point of view would have been to assume
that the poleis in their turn imitated the habit of the Hellenistic kings.
In this strong version, there are difficulties with the argument. First, the
distinction between official letters sent to individuals and official letters sent to
communities seems to complicate matters unnecessarily, since there is no real
difference in the way both are couched, either in the source of the letter’s authority
or in how it will work.88 Secondly, the important point is surely the hypothesis of a
derivation of the royal letter from the city-decree. We have seen above a number
of experiments by kings and other communities with the form of the city-decree,
adapted with greater or lesser success and skill to the needs of the moment. Are
Hellenistic royal letters the result of a similar development, or could the Hellenis-
tic chanceries rely on a stable epistolary format? Although the Greeks sent letters
at the latest from the mid-sixth century onwards, a stable format for epistolary
correspondence is traceable in our evidence only from the mid-fourth century
onwards. The first normative statements concerning the nature and the form of a
letter are in Isocrates. This means that there was no standard model earlier; at the
same time, we know, because our texts mention them, that letters were already
being written, and in reasonably high numbers—numbers which will only in-
crease in the course of the fourth century. This is when the various royal
chanceries develop their specific models.
In general, it is reasonable to assume that a royal answer to a decree will have
been couched in terms that followed quite closely those of the decree which
occasioned it. The model of the Empfängerformular will have worked also in
this case, with royal chanceries looking at the texts of decrees and modelling their
answers on them (and poleis looking at royal letters and deciding how to deal with
them). This explains the existence of a common, shared language of euergetism.
However, we have also noted above the existence of important structural differ-
ences between royal letters and decrees, as well as of specific choices in the use of
some key words ( Œø, K Ø ). I suggest that on the level of the semantic
choices, these differences can be seen as instances of inverse modelling and
distantiation, while at the structural level they reflect the continuous thread that
runs from private letters to royal letters. Although it is correct to say that the
Hellenistic chanceries will have adapted forms and elements of the city-decrees for
the varied necessities of the royal correspondence, at a more fundamental level

Amphictions, and the Greek world at large. On Adeimantos and his role in liaising between king and
cities see Wallace 2013: 149–51, with a slightly different interpretation of the letters, which he takes to
be dogmata of the league.
87
Welles 1934: pp. xlii f. He partly contradicts himself, since elsewhere he states that ‘the letters
inherited from both their parents, though less from the city decree than from the private letter, a taste
for brevity and simplicity’ (p. xlvi), thus acknowledging the strong influence of the private letter.
88
Moreover, at least in Macedon, royal letters concerning a community would be addressed to the
epistates (see Mari 2006).
330 Letter Writing and the Polis

they were continuing a practice of letter writing for personal purposes that was
much more ancient, and one that through Macedon could possibly be traced back
to Achaemenid practices; the city-decree marked something completely new in
terms of communicative strategies.
What is the place of the letters of poleis in this picture—should we see them as
the consequence of the influence of royal letters on the political life of the
Hellenistic world? Because of the high number of unknown variables occasioned
by the accidents of transmission and the scarcity of the epigraphical material, it
would be dangerous to insist on excessively clear-cut distinctions. Yet it remains
the case that there is a strong opposition between the language of the kings (royal
letters) and that of the polis (decrees). At the same time, some poleis made use of
letters for official interstate communication, and most certainly did so for official
internal communication. This means that the epistolary language cannot be
considered a format totally external to the language of the polis—all the more
so since those cities which used letters to transmit decisions did not see any
problem in combining the structure of the letter with the vocabulary of the
decree. At least in the case of Sparta, the use of the epistolary format for
communications seems to be traditional. It is attested already at the beginning
of the fifth century bc, and cannot therefore be interpreted as an imitation of
the habit of Hellenistic monarchs of sending letters. The other place where we
find a high number of official letters is Crete: already Aristotle had remarked
on the similarity between the two constitutional systems of Crete and Sparta.89
It must be admitted that in the case of the Cretan poleis it is difficult to speak
with certainty of a ‘traditional’ use of the epistolary format: the first preserved
letters, both private and public ones, are to be dated to around 200 bc; the
literary tradition does not provide us with any further information. Moreover,
the existing decrees show modifications in the formulary in the Hellenistic
period. But for Sparta at any rate, and very probably for some of the other
poleis for which the sending and/or inscribing of official letters is attested, the
use of letters for official communication (and the relative ease with which a
monumental status was conferred on them) has to be explained by the use in a
non-democratic regime of formats typical of private communication also for
the sphere which we should rather define as public.90 If this is correct, then the
royal chanceries had their models for official epistolary writing already within
the world of the Greek poleis, models that they then developed further. It is a
sign of the strong democratic feeling in other Greek poleis that this format was
for a long time eschewed (or at least not monumentalized), even in those
situations where it might have seemed appropriate, such as for correspondence
with other poleis.

89
Arist. Pol. 1270b28–31 criticizes the tendency of the ephors to judge on the basis of their opinions
rather than according to written laws; a similar criticism is made also of the Cretan kosmoi (Pol.
1272a35–9). See Millender 2001: 132–3 and nn. 52–3.
90
Interestingly, in the examination of the types of writing most used in archaic and early classical
Greece made by Stoddart and Whitley 1988 and Whitley 1999, Crete had appeared as a society in which
a restricted elite made a public use of writing, as opposed to Attica, in which personal writing by
craftsmen and aristocrats competing for prestige would have been widespread. See above, 30–1.
 Epilogue

I set out to investigate the use of writing for long-distance communication in the
ancient Greek world. In particular I wanted to explore both the development of
letter writing until it became the instrument of royal communication and the
connotations acquired by official correspondence, as opposed to public speech or
other forms of writing. The preceding chapters show, I believe, that there were
definite differences in when letter writing began to be used, and in how letter
writing (and writing tout court) was viewed within different parts of the Greek
world; but also, that things changed, and that the important watershed in Greece,
so far as letter writing is concerned, is the period at the end of the fifth and the
beginning of the fourth century. The evidence shows that writing was used in
Greece to send messages over a distance already in the mid-sixth century bc, from
the Black Sea to Spain (ch. 2); and given the unpredictability of epigraphic finds
and the likely use of perishable supports, we may wish to posit the same practice
for other areas of the Mediterranean. However, these early letters are couched in a
very individual language, reflecting closely the spoken language; they almost
entirely lack the formulae that will characterize later Greek letter writing. Such a
lack of standardized formulae is evident also in the letters transmitted through the
literary tradition in Herodotus and Thucydides; clearly, letter writing was not yet a
widespread activity. It is only in the last quarter of the fifth century that letters
enter the ‘literary’ scene (chs. 4, 5). In Herodotus’ work, they are (mainly) the
prerogative of tyrants and oriental kings; in the Greek world, their use appears
linked to a desire for secrecy, usually unhealthy—or to the desire of avoiding face-
to-face confrontation, which is again something deemed to be rather unhealthy.
Of course, the Herodotean letters (and those that appear in Thucydides’ and
Xenophon’s work) are not the everyday private messages that individuals would
send to each other: they play a role in interstate relations. A first crucial point
emerges here: in itself, a letter is but the transcription of an oral message (and this
remains valid, even from the formal viewpoint, until the mid-fourth century), and
as such its use is not viewed in any way differently from that of an oral message
(chs. 3, 4). A letter and an oral message may both be truthful or deceitful.
However, for various reasons, the medium acquires negative connotations when
use of it is made for transactions concerning not the private but the public sphere,
and more specifically interstate relations. That this should be observable especially
in Athens is not surprising, nor is it just a problem of sources: Athens is the polis
which, already in Pericles’ day, the third quarter of the fifth century, had advanced
most towards differentiation between a private and a public sphere. In areas of the
Greek world where this differentiation was less marked, such as Sparta or Crete,
332 Epilogue

the letter may have been felt, and probably was felt, to be an unproblematic
instrument for managing public policies (ch. 7). Along the same lines, it is
difficult—not to say impossible—to ascertain what exactly lies behind Lucian’s
story concerning Cleon’s use of a written format, preceded by the apostrophe
åÆæØ, to liaise with the Athenians, but passages from comedy and tragedy testify
to an unease, an underlying worry, about epistolary communication. This concern
is particularly evident in Euripidean tragedy: interestingly, Euripides’ questioning
does not involve writing as such, but is specifically focused on communication by
letter. Finally, one more general reason for the often negative Greek perception of
letter writing may lie in the use made of it by the Persian king for controlling his
empire. However, it seems more likely that it was the socio-political conditions
(the fragmentation of the Greek world and the lack of a unified postal system,
implying that a messenger was anyway obliged to cover the entire distance) that
almost by necessity rendered letter writing in official contexts something often
unnecessary, and always open to the suspicion of involving some unsavoury
activities (here Bellerophontes’ letter functioned as a model, on which variations
would be played).
The very nature of the letter would have implied, in the Greek world, this kind
of connotation: a letter is a personal message, strongly marked by the presence of
an ‘I’ addressing a ‘You’, and sealed, in other words, excluding others from the
exchange. In a private context this is by and large unproblematic, unless these
private dealings end up in a law-court; for in this context the choice of a private,
sealed means of communicating may be given a negative slant (ch. 6). But when
used at the public level, for the polis’s affairs, sealed, personal communication may
be problematic. Here, however, geographical differences acquire importance.
Most of our evidence is Athenian, and so can be used only with caution for
drawing inferences about the rest of the Greek world; but it would seem that some
communities (Sparta, for instance) had fewer difficulties with written communi-
cation than others (such as Athens). There is nothing like the traditions surround-
ing the Spartan skytale for Athens. Of course Athenian generals did send reports,
which would be read by the council and in the Assembly, and then kept in the
Metroon: this shows that in Athens, the letter (the long-distance message) had to
be open and unsealed for it to be accepted at the official level. It may not be an
accident that the format of the ‘open letter’, as used, for instance, by Isocrates,
developed in Athens, and exactly in the first half of the fourth century (ch. 6).
In the preceding paragraph I have focused on the uses of letter writing for
official communications internal to a polis; at that level, the ‘secrecy’ was the
aspect that might have been negatively focused upon. When used for exchanges at
the interstate level, the other feature of the letter, its being addressed by an ‘I’ to a
‘You’, would have brought further uneasiness. Comparison with the city decree
brings this into sharper focus: a decree is expressed in the third person; the polis
speaks, in the past, and there is no addressee. The decree is a type of format that
highlights (and suits perfectly) the strong ideological tendency to autarky typical
of all Greek poleis. A letter, however, must have an addressee; moreover, a feature
of epistolary correspondence is the fact that the acceptance of a letter implies
entering into an exchange, into a discourse initiated by the other. This goes a long
way towards explaining the strange conversations between cities and kings in the
Hellenistic period, with kings writing letters, and cities sending back decrees.
 Epilogue 333

In the fourth century, an epistolary language appropriate to, and standard for,
private communication takes shape; in this same period, the chanceries of the
Macedonian kings, and then of their successors, adapt this private form of
communication to interact with the Greek poleis. Just as they had accepted
entering into discussion with the Persian king, the Greek cities entered into
discussion with the Hellenistic kings; and this may in a certain measure explain
why from the mid-third century bc on we find some cities writing letters for
official communication that get inscribed on stone. The work of Polybius reflects
this evolution: while he (still) tends to associate epistolography mainly with
Hellenistic kings and the Romans, thus continuing the tradition of the letter as
ideologically marked, his history also recognizes the writing of letters as a standard
medium of diplomacy throughout the Mediterranean world. From early on, the
letter had been a widespread means of personal communication; but by now, it
has also become an accepted instrument of interstate communication, even if
some areas appear to have resisted longer than others.
 APPENDIX 1

Archaic and Classical Documentary Letters

The arrangement followed is regional, with the two regions that provide the most instances
of private letters, the Black Sea area (nos. 1–22) and the gulf of Massilia (nos. 23–31),
coming first, then Sicily (no. 32), the Chalkidike (nos. 33–4), and finally Athens (nos. 35–42).
Within each area, inscriptions are listed in rough chronological order, with uncertain cases
or references to unpublished documents at the end of each group. I have indicated the first
edition and the main contributions, but have not attempted a full bibliographical coverage for
each document; the apparatus is likewise selective.

A. Certain and Probable Letters

Black Sea Region
1. Private letter, dated 550–500 bc, on a tightly rolled-up thin rectangular lead sheet
(c.15.3
6.5
0.1 cm), inscribed on both sides (opisthographic), found in Berezan
(Olbia) in 1970.
Vinogradov 1971 (with photograph and facsimile; cf. SEG 26, 845); Bravo 1974 (cf. Bravo
1977); Dubois 1996, no. 23 (with facsimile); Wilson 1997–8: 35-8; Trapp 2003: 50–1, 195–8;
Bravo 2011 (with BE 2012, no. 306); cf. Eidinow and Taylor 2010, E1.
Outside:
åغºæ- e º-
Ø Ææa e ÆEÆ
ŒI Æ Æª æÅ .
Inside: t —æøƪ æÅ, O Ææ Ø K غº-. IØŒÐÆØ
P e ÆÆı· -ºÐÆØ ªæ ت ŒÆd Ð
çæŪ- I æ . KºŁg Ææ’
Æ Æª |(<-)æÅ
I ªÅÆØ· çÅØ ªaæ ÆPe
Æ Æª æø
5 к ’ ÆØ ıŁ  · “¼’
Æ< Æ>ª æÅ å-
ήd 
-º- ŒÆd 

-ºÆ ŒNŒÆ”· O b I Æ HØ 
ŒÆd h çÅØ ’ ÆØ Pb KøıHØ  ŒÆd ÆÆ<ıØ>
ŒÆ çÅØ ’ ÆØ Kº Łæ ŒÆd Pb ’ ÆØ Køı<H>Ø
ŒÆd ÆÆ{Æ}ı<Ø>· C-  Ø ÆPHØ  ŒI Æ Æª æfiÅ, ÆPd
10 YÆØ ŒÆa çA ÆP
-. ÆF’
Æ Æª æfiÅ ºª-
fi ªı ÆØŒ. æÆ  Ø K غº-· c ÅæÆ
ήd B
ŒÆd -̀ I<º>ç , <>Y KØ K
æ Ø ÅØØ , ¼ª- K c ºØ ·
ÆPe  ª’ O -æe KºŁg Ææ Ø <N>ŁøæÆ ŒÆÆ ÆØ.
_ _
Outside : Achillodoros’ piece of lead, to his son and Anaxagoras.
Inside: Protagoras, your father sends instructions to you. He is being wronged by Matasys,
for he is enslaving him and has deprived him of his cargo-carrier [or: of his position as a
carrier; or: of the shipment]. Go to Anaxagoras and tell him the story, for he [Matasys]
asserts that he [Achillodoros] is the slave of Anaxagoras, claiming: ‘Anaxagoras has my
property, slaves, both female and male, and houses.’ But he [Achillodoros] disputes it and
denies that there is anything between him and Matasys and says that he is free and that
there is nothing between him and Matasys. But what there is between him and Anaxagoras,
they alone know. Tell this to Anaxagoras and the (his?) wife. Besides, he sends you these other
instructions: take the (your?) mother and the (your?) brothers, who are among the Arbinatai,
336 Appendix 1
to the city. The ship-guard himself [or: Euneoros himself], having come to him, will go
directly down [or: down to Thyora].
2. Private fragmentary letter from Berezan, c.540–535 bc, found in a pit with pottery
dated not later than the third quarter of the sixth century bc. The text is inscribed
boustrophedon on two lead fragments, measuring c.4.5
5.5
0.05 and 4.8
8.7
0.05
cm respectively and folded three times over in ‘concertina’; the central part is lost.
(A. S. Rusjaeva, in Kultura naselenija Ol’vii i ee okrygi v arkhaičeskoe vremja (Kiev, 1997),
152, citing Vinogradov); Vinogradov 1998: 154–7 (with photographs and facsimile; SEG 48,
988); Dana 2007: 70–2; Bravo 2007: 58–66. Cf. Eidinow and Taylor 2010, E3.
—Ææa [F E  vel e E Æ - - ]`[ ][ - ] r -
ÆØ, ŁºØØ[ - - - - - - - - - -‰ ]åØÆ__’ -_
ÆØ : ˜[ - - _- - - - - - - - - I ?]  : H ªaæ
K Ł  ¼Æ [ -ºÆØ ? I ]ºº- , P-
5 _ 
Æ ÆØ [ - - - - - - - - - - ] ÆEÆ _ Ø ¼ªØ
Ææa ºÆ [ - - - - - - I ?]Ð ÆØ
_
1: [?] Vinogradov || 1–2: r ÆØ = r , ed. pr.? (in SEG 48, 988); recent Ionian imperative (Dana);
_ infinitive, used in place of the second aorist imperative N ? || 4: ¼Æ=¼Æ (with psilosis; the
aorist
second letter is a sampi, ed. pr., with parallels. || 4–5: P Æ ÆØ = P Æ EÆØ, middle/pass. perf. of
 Æ ÅØ: ‘to slacken, remit a little’ (‘he gave in’?) or impersonal passive form (‘given in’), ed. pr. || 5: c
Vinogradov 1987, and Dubois in Dana 2007;  X Vinogradov 1998.
From X [to Z] say (or: To X, Z said), I should like . . . to be as soon as possible . . . you return;
for the things [property, money, goods] that (I want?) to send from here, he relinquished
. . . the slave-girl whom he brings to you from Melas . . . send back.
Vinogradov’s understanding of the first two lines, ‘From such and such (or to such and
such) . . . what would he have said? I would like . . . ’ (1998: 156), is not very convincing. The
aorist infinitive r ÆØ at ll. 1–2 (a form familiar from Herodotus) may be here used in place of
the second aorist imperative N , as suggested by Dubois. This has then to be understood as a
message to one person (see also 5: Ø), who is asked to transmit a request to others (ŁºØØ . . .
’ ÆØ . . . , and the second person imperative plural ]  at l. 3), as well as to do something
himself (l. 5: Ø). Alternatively, on the first editor’s proposal that r ÆØ = r , we might assume
that the first line contained the name of the addressee (—Ææa [e E Æ) and that of the sender
in the nominative, followed by a verb of saying in the past (To X, Z said: ‘I would like’); one
could compare with the early literary letters (‘Amasis to Polykrates thus says’). The oscillation
between plural and singular adressees in what follows (I ?]  l. 3 and Ø l. 5) is not really
problematic, as it can reflect a situation in which a person is the main representative of a group.
A further problem is caused by the difficult P Æ ÆØ at l. 5: Vinogradov suggests that this may
be a third person singular perfect middle ( Æ ÐÆØ =  Æ EÆØ ‘he has relinquished’, i.e. he has
no further claims on), which would introduce yet another person in the transaction. Bravo 2007
proposes a very different interpretation, based on the hypothesis of a religious text (three
‘Orphic’ bone tablets dated to the fifth century bc have been found in Olbia, SEG 28, 659–61);
the phraseology could be compared to that of other ‘Orphic’ lamellae: at 1. 1–2, Ææ Æ[Pc c
ºıc EØ Œø ]Æ[]Ø[ ] r - | ÆØ, with the injunction to ‘tell’ frequent in the_ lamellae;
at 1. 5, [ - - ˆB ŒÆd  OæÆ_ Ð_] ÆEÆ could refer to the mystic identity of the dead woman. If this is
correct, this piece of lead would be the earliest Orphic instruction text to date, and not a letter.
3. Private letter on a lead sheet broken on both right and left margins (3.8
3.8
0.05
cm), opisthographic, from the agora of Olbia, c.525–500 bc.
Vinogradov 1998: 157–60 (with photographs and facsimile; SEG 48, 1011); Dana 2007: 72–5;
Bravo 2007: 66–74; cf. Eidinow and Taylor 2010, E4. Punctuation marks (:) are present
only in the recto; this contrasts with the fact that the two sides seem to have been written
 Appendix 1 337
by the same, very careful hand (so Vinogradov 1998: 157–8; the photographs bear this out).
As pointed out by Eidinow and Taylor (2010: 55), these are indeed probably two letters:
the lead has been reused, possiby by the same writer. Bravo however recognizes hexameters
in B and consequently interprets the tablet as containing Orphic religious and moral
recommendations.
A. [ - - - ]ªÆ ŒÆd[ - - ]
[ - - - ]K NØ : /[ - - ]
[ - - - ]æ æ ¯ (or ˜) [ - - ]
[ - - - ] ŒÆd : Œ _ -º[_ ?-- ]
5 [ - - - ]ŒÆØæe : ÆFÆ/ [ - - ]
[ - - - I]_ Å : I Æ[ - - ]
[ - - - ]Ka ÆP[ - - ]

B. [ - - - ]ÆØ ıNE[ - - ]
[ - - - ]Œ OŒH NÆ[- - - ]
_
[ - j I ]ØH, ŒŁ [ - - ]
[ - - Ø]æ ŒÆNÆ-[ - - ]
-
5 [ - - I ]- Å[b ?- - ]
[ - - - ]K ø [c e.g. ºø - - ]
_ [Æ- - ]
[ - - - ]ŒÆd [Ø]æ
[ - - - ]Ø. Vacat _
B 2: Vinogradov; [Ø ªa]æ oø Ær [Æ Bravo k 6–7: Dana; []ıæ [Æ ed. pr
_ _
A. . . . big and . . . for the blue dye . . . before . . . and black clothing . . . opportunity, these . . .
good value . . . if . . .
B. . . . to the son . . . and equally eight . . . if I give back, receive . . . of iron and as many . . . not
give anything back . . . being not [? more than?] . . . and woollen coats . . .
A. The fragmentary state of the document makes it difficult to understand its content;
comparison with the letters on lead of Artikon and Mnesiergos (below, nos. 12 and 35), as
well as with those scratched on ostraka from Kerkinitis (below, no. 8), from Gorgippia
(below, no. 13), or from the Athenian agora (no. 32 and 34) leads one to suppose that this
text too probably referred to commodities and household goods (so Vinogradov 1998: 160,
Dana 2007): isatis indicates a blue dye; kandolos might refer to a Lydian dish (Vinogradov
1998: 158), or to black clothing (Dana).
B. At l. 1 ]ÆØ may be a verb-ending in the first person. Eidinow and Taylor 2010: 32 and 55
suggest that these are ‘directions to the writer’s son’; this is possible (it would work better if
]ÆØ were the ending of the writer’s name, which is impossible, or the son’s, which is hardly
likely in Ionic). Other interpretations, in which the son is a third party, are also possible. At
B. l. 3 the two verbal forms I ]ØH and ŒŁ’ refer the first one to the sender, in the first
person, while the second must concern the addressees. Objects in iron are mentioned in B l. 4,
and fur coats in B l. 7. Vinogradov 1998: 159 points out that ŒÆNÆ- corresponds to ŒÆd
NÆı, and that this form is attested in the Milesian lex sacra of the Molpoi (LSAM 50. 9/10).
4. Private letter (or tag?) on a lead sheet (3.4
7
0.05 cm), folded four times, and
damaged at the bottom, from Phanagoria, c.530–510 bc.
Vinogradov 1998: 160–3 (with photograph and facsimile; SEG 48, 1024); Dana 2007: 87–8;
Bravo 2007: 55; cf. Eidinow and Taylor 2010, E11.
 ˇ ÆE : y | K BæıŁ  | K æŁÅ : Z Æ : | ÆPHØ :
ƺºÅ, |  Æ : Łº : | [ - - I ][ ]ŁÆØ
_ _
2: K=K ed. pr.; K=Kª?; K=K Dana (genitive locative). Borysthenes: Berezan or Olbia, ed. pr. || 4: the slave
has a Greek name; he either was enslaved in Olbia’s hinterland, or was a threptos in an Olbian house,
ed. pr. || 5–6: [åæB ? I ][ ]ŁÆØ. ‘we wish (all debts?) to be paid’ ed. pr., who proposes to restore åæB
_
338 Appendix 1
for åæÆ and suggests that the buyer from Phanagoria had paid a deposit and that he was now requested
to pay the entire sum.

This slave here was sold out of Borysthenes, his name is Phaylles. We wish all (debts?) to be
paid.

5. Private letter on a rectangular sheet of lead (15.8
8.5 cm) probably folded three
times over, from Olbia, c.500 bc (mentioned in Vinogradov 1981: 19; Vinogradov
1998: 157; Wilson 1997–8: 38–40; SEG 48, 1012).
Dana 2004: 1–14, on the basis of a facsimile (BE 2005, no. 366; SEG 54, 694); Santiago
Álvarez and Gardeñes Santiago 2006; Dana 2007: 75–6 (with facsimile); Bravo 2011 (and
BE 2012 no. 306). Cf. Eidinow and Taylor 2010, E5.
Outside:
Æ -æØ
¸ ÆŒØ
Inside: ¸ ÆŒØ
Æ -æØ· : a åæÆÆ ØºÅÆØ P   HæÆŒ-
ºø Ð ¯[O]ŁæØ : ŒÆa  ÆØ c c : c I º-
ø a åæÆÆ: a ªaæ åæ[]ÆÆ a KçÅ ’ ÆØ : ŒÆd  ø-
_ HØ : ŒÆd [¼]<º>ºÆ ZÆ f ÆPHØ K -
: çÆ :  K [Ø]ŁÐ ÆØ Køı
5 ŁÅŒÆ ŒÆd [æe] çÆ { }_ Kd
_ _ a a åæÆÆ K B : Na K ØŁ-
Å ØçŁæØÆ [æe]  HæÆŒºÅ ŒÆd ¨ÆŁÆÅ : a åæÆÆ -
 ! O [-, I ] --· ŒÐ Ø ªæ çÆØ : Z Ø e å-Ø I -
_ - _] _e : _ıºÅ[Łb Zº] K a ŒÆd YŒØ ÆBæ. vacat
[
[? Ł]º-; æd H NŒØÅø ¨ıºø (sic)
10 ( - - c.10–11 - - ¯O?]ŁıøØ : ÆPHØ Ø PŒ r Æ j {} ª ÅÆØ ’
Vacat.
6: the name could be masculine (from Thathaies, as Wilson 1997–8) or feminine (Thathaie, as Dana
2004 and Eidinow and Taylor 2010). || 7: SEG 54, 694 (‘a verb in the future, third person singular,
followed by the Ionic ŒE Ø is probable, but T O [-] is problematic; the reading of an omega on
Vinogradov’s drawing is not certain, and one _would expect Iç’/Ø’ › -’, Avram); O [Æ? c.4 ]·
KŒE Ø Dana; ‰ O [-, I ] · ŒE Ø Santiago Álvarez and Gardeñes Santiago (cf. SEG _ 37, 838 l.
_ _ of value) ||_ 8-9: a line marking the end of the text was traced, possibly for the
13: OΠ; a genitive
purpose of cutting the lead (Dana 2004); but more text was then added.

Apatorios to Leanax.
To Leanax, Apatorios. I have had the goods confiscated by Herakleides, son of Eotheris; by your
influence I will not lose the goods. For I said the goods were yours, and Menon said that you had
entrusted to him also the other things that you have entrusted to him, and he added that the
goods that I had were yours. If you would send the records to Herakleides and Thathaie(s), how
much(?) your goods(?) . . . for they say they will hand over what they have of yours. The total of
goods seized is twenty-seven staters. What will you decide? Concerning the domestic slaves of
Thymoleos . . . to Eothymios; as to myself, I do not know if it is going to end well.
The situation is similar to that described in no. 1. Particularly interesting here is the
mention of ØçŁæØÆ, a document written on leather (here, ‘records’). The accusative
‘Thathaien’ has given rise to speculations: it is a ‘Lallname’ similar to the one attested,
certainly for a woman, in Cumae (ÆÆÅ, SEG 47, 1475). If this is a woman, she might be
the wife of Herakleides; at any rate, she would seem to have a fairly important role
(a woman, the wife/mother, was mentioned also in no. 1, but her position was different,
as she seemed to depend on Protagoras). However, as ¨ÆŁÆÅ can be the accusative of a
male name, it may be preferrable to see here a male associate of the opponent (so Wilson
1997–8; Chaniotis, in SEG 54, 694).
6. Private letter on a ceramic fragment (amphora or oenochoe, dimensions 6.7
9.5 cm)
of Fikellura style from Olbia, c.500 bc.
 Appendix 1 339
Rusjaeva and Vinogradov 1991: 201–2, with date to c.550–525 bc (with photograph and
facsimile; cf. SEG 42, 710); Dubois 1996, no. 24, with date to c.400 bc (facsimile); Dettori
1999: 283–5 (with date ante 500 bc); Bravo 2000–1: 162–4 (cf. EBGR 2001, no. 26); Bravo
2001: 253–64 (SEG 51, 970). Cf. Eidinow and Taylor 2010, E12.
[ - - - c.9 - - HØ Iªø] ŁfiÅ ºØ ŒÆd ŒæØe[ - - c.11 - - - ]
[ - - - -c.14 - - - - ]F ‰ K غºØ  [ø - - c.9 - - - ]
[ - - c.10 - - - Æ ]e  ı Ł Øı_ æØ[ - - c.9 - ]
_
[ - c.9 - - Øa ªa]æ ¼ŒæÅ ºÅ< >ŒÆ  Iæ çø_ Å[Å ŒÆd]
5 [ - - - c.13 - - ] ØŒÆø. K Bfi "ƺŒÅfifiŠƃ ªı[ ÆEŒ NØ]
[K ÆæÆåBØ º]ºBØ. K ŁFŁ K c  ºÆÅ[ Ø Å ]
[ - - c.12 - - ] ÆsØ ƒ ød  ºÆ [Ø_ N]
[ - - c.10 - - ]Åæe ¨H ŒÆd BæıŁ< > ŒÆd  HæÆŒº[ı - - ]
[ - - c.8 - ]a e ÆıªØ ƒ FºØ ŒÆÆæÆ[  - - ]
10 fi ƒæıæªÅØ Åæç  ƒæe Kº _ [- - - ]
[ - c.9 - - ]B
_
[ - c.10 - - ]H Øø ŒÆŒÆd H  æø ØÅŒ ØÆ
[ - - c.12 - ]ƒ ŁÅæıÆd H ¥ ø ÅæŒÆØ a ŒØ _  ø º[ºH
__
1: Bravo; [- -] ŁÅ SEG 42, 710 and Dubois || 2: [ŒÆd Æ]F ? and  [Ø - - ] SEG 42, 710; [- -]ı and
 [ø Dubois || 3: Æ ]e (for  Æ) Bravo; [ - - ‹ ]ø  ı Ł_ Øı æØ[æÆBÆØ] SEG 42,
710;_[- - K] and æ[- - ] Dubois;
_ in fine æØ[ ºÅÆ (or_ - ºøÆ) ÆP ] with a first_ person, Bravo || 4:
_
 Iæ çø (possibly  IæçH _
) Bravo; perhaps  IæçH{!} Å[ - - ] Avram, SEG 51, 970; [ŒÆd a]æ’ ¼ŒæÅ ¼ªfiÅ,
ŒIb ƒæÐ çH Å[ - - ŒÆd] SEG 42, 710; [ - - ]PAKPHNAIHKAME ƒæg çH ˝˙Z[ - - ] Dubois, who
proposes at the end [ı] (‘light of the sanctuary of the island’?) || 5: Bravo, who proposes as the
beginning of the line [ ºE PŒ XŁº] or [ ºE I Œ Å] and considers for the end also qÆ instead of
NØ; [ŁøØ ?] - - "ƺŒ fi Šƃ ªı [ÆEŒ SEG 42, 710; no restoration before ØŒÆø and after ªı [ÆEŒ - ]
in Dubois || 6: Bravo; [› b? KŒ ]ºBØ SEG 42, Dubois || 7: [ - - ] ÆsØ - - [ - N - - ] SEG 42, 710 and Dubois;
Bravo suggests [K fiH  Ø] ÆsØ or [K HØ ƒæHØ K ]ÆFØ || 8: BæıŁ< ø> (or < Ø>) ŒÆd
 HæÆŒº[B] SEG 42, 710; BæıŁ< > ŒÆd  HæÆŒº[- - ] Dubois || 9: [ - - ]Æ and ŒÆÆæÆ[ 
- ] Dubois; [ - - ]a and ŒÆÆæÆ[E ] SEG 42, 710; [KŒÆØ b ]a, [çÆ  ÆØ b ]a, [ºª _ ÆØ b
_
]à or [Kºª  b ]a and ŒÆÆæB[ ÆØ]Bravo || 10: [ - - å]ØæıæªÅØ SEG 42, 710; [ - - ] ƒæıæªÅØ
- - - ºØ  [ - - - ], Dubois; [F  ’ K ]B fi _Bravo || 11: SEG 42, 710 and Dubois; ]H
fi or [F  ’ K d ]B
Øø {ŒÆ}ŒÆd H <¼ººø >  æø ØÅŒ ØÆ Bravo || 12 beginning: SEG 42, 710 and Dubois; Bravo
proposes to restore [ŒÆÆŒŒ Æ (or ŒÆÆŒŒÆı _ Æ)]; end of line, Bravo;  %` SEG 42, 710, Dubois

. . . I am sending to the agonothetes honey and a ram . . . as you request . . . (I have inspected?) all
the places made by the gods . . . because, following an excessive fasting, Hirophos was ill . . .
rightly so. In the Chalkeie the women (are in great agitation?) . . . Thence (I moved) towards
Hylaie . . . the altars have again been damaged . . . of the Mother of the Gods and Borysthenes
and Herakles . . . after the shipwreck the slaves have run away (?) . . . (only) the sacred slave of
Metrophanes was left in the religious service . . . of the pines and (?) of the (?other) trees two
hundred (had been damaged?) the hunters of the horses have found with many dangers . . .
According to Bravo, this is an official letter, sent by an Olbian magistrate, or possibly a
priest, to another, higher magistrate; the sender was probably writing from Berezan
(BæıŁ ), while the addressee lived in Olbia. It would seem that the writer begins by
detailing his next action, sending honey and a ram to an agonothetes (presumably in charge
of organizing competitions and a sacrifice in honour of Achilles). The deictic F , the
second person present K غºØ (‘as you request’), and the verbal form  [ which,
although fragmentary, speaks for a present tense, all point to this. Then, begins a report on
events past (too little of the end of l. 3 remains, and l. 4 is restored; but ll. 6 to 12 clearly refer
to events in the past). For a detailed commentary on the religious and historical aspects of
this letter, see Dubois 1996: 55–63 and Bravo 2001, who both highlight a number of
Herodotean parallels. The text is very lacunose, and as a result any interpretation must
remain hypothetical; even so, the second person singular at l. 2 makes it virtually certain
that this is a letter, and a very elaborate, lengthy one. It is all the more regrettable that
beginning and end are both lost.
340 Appendix 1
A further hypothesis advanced by Bravo (2000–1) is of interest here. The sherd presents
lines criss-crossing it, and the letters are covered in reddish ochre; comparison with other
documents show that these features may have a magical significance. In this specific
instance, Bravo suggests that the letter was reused by an enemy of the writer for a curse,
as a magical object representing the victim. This fits nicely with the widespread perception
of the letter as very literally representing the person writing it (see ch. 1, 3 and n. 8 for the
ªæÆç as the image of the correspondent).
7. Private letter on a rectangular lead sheet (3.4
6.7 cm) folded in two, from Zhivakov
Hill (on the coast of the gulf of Odessa), found in a pit-dwelling dated to c.450–400 bc
(but according to Dubois in Dana 2007, not before 400 bc).
Vinogradov 1998: 164–6, no. 4 (with photographs and facsimile; cf. SEG 48, 1029); Dana
2007: 76–8. Cf. Eidinow and Taylor 2010, E7.
—æøƪ æÅ· i I ØÅ[åÆ]-
A K, Iºø, åø æd -
f n i æØ ŒÆd I ŒÆŒ Oз
Ka q   Å ‹Ø åı, []-
5 _
åØ( ) Nåø aæ <><ı> åæØ[ ]
b ¼ª- Æ›e læ
vacat
[ - - - ]¯[ - - - - - ]
1–2: Dubois, Dana (=I ØÅ[åÆ]-| AÆØ); I ØÅ[åÆ]| fi A  Vinogradov; I ØÅ[åÆ]-| A? || 3:
I ŒÆŒ Dana; I Œ< >ÆŒ Vinogradov, SEG || 4: Vinogradov. Ka {Å}   fiÅ{ } Dubois, Dana. ||
5: <><ı> ed., ˇ˝ˇ˝ lead _

Protagoras! If he is scheming against me, it will not bother me: I have at my disposal
something [proof, argument, etc.] regarding your kin, and I can easily refute (or remove
them). If he is really going to go to much trouble that quickly, then may he leave from here
quickly, and he may be thankful that a civilized . . .
The letter opens with a direct address (a nominative used as a vocative), as in no. 1; in what
follows the sender speaks in the first person, as expected (K, Iºø, åø), but it is
unclear whether he is referring to the addressee with a third person singular or whether we
should read in the verb a second person plural (I ØÅ[åÆ]-| A’, with the meaning ‘if you
people start to use your cunning against me’); the following æd , ‘concerning your
people’, is problematic in both interpretations. It may also be that the scheming person is a
third party, and not the addressee. Dana (2007: 78) states: ‘texte rédigé à la troisième
personne’, presumably meaning that the first person speaker (the sender) speaks of the
addressee in the third person. But this is not the case: cf. æd - | f at ll. 2–3. The text as is
restored currently presents an intriguingly high number of ‘persons’: a speaker in the first
person singular; an implied second person singular ( at l. 2); a third person, or a second
person plural, depending on how I ØÅ[åÆ] ÆØ is understood; a third person, with
Nåø. A further intriguing feature of the object is that the lead was folded, after being
inscribed, first along the horizontal axis, and then along the vertical one, in such a way that
the inscribed text remained on the outside (Vinogradov 1998: 164). One wonders whether
this is a result of a ‘refolding’ by the addressee, after the message had been delivered and
opened, or whether it was an original feature (the lead is inscribed only on one side, thus the
opening line, with the personal name ‘Protagoras’, might have been meant to function as an
address).
8. Private letter on an amphora fragment (dimensions: 9.5
7 cm), from the city of
Kerkinitis, near Olbia (but the text is in Ionian dialect, and thus not necessarily
written by a trader from Olbia), dated to the fifth century bc.
 Appendix 1 341
Solomonik 1987: 114–25 (with photograph and facsimile; SEG 37, 665); BE 1990, no. 566;
Vinogradov 1995: 66; V. O. Anochin, in Archeologia (Kiev) (1998) 1, 136–42 (non vidi: cf.
SEG 48, 1004), proposing a date c.400 bc; Dana 2007: 83–4 (with facsimile); Bravo 2011
(and BE 2012, no. 306).

Æ æØ ˝Å øØ·
e Ææå K r Œ
ı Œ Ø ŒÆd 猒 YÆ·
ŒÆd ’ ¼ªfiÅ Åb ¼æ
5 K· ŒÆd H H
I ÆŒH å·_ ŒÆd ‹Ø
ºB ª øŒ __ _
K e() !ŒŁÆ.
4: ŒÆd ’ ¼ªfiÅ ed. pr. (= ŒÆd a ¼ªfiÅ Åd, ‘let no-one take care of your affairs’); ŒÆd ªfiÅ, with apheresis,
Dubois in Dana, ŒÆNªfiÅ (ŒÆd KªfiÅ) Vinogradov 1995 (‘and no-one may import without me’);
Holford Strevens’ understanding of the passage, ‘let no-one lead you’, seems to me much better || 5:
ªH ed. pr.; H Pleket (SEG 37, 665), Dubois (SEG) || 6: ‹[Ø] ed. pr.; ‹Ø Dubois; ‹[Æ] Vinogradov
(sampi). _ __ _

Apatorios to Neomenios: bring the dry fish home and planks of equal size (? ‘bundles of
salted fish’, Vinogradov 1995); and let no-one lead you without me. Also take good care of
the oxen and learn who pays (a tax?) to the Scythians.
9. Lead letter, broken on all sides except the right, measuring c.5.8
5.9 cm, from
Panticapaeum, end of the fifth–beginning of the fourth century bc.
Saprykin and Fedoseev (2010a): 50–8 (in Russian; photograph and facsimile).
[-------------------]
1 [ - - - ] åÅÆ[Ø - - - - - - - - - ]
[ - - - ]BØ æHØ _ F [- - Ææ ÆP]-
[]F & Æ[] æÆå[ - - - - ]
ÆPe  ªø K [ -_- - - - - - - ]
5 Æ  – ŁæÆ Æ[Æ - - æ]-
_   , oºÆ [ - - - - - - - - ]

[]ƺ [ø] , º[ı - - ƺ]-
 ø KŁ[][Å -_- - - - - - - - ]
10 []ƺ _ ø [ _- - - - - - - - - - - ]
[]ƺ [ø _ -----------]
[ . . . ] åÅ[ÆØ]˜[ - - - - ]
_
[ . . . . . . . . . . ]ƺ [ø - - ]
[-------------------]
3: edd. pr. propose & Æ, aorist participle from ¥ ÅØ, with the meaning of ‘sent’, agreed with the
following dæÆå; a compound form would, however, be more acceptable here. || 4:  ªø K
ıºÆ ? K ŒÅ ?edd.pr. || 6:  > ø edd. pr. |

I am taken . . . coffin of the son . . . drachmas . . . myself I bring in (or under) . . . preparations
. . . cable, wood . . . talents, slaves . . . talents . . .
The first person  ªø (l. 4; cf. åÅÆ[Ø , ll. 1 and 12,  o? l. 6) make it clear that this is a
letter. But beginning and end are lost, and the body is difficult to understand.
10. Lead letter (c.6.5
4.5 cm), preserved on top, right and bottom, left side missing,
from Panticapaeum, c.400–350 bc.
Saprykin and Kulikov 1999: 201–6 (with photograph; cf. SEG 50, 704); Dana 2007: 86. Cf.
Eidinow and Taylor 2010, E9.
342 Appendix 1
 EæÆE[ fiH E Ø åÆæØ - - ]
 Œ [Æ - - ]
__ - - - ]
K غÅ[
Ø ªaæ [ - - - - - ]
5 _ [----]
I غÆ
øŒÆ Kç_ Æx  [c] K Ø[ºØÆ - - - - Œ]-
_ _
ÆØ Ææ ÆPF· ææø[].
3: Dana; K غ[ŁÅÆ ed.pr.

Hermaios (to such and such, greetings). Fifty . . . take care . . . for it is . . . I have sent . . . I
have given, over which the care . . . recover them from him. Farewell.
A small hole in the top left corner has been interpreted as implying that the tablet was
nailed to a wooden support; this is slightly surprising, as the information contained in the
text does not seem to warrant display. The hole might have preceded the incision of the
text: while lines 2 to 7 begin very regularly one under the other, the first letter of the first line
is slightly indented, as if to accommodate the hole.
11. Private letter on an amphora fragment of triangular shape (max. dimensions 6

2.5 cm) from Kozyrka (chora of Olbia), second half of the fourth century bc (350–
325, on the basis of the archaeological context).
Vinogradov and Golovačeva 1990: 15–30 (= Vinogradov 1997: 323–35, with photograph;
cf. SEG 42, 711; Dubois 1996, 49 n. 1); Dana 2007: 79–81 (with photograph).
[˝Ø]Œç Æ
æı | Zø ıæø Ø ¥ - | øŒ · I - | غø Ø | N ºØ | ŒÆd
 ø_ | [ÆP]HØ a ªæ/- _
verso ÆÆ. _ _
_
Nikophanas son of Adrastos gave Zopyrion a horse; let him send (it) to me in the city, and
give him the document.
Vinogradov has proposed to see in Zopyrion the general of Alexander who unsuccessfully
besieged Olbia in 331/330 bc—but the name is a fairly common one, and this is most likely
a private affair.
This has been interpreted as a letter; but it makes more sense to think that the document
registers the change in ownership of the horse, and that it is meant to be handed back in
exchange of the horse.
12. Private letter on an opisthographic lead sheet (13
3 cm), folded vertically in the
middle, from Olbia (or Berezan? Jordan 2000: 91). Fourth century bc.
B. B. Latychev, Izvestiya Arkheologicheskoi Komissii 1904: 10–13 (non vidi); Wilhelm 1909:
118–26 (with photograph) (= Wilhelm 1984: 338-46; Syll.3 1260); Dubois 1996, no. 25 (with
facsimile; date to 350 bc). Cf. Eidinow and Taylor 2010, E2.
side a:
æØŒH : E K YŒø<Ø>
åÆæØ : j Kª ºØ: KŒ B
NŒÅ :[A] ıººø :
Ææa
ÆŒı, [N] e YŒÅÆ,
side b: j ªaæ ØHØ· N b ,
Ææa
ªŁÆæŒ : N a
Ææa ˚æø[ ] Kæø
e æ ŒØŁø.
Artikon to those at home, greetings. If Myllion ejects you from the house, go to Atakes, to
the room there, if indeed he gives it (to you). But if not, go to Agatharkos. As for what we get
from Kerdon, let him take his(?) part of the wool.
 Appendix 1 343
13. ‘Letter’(?) on a triangular ceramic fragment from Gorgippia (Bosphoros), 350–325 bc.
Vinogradov 1997b: 232–44 (with photograph; cf. SEG 47, 1175); Dana 2007: 89–90 (with
photograph). Cf. Eidinow and Taylor 2010, E14.

ººø [ - - - - - - ]
ºŒ _ .[ - - - - ]
e HØ Æ()[ - - - ].
 ŒÅ -æ , ›Œ[ ?]
5 YŒøØ qºŁ I [E-?]
 r Æ Æçıºc
ŒÆd ºåÆ Æ.
1: ostrakon, Dana;
ººø [][Å HØ ? Vinogradov || 2: çı[æªÐ ? Vinogradov || 3 —`˜ ostrakon;
_
Æ(Ø)[øØ] or Æ(Ø)[ŒøØ] Vinogradov; Æ(Ø)[øØ] Dana _|| 4: the scribe initially had written ˇ—
(probably for the koine form › ), which he then corrected to Ionian ›Œ , Vinogradov, Dana. || 5:
the º in qºŁ has been inserted above the other letters. || 5–6: Dana (who comments on the psilosis
here) ); I [E] x Æ Vinogradov.

Apollon-[ ] give the axe to the slave. The gardener, when he came to the house, wine grapes
and vegetables had ripened (?).
14. Two joining fragments of a lead sheet (8.2
4.1
0.1 cm) found in a house in
Olbia in 2010, in an archaeological context of fifth or fourth century bc.
V. V. Nazarčuk, ‘Un nouveau fragment d’inscription sur lamelle de plomb d’Olbia’ (in
Russian), in The Phenomenon of the Bosporan Kingdom: Population, languages, Contacts,
St. Petersburg 2011, 471–4 (drawing and bad photograph); cf. Avram, BE 2012, no. 310.
. . . ˝%˜ `(?)%˙(?) . . .
 —˛ˇ  ˚`ˇ . . .
`˙!˙%ˇ ˙ ˙˝—¯˝ . . . .
¯˚`!˜¯˙ `  —ˇ˙(?) . . . .
5 ¯!!ˇ˝ˇ˚ˇ˝˛ ˝ . . . ...
TA
As Avram points out, none of this makes much sense, with the possible exception of the
fifth line, [- - ]Ø e r Œ f [-], which orients the interpretation towards a letter. An
intriguing feature is that the text seems to have been inscribed within a frame.
15. Fragment of right side of ceramic sherd (dimensions 7
7 cm.) from the agora of
Olbia, third century bc.
Shebalin 1968: 296–9 (non vidi); Dubois 1996, no. 26 (with facsimile; cf. SEG 46, 943). Cf.
Eidinow and Taylor 2010, E15.
[ - - - - - - ]E ÆŒºæ-
[Ø - - - ]E æe )H -
[ - - - - - j] c I H
[ - - - - - ]  æı Kå -
[ Æ - - - - ] Æ[ . . ]
Dubois. l. 4: [a F (or a  e) ]  æı Kå k[ Æ? ed. pr., SEG)

To the naukleroi . . . those before us . . . if you do not give back . . . what is held by
Menander(?)
The first person plural pronoun )H (‘before us’) certifies that this is a letter: Dubois 1996: 66.
16. Unpublished ‘letter’ on a rolled-up sheet of lead, from the cemetery of Olbia.
Mentioned in Vinogradov 1998: 154 n. 4, no. 3; Dana 2007: 81–2; cf. Eidinow and Taylor
2010, E6.
344 Appendix 1
Dana dates the document to fourth–third century bc, and considers it a letter, sent by one
Batikon (BÆŒø ) to one Diphilos (˜Øçº); the sender would be writing about difficulties
that he and his family had had in a trial. Bravo 1987: 206 suggests that the lead sheet should
be seen as a magic text, because of its provenance from a tomb; it is a letter sent by Batikon
to a dead man, Diphilos, who should use his power on the sender’s behalf.
17. ‘Letter’(?) on a ceramic sherd (an amphora fragment, 6.5
8.5 cm) from Panskoje
I in the Chersonesos, dated to c.350 bc.
Stolba 2005: 76–87 (with photograph and facsimile; cf. SEG 55, 859). Cf. Dana 2007: 82–3.
[ - - - ]`!ˇ˚¯—¯
[ - - - ]˝ˇ˚¯˜˜ˇ
[ - - - ]% ˚ıø 
[ - - - ]æÆØ K ØŁ-
5 [ - - - ]ıÆæØ
1–3: [acc. plur. fem.]Æ OŒ  - | [łÆ–? æ]d nŒ KØ/[ı - - b]æ ed.pr.; but it would be possible to
_ the second || 4–5: [ÆE ]æÆØ
read nΠfor doric nή also in the first line, or conversely OΠfor PΠin
K Ł- | [ K d ł]ıæØ ed.pr.; perhaps [e Æ]ØæØ , Avram in SEG].
_
The editor, on the basis of rather adventurous restorations, suggests that the text concerns
the making of sacrifices in the local cemetery. In that case, ‘letter’ might not be the right
designation for this document (but it is difficult to see what else it could be; it might be
better to dismiss the sacrifices). A further peculiarity is that this is the first document in
Doric dialect from the region. Dana 2007: 82–3, who does not attempt to propose a text or
an interpretation, concurs in considering it a letter.
18. Letter on a rolled lead plate found on the seashore near the ancient city of Hermo-
nassa, and now sold to unknown person; the publications are based on a photo-
graph. Dated by the editors on the basis of the letter-forms to the end of the fifth-
beginning of the fourth century bc. The lead (dimensions: c.3.2
20.4
0.15 cm.)
had been divided in two; the writer began writing on the left half, and when this was
full moved to the right half.
N. A. Pavlichenko, S.V. Kashayev, ‘Novaya epigraficheskaya nakhodka iz Germonassy’,
Drevnosti Bospora 16 (2012) 288–99 (non vidi); S. Yu. Saprykin, A. V. Belousov, ‘Pis’mo
Kledika iz Germonassy’, Drevnosti Bospora 15 (2012), 348–59 (non vidi); Belousov and
Saprykin 2013: 153–60 (with photograph).
Col. I Col. II
ὮæØ ŒæÆ : K غº ̄ Ø ı º Æ  : a ººØŁ Æ
̄
˚º ØŒ : ı Ł $ ªaæ K Æ ªÅ : ŒÆÆçæ(Æ)ŒÆ
Æ æ åÆæØ : a K ŁÆıÐ
K Æ : ØÆØA : ŒÆŒH<>
5 KºŁg : Ææa !øŒæÆ
K a NŒ[– – – – – – – – ]
Text: Belousov and Saprykin 2013. I 3: Pavlichenko;  ǣ åæØ Belousov and Saprykin. || 4 ØÆØA
Pavlichenko; ÆØÆ Belousov and Saprykin. ŒÆŒH Pavlichenko, Belousov and Saprykin, lead. || 6: K
a NŒÅ I ºøº Æ Belousov and Saprykin (for NŒÆÆ); K A NŒ[Æ ºÆ  e.g. - - Pavlichenko. II
_ __ _ _ ___
1: ººØŁ(ø) Æ Belousov and Saprykin.
O Aristokrates! Kledikos sends to you. For I learn that Mandrocharis administers what is
there badly. Going to Socrates, to the ruined buildings, collect what has been left and seal it
in one room.
 Appendix 1 345
The letter opens with a vocative, a feature that finds parallels in other letters (cf. nos. 1, 7,
31, 35, and 37, as well as possibly 13 and 25); the writer then presents himself in the third
person, again in a movement that finds parallels in numerous other early letters (note
moreover the use of K غºø, which finds a close parallel in no. 1 above). This is followed
by an enunciative shift to the first person; there is no greeting formula.
19. Letter on a rolled up, damaged lead sheet from Hermonassa, containing on the
outside c.12 lines of Greek letters.
Unpublished; preliminary discussions in Finogenova 2003 and Dana 2007; see also Eidinow
and Taylor 2010, E10. Finogenova 2003: 1019 reads only the word ªı ; photographs at
p. 1044, fig. 9. Dana 2007: 86–7 disputes the reading ªı  on the basis of the photograph;
she recognizes letters of different sizes, but cannot offer a transcription.
20. Letter on ceramic sherd, broken into two joining pieces found at different times
during the excavations of a house, from Nikonion, in the region of Odessa, Ukraine,
now in the archaeological museum of Odessa. The letter is dated, also on the basis of
the archaeological evidence, to the second half of the fourth–first half of the third
century bc. Dimensions of the two sherds: 9.8
10.5 and 9.4
7.5 cm.
Awianowicz 2009: 196–8 (A. Avram, in BE 2009, no. 360) and Awianowicz 2011: 237–9
(for the second fragment, and a text of the whole).
˜Ø Ø E K YŒø[Ø] åÆæØ . ø ı +æ<æø>ÆØ ŒÆd
›  . E b ƺƌÆ[Ø] ÅŁb K H, Iººa I -
غ Ø Æ ŒÆd  غÆØ ÆæÆŒÆÅ e  -
ıº ¼ı K  ºÅÆ. K æÆ øæ
5 ÆPe I Æ Æø, ‰ ºc KŒ ı Kº d
PÆ K . ŒÆd ŒæØŁH NØ Ææa —ØŒæ[]-
ÅØ Ø Ø K Æ. ææøÆØ b ŒÆd
—ØŒæÅ. {Ø} Œ[ ]ØÆØ b Ææa H
¨Æłø )Ø[]Åæ I F-
10 e ƒ[Ø .] Æ
Text: Awianowicz 2011. 3: Awianowicz 2011. ÆæÆŒÆÅ Awianowicz 2009; Ææa ˚ÆÅ or Ææ
΁
Avram (names); possibly a term linked to Iί , light boat?

Dionysios sends his greetings to those at home. Until now I am in good health, and so the
son too. You should in no way give in to weakness, but dispatch someone and instruct
Marakates (?), (that) I have loaded the light boat with sand. After emptying it out, he should
haul it up, for apart from this there is no hope at all. And there are nine medimni of barley
with Possikrates. Possikrates too is in good health. And get from the Thoapsoi(?) a half
stater, after returning the cloth.
As Awianowicz 2011 notes, Dionysios begins by greeting the members of his household, but
from ll. 2–3 onwards he switches to a singular, which must refer to a woman, as shown by l. 9.
The letter gives practical instructions, of a commercial nature (the sand may have been a
commodity, unless it served the purpose of preventing the boat from being carried away).
The mention of Possikrates attracts information on the health of the latter (l. 7, ææøÆØ);
then, the letter goes back to business matters; there are no final greetings. Remarkably, the
verb used at the beginning of the letter (if the restoration is correct) for the purpose of
indicating the good health of the writer and the son, as well as later, at l. 7, for Possikrates,
again to convey his good health, is a form of Þ ıØ, the verb usually employed in the final
valediction (‘farewell’, ææø), while for information concerning one’s own health one
tends to find forms of ªØÆ ø.
346 Appendix 1
21. Letter on a squarish fragment of amphora from Sinope (dimensions: 6
5.5
0.4/0.6
cm), found during excavations in the northern part of the Chersonesus Taurica, dated
by the editor to 375–325 bc.
I.A. Makarov, ‘A Greek Letter Found in Chersonesus Taurica and the Interpretation of the
Term !`!˙% (IOSPE I2 401)’, VDI 2009/2, 49–61 (photograph); cf. BE 2010, 462.
ØŁ ÅØ
åÆæ e
Ł ºÆŒ
ºF ŒÆd ºÆ-
5 _
b e KÆæªæØ
KŒ Ð ŁıºŒ
..A.H
The editor considers Ł ºÆŒ here = ŁºÆŒ ; he proposes: ‘To Timosthenes, greetings.
Open the bag and take the money from the bag . . . ’.
22. Letter from Nymphaion (now in the Hermitage).
Unpublished. Mentioned by Vinogradov 1998: 154 n. 4, no. 9 (cf. also Dana 2007: 85;
Eidinow and Taylor 2010, E8).

Gulf of Massilia
23. Private letter on a rectangular sheet of lead (dimensions: 9.5
14.2
0.1 cm)
found rolled up during the excavation of a building in Emporion, c.500 bc.
Sanmarti and Santiago 1987: 119–27 (photograph and facsimile; cf. SEG 37, 838); Sanmarti
and Santiago 1988a: 100–2; Sanmarti and Santiago 1988b: 3–17; Jeffery 1990: 78, no. 3;
Santiago 1990a, 79–80; Slings 1994: 111–17; van Effenterre and Ruzé 1994–5: ii, 268–71, no.
74; de Hoz 1997, no. 2. 14; Rodríguez Somolinos 1998, no. 1; Santiago 2003: 167–71 (SEG
53, 1153). Cf. Wilson 1997–8: 46–8; Eidinow and Taylor 2010, D2.
[ - - ]ø K !Æت ŁÅØ ÅØ, Œi [ - - - - ]
[ - - ]  _¯  æÆØØ P K Ø Æ[ - - - - ]
[ - - ] _ j  ŒØ Œr  PŒ KºÆ_[- - - ][ - - ]
_
[ - K !Æت] _ ŁÅØ O ø BŁÆØ BÆ
_ _ _ [ . . . ] [ - - - ]
5 -
[ - - ] Æ ¼æÆ ÆæÆŒ Œi [ . . . ]  [ - - ][ - ]
[ - - ]º. .ø Ø  ø Å [ . . ] [ - - ] _
- _
[ - - ]Æ ŒÆd Œº b BÆ [ . . ] KºŒ[ -- ]
[ - - ]ŁÆØ [Y] Ø Ø k º Ø K ˜[ - - ][ - - ]
[ - - M]æ Œi  ørØ,  æ[]Ł[ø - - - ]
10 [- - ]º  ø Œi ÆPe Łº[ÅØ . . . . . ]ŁÆØ[- - - ]
[- - þ]ıı åø. Œi c O[ - - - - - ]
[ - - ]ø ŒI غø OŒ  ¼ [ - - ]
[- - ] T i  ÅÆØ åØÆ[ - - - ]
[- - ŒŒ]ºıŒÆ. åÆEæ.
1: [ - - ]Z[Œ]ø Sanmarti and Santiago 1987 || 2: K Ø Æ[ - - - Slings; K Ø Æ[ ÅØ Santiago 1990a, de Hoz,
_
K Ø [ÆØØ _Sanmarti and Santiago 1987 || 3: º] _ van Effenterre–Ruzé
_ || 6: Santiago 2003; Æ.ø Ø
_
van Effenterre–Ruzé _ 1987, Slings; KæÆ ŒÆd Œºı Sanmarti _and
|| 7: μ- Sanmarti and Santiago
Santiago 1988a || 11: O[]º[ ªÅØ Santiago 1990a; van Effenterre–Ruzé || 12:  ]ø Sanmarti and
_
Santiago 1988a, van Effenterre–Ruzé

When (?) you are in Saiganthe and if . . . with/for the Emporitai, do not embark [ . . . ] (not
less than ?) twenty, and wine (?) not less than . . . in Saiganthe, that Basped- may buy a
ship . . . anchor to transport . . . what of this should be done . . . and he tells you . . . Basped-
. . . if there is someone who will tow until . . . ours. And if they are two, that he foresees
 Appendix 1 347
two . . . but that it (he) may be . . . ; and if he wants . . . that he may share for a half. And if
not . . . and that he may send instruction of how much he would . . . as rapidly as can be.
These are my instructions. Greetings.

24. Private fragmentary letter on a rectangular sheet of lead (dimensions 6.3
4
0.1
cm) found during excavations along the inner side of the Greek wall at Emporion,
opisthographic, end of the fifth century bc.

Sanmarti and Santiago 1989: 36–8 (photograph and facsimile; cf. SEG 39, 1088); Santiago
1990b (cf. Slings 1994: 111–17); de Hoz 1997: 41–2, no 2. 15; Rodríguez Somolinos 1998,
no. 8. Cf. Eidinow and Taylor 2010, D1.
A [ - - ]¸ [ - - ]
[- - ]d [] [ - - ]
[ - - ]þ [Å]Ł[- - ]
_ --]
[ - - ] O B[ÆØ
5 [ - - ]X Ø[ - - ]
[ - - ]ÆPHØ [b - - ]
[ - - ] P[[Å]]Œ Mı[ - - ]
[ - - ]  K
[ - - - ]
[ - -  æÅ]  ŒE [ - - - ]
10 [ - - - ]¼ººØ_ ŒÆd  O[ - - - ]
B [ - - ]`¯¸`% [ - - ]
[ - - ]`![.]`!¸ [ - - ]
_
[ - - ] ¼ æÆ Æ [ŒºÅæ ]
_
A. 2: Santiago 1990b, de Hoz || 3: Santiago 1990b, de Hoz; -]ø [..]Ł[—] edd. pr. || 4: Santiago 1990b, de
Hoz; - ] Å[ - a form of O [Ø? edd. pr. || 5: edd. pr. - - ]X Ø_ [ŁºÅØ Santiago 1990b, De Hoz || 7:
_
Santiago 1990b; P[[Å]]Œ Mı[ Å ?] edd.pr.|| 8: r]  K
[ edd.pr.; ]  K
[ - - qºŁ, Santiago
1990b || B. 1 [ - - Ææ]a غÆæ[ - - ] Santiago 1990 ||

A. . . . twice as much . . . you may have earned . . . to be advantageous . . . if someone . . . but

to him . . . not . . . he managed . . others and O– . . . B. . . . a shipowner and merchant
A. While this clearly has to do with buying and trade, and while the involvement of ‘others’
at the end is certain, the text is too fragmentary to allow any serious discussion.
B. This may have contained the address, possibly introduced by Ææ, if we see in Tielar– a
name (Etruscan?); but again, too little is left to allow solid conjectures.
25. Letter on clay tablet, dated c.500 bc, written in Attic dialect but found in the gulf of
Roses, between Ampurias and la Escala, now lost; boustrophedon.
Dunst 1969: 146–54 (with photograph); Johnston in Jeffery 1990: 464 n. A, who remarks:
‘possibly ancient text, known only from a bad copy with a most unusual alphabet’; de Hoz
1997, no. 2. 56; Rodríguez Somolinos 1998, no. 4.
"ÆEæ  ‚ æª Iºç 
K  ¯  ıºø
fi · PŒ N-
ı źF E Y I<> -
ºı· Iººa ªaæ åæÅ<>a
5 ŒŒÆıŒÆ Iº<ç>Ø , <þ> `
ŒÆd fiH åƺŒfiH Iæªı-
æ ¸—[c.5] ˇ H Œ-
_ _ æı
æø _ _ _
_ _ __
IŒ ı ÆØ e Ø K
10 fiH ø fi — ·¯%¯ Øa f
ıææÅ f_ N · ›
348 Appendix 1
F Œı æ ı çº,
Ø Œ, çæø c K Ø-
ºc ¯%¯¯˝ c Iæåc
15 K BØ{ } I _ ØŒÆ
fi · ƒ Ø-
çæØ ªºÆıŒd ŒæÆ-
æd ŒÆd ƃ ºÆØ ÆØ ˚`¸

Greeting, Energos brother in Empylion. There is no need of a particular clay or soot; for
I have cooked some excellent (vases) to paint over, in the same way as with bronze in silver
(is done). The cooking of the vases in the usual way (?) broke a number of . . . Ask about
Etruscan ovens; the friend of the pilot of the ship, Tibekos, bringing the letter . . . authority
in the colony; the excellent green strong ones (or: the excellent green are strong) and the
black . . .
26. Fragmentary private letter on on a tightly rolled-up thin rectangular lead sheet
(dimensions 17.3
15.5 cm), from Emporion, end of the fifth century bc.
Almagro Basch 1952: 33–5, no. 21 (photograph, facsimile; cf. BE 1955, no. 282); Jeffery
1990: 287; Jordan 1985: 183–4, no. 134 (SEG 35, 1071); de Hoz 1997: 42–3, no. 2. 16 =
Rodríguez Somolinos 1998, no. 5, whose text I follow (Eidinow and Taylor 2010, D3 must
refer to this letter).
[ - - ]ŒÆØ[ - - ]Å [ - - ]
[ - -] Ø ø Å[ - - ]
[ - ŒÆ]d —ıŁÆª æ[Å] K[ - - ]
[ - -]
ªÆŁŒºB[] [ - - ]
5 [ - P]   Ø c [ - - ]
[ - ‹]Œ KŒÆØŒÆØ[ - - ]
[ - -]  Æ ı[ - - ]
[ - -]ŁÆ ŒÆd B ªæ[ÆçB - ]
J. and L. Robert, BE 1955, no. 282, advanced the hypothesis that the text might be a letter,
private or public; Jeffery 1990 and Jordan 1985, no 134, prefer to see in it a curse. Because of
the presence of a verbal form in the first person plural and of pronouns in the second
person plural, a letter seems more likely (so de Hoz 1997: ‘una carta publica o privada’).
27. Lead letter (?) dated to the third century bc (?), found under the church of Santa
Maria at Roses (Rhode) and destroyed during the Spanish civil war. Eight lines of
extremely fragmentary text.
Oikonomides 1983: 107–9 (edited from a photograph published in M. Oliva Prat, Arqui-
tectura Romanica Ampurdanesa: Santa Maria de Roses (Gerona, 1973), 12), cf. SEG 33, 841;
de Hoz 1997: 32–3, no 1. 1.
[ - - - ] Ø | [ - - - ].`˝ ˜ | [ - - - ].¸˚? | [ - - - ]`¯¯[?] k [ - - - ] ˝˜¯ | [ - - - ].
˚`—ˇ[?] | [ - - - ].ˆ¯?˝˝? | [ - - - ]åØ
Text: de Hoz. 2: ÆEÆ Oikonomides || 3: K] [ Oikonomides|| 4: ]º  Oikonomides || 5: ]ı -
Oikonomides || 8 P]åØ Oikonomides
Oikonomides based his restorations (‘fanciful’: SEG 33, 841) on the assumption that this
was a letter. The only element that might speak for it is the possibility of recognizing a
greeting at the end, P]åØ, ‘farewell’.
28. Letter on a lead sheet folded in two, found in 2005 at Lattes (Hérault), ancient
Lattara, in the refill of the floor of a building constructed in c.430 and destroyed in
c.415 bc. Opistographic; dimensions c.5.5
5.6 cm. Inscribed ‘quasi stoichedon’ by
the same hand in regular letters, and dated to the first half of the fifth century bc
(and at any rate before 430 bc).
 Appendix 1 349
Bats, in Py 2009: 303‒5 (photographs and transcription); Bats 2010: 749–56 (photographs);
Bats 2011: 202–4; cf. Decourt, BE 2012 no. 512. I am grateful to M. Dana for pointing out
the existence of this document; I give the text offered by Bats.
A. t Z ª[ - - - ]
Ø º[ _ - - ˚º-]
Ł Å[ - - - - ]
 Æ[BæÆ - -]
5 t Z ºÆ[ - - ]
 [ -]ÆØ ª[ - -]
˚º ÆŒÆ[ - - ]
 ÆB[æÆ - ]
,—¯¸`ˇ˝[ - - ]
B. I ÆØBÆØ
ŒEŁØ ªÆæ-
ºÅ  OŒ-
 Æ
A 4: an adjectival form linked to the following Æ[BæÆ, such as Ø or æØ , Bats;
ending of an imperative aorist? So also 8.
A.: O Zeus! . . . more . . . (K)leosthenes . . . a stater . . . O Zeus! six and one eighth (?) . . .
Kleanax . . . stater . . . for the oil . . . B.: Demand back from there garelaion (? garos with
olives, Bats), two octana (? a measure, Bats)
About one half of the length of text on A is missing, and this renders attempts at
restoration problematic. The side A presents a regular pattern of repetition (the sequence
,Z˙˝, personal names in the accusative, reference to a stater); the sequence of letters
,Z˙˝ is difficult to explain otherwise than as an invocation to Zeus (see Michel Bats’
commentary). Both sides of the lead concern transactions having to do with olive oil; but
the repeated invocation (if that is what ,Z˙˝ is) and the lack of verbal forms comparable
to I ÆØBÆØ on B (if the two – at l. 4 and 8 are not to be considered as aorist
imperatives) may mean, as suggested by Bats, that A is an order for payment, while B is
a further request.
29. Private letter on a rectangular rolled up lead sheet (c. 18.2
c. 2.7 cm.) inscribed on
both sides (and never opened), found in the harbour of Marseilles, third century bc.
(Hesnard 1999: 44); Decourt 2004, no. 4 (BE 2000, no. 751; SEG 54, 983).
A ¸Œø Ø
B ªØB ¸Œø Ø åÆæØ · N ªØÆ Ø, ŒÆºH E·
ªØÆ _ b Œ[Æ]d )E. ˇsºØ
_ Ø K ıå  bæ B
I ŒæÅ I ØH _ÇÅE KŒ Æ e æ ı ‹ ø i ºıŁÅ·
› åæ  · ø ›
ÆıæØ · ŒÆd ÆPe çÅ æ[ . ]CE[ .1–2]C¸
`   ¸ (?)¯ˆ¯˝ ˆ%[ . . . ]``¸ˆ[ . ][ . ][ .1–2]I[ - - ]· PåØ
_
To Leukon.
Megistes to Leukon, greetings. If you are in good health, you do well; we are in good health
as well. Oulis has contacted me, asking, concerning the anchor, to try in any way to leave.
The moment: let it be Apatourion. And he himself said . . . Farewell.

30. Opistographic lead sheet from Agde, broken in seven fragments (dimensions of the
whole c.8.5
4.5 cm), of which three could be joined; dated to the fourth century
bc. Now lost.
350 Appendix 1
BE 1955, no. 282; BE 1956, no. 357; Lejeune 1960: 62 (BE 1961, no. 844); Decourt 1993: 134,
no. 13; Decourt 2004, no. 130 (photograph and facsimile). Eidinow and Taylor 2010, C1
mention this letter, but do not take into account Decourt’s publication.
side A: 1. ŒÆd l. 2. ÆªÆŁÅ and possibly ÆPe æH l. 4 Œı l. 5 åÆæØ ŒÆd -
l. 6 Æ تÆæ-
side B: only a few groups of letters can be recognized: ı, PŒ. Decourt was not able to
confirm Lejeune’s readings åÆæÆ and æ ø.
31. Private letter on a ceramic sherd, from Olbia in the Var (southern France), close to
Hyères, dated to third or second century bc.
Coupry 1968: 244; Coupry 1971: 141 (B¯ 1971, 728); Koumanoudis 1986: 158–9 (SEG 36,
949); Decourt 2004, no. 71. Cf. Eidinow and Taylor 2010, E13.
¯PåÅ | I ºH Ø | K غc | æd H | øÆø

Eutyches, I will send you a letter concerning the slaves.

Sicily
32. Opisthographic lead sheet (8.5
5 cm.) folded in two lengthwise, from Himera
(Sicily), c.470–450 bc. The characters on A (outer face) and B (inner) are of similar
dimensions (0.3–6 cm.), but while A is very legible, the inner face B is very damaged.
Manni Piraino 1969: 301–4; 1970: 385 (A only); 1972, 107–9 (A and B, photographs,
facsimile); Dubois 1989, no. 11 (A only); Arena 1994, no. 51 (A only; pl. 15.2) (SEG 44,
753); Grotta 2008: 255–64, no. 3.
A: ¯P- Æ hƺ-
˜Øå- : ºåǢª 
˜ÆEØ.
B: [ - - - hå]ÆØ Y Æ  -
-· [ - - - - - - - - ] KØ
[ . . ]  [ - - - s ¼] ŁÆØ
_
[ . . . ]ØØ [ - - - - - - - ]
_
A. 1: hغ Manni Piraino, Dubois, Grotta; hƺ- Arena || 2: Ø På- Manni Piraino; ˜Øå- Dubois,
-
Arena, Grotta. ¸ˇ"`ˆˇ! Manni Piraino; ¸ / ºåƪ Dubois; ºåǢªe Arena, Grotta. || 3: ˜`!
Manni Piraino; ˜ÆEØ Dubois, Arena, Grotta. B.: Manni Piraino. The photograph does not allow to
check the readings of the B side; but Manni Piraino’s own facsimile indicates nothing at the beginning
of B. l.2; and E˛¯ at the beginning of B. l.3.

The text has been variously interpreted as a letter of recommendation (Manni Piraino:
‘Euopidas was sent through your request: lochagos Daitis’); as a military message (with
hغ as an aorist passive from ƒººø, Dubois: ‘Euopidas has been sent, and Dieuches;
commander of the company, Daitis’); or as a defixio, later reused as a projectile (with hƺ
as aorist active, Arena: ‘Dieuches Euopidas sent; commander of the company, Daitis’). The
interpretation of the whole depends on whether the text on B is considered contempor-
aneous and connected to that on A, as Manni Piraino does, or not. If the two texts are taken
together, then the second person in B l. 2 (KØ) is indeed best interpreted as part of a
message, a letter: ‘I ask(?) that you do not allow . . . foreigner(?)’ (as Curbera 1999, 181
no. 51 points out, a second person in a defixio of this date is unlikely); however, the reading
of part B is extremely uncertain. See further the careful discussion by Grotta 2008.

Chalkidike
33. Rectangular rolled-up lead sheet (5.2
15 cm.) from Torone (Chalkidike), c.350–325 bc.
 Appendix 1 351
Henry 1991: 65–70 (SEG 43, 488); Henry 2001: 765–71 (photograph and facsimile). Cf.
Eidinow and Taylor 2010, B4.)
[ - - ] ªÆØ åÆæØ . [˛]ºÆ PŒ åø K  [ ÅØ?]
[T ]EŁÆØ. !f c I غ )E P[Łø] N_ º[E åØ],
[& ]a æØ  N K ı ÆHØ KØ _ºÆ Æ [ . . c.8 . . ]
[c] c Kº<>ø Åb  _ [Ø Ææ][å]ø ºØÆ b [ . . c.8 . . ]
5 [± ?]_ ø , N b [c. . ._. c.8_ . . ] ŒÆd ÆFÆ ØH åÆæØ[E )E ].
[ ]æÆ   T [] _ [ÆÆ &] a [)æH ] j ŒÆÆŒøº[]ø vac.
_
[ææ]ø?[]. _ _

[ . . . ]tos to Tegeas, greetings. I am unable to buy wood in M[ende?]. So you dispatch some
to us immediately if you have a boat, buying seven talents if it is possibile [from X]. Let him
not provide you with any fewer at all, preferably [of . . . ]; but if not, [ . . . ] And if you do this,
you will be doing us a favour. Complete purchases within seven days or I shall put a stop (to
the arrangement). Farewell(?) (Henry’s translation).
34. Unpublished letter from Mende in Chalkidike, fourth century bc; Jordan 2000: 92 n.
3; cf. Eidinow and Taylor 2010, B2.
Attica
35. Letter (?) on ceramic sherd from the agora of Athens, mid-sixth century bc.
Thompson 1948: 160 (pl. 41. 2); Jeffery 1990: 135, 137, pl. 22, no. 1; Guarducci 1974: 320–1;
Lang 1976, B1; AVI no. 540. Cf. Pébarthe 2007: 81; Eidinow and Taylor 2010, A5.
[¨Æ ]F : ŒŁ : hı e ÐØ hÐØ A ŁæÆ Ð Œ  : æ (Æ)
Thamneus, put the saw under the threshold of the door of the garden.
The letter forms correspond to those in use at Megara; the writer must have been a
Megarian. The name is almost entirely restored by Lang, on the basis of the fact that in
the same pit two vases belonging to Thamneus (in Attic letters) were found.
36. Letter on ceramic sherd from the agora of Athens, early fifth century bc.
Lang 1976, B2; Oikonomides 1986, no. 9 (SEG 36, 124). Cf. Pébarthe 2007: 81–2; Eidinow
and Taylor 2010, A6.
—ÆE, ÐØ ƺ[ ŁØ] | ¼ºº ŒÆ<Ø> e Œº[Ø Ðæ]- | Æ ç æØ.
Lang’s restorations, exempli gratia.
Slave, bring to Phalanthos other new beds.
Oikonomides 1986, 51–2, no. 9 proposes a very different restoration: —ÆE, ÐØ ƺÆ[Ø ª]-
| º ŒÆ e ŒÆ[d hØ ]- | Æ ç æØ, ‘Slave, put on Phalias (a donkey) big baskets and leather
straps.’ If this changes the details of the message, it does not modify its overall structure.
37. Letter on an amphora sherd from the agora of Athens, second quarter of the fifth
century bc.
Lang 1976, B 7; AVI no. 496. Cf. Pébarthe 2007: 82; Eidinow and Taylor 2010, A7.
¯Pºd wŒ[] | › å |
æŒØ
Eumelis come as quickly as possible. Arkesimos.
The above is Lang’s understanding of the text. As Immerwahr (AVI no. 496) rightly points
out, however, ‘the problem is that the ancients do not sign their letters on the bottom’.
352 Appendix 1
38. Graffito on the inner part of handle and adjacent part of lip and body of a black-
figure skyphos from the agora of Athens, last quarter of the fifth century bc.
Lang 1976, B 9; BE 1977, no. 122 (SEG 26, 67); Jordan 1978: 92–4 (SEG 28, 41); Oikono-
mides 1986: 49, no. 6 (SEG 36, 121); AVI no. 272. Cf. Pébarthe 2007: 82; Eidinow and
Taylor 2010, A9.
! () | K º | ˆºÆŒØ | K ¼ı |  ( )
_
Sosineos sends to Glaukos in the city a bundle.
For Lang, this is a tag (‘Sosineos sent a bundle to Glaukos in town’); but K º here
implies that this is a letter (J. and L. Robert, BE 1977, no. 122, who assume that we have here
only the beginning of a letter). However, the sherd seems to have been broken before the
inscription: there probably was not much else. Hence the proposal of Immerwahr, AVI no.
272, to translate ‘Sosineos [sends] to Glaukos: a bundle into the city (sc. I have sent)’; but it is
necessary to supply information (the intended ‘I have sent’, or possibly also an order, ‘send’)
for this to work. As there is no further verb, the letter may have travelled with the bundle.
A further possibility has been aired by Jordan 1978, who reads the final line as a dative
(K Ð[Ø]), and takes this to mean ‘a scroll-like parcel’: ‘Sosineos sent a letter in form of a
scroll to _Glaukos in town’; this seems extremely unlikely (see discussion above, 44 n. 81).
39. Letter on lead sheet (4
7 cm), from Chaïdari near Daphni (Athens), fifth–fourth
century bc.
IG iii 3 App., pp. ii–iii; A. Wilhelm 1904a: 94–105 (photograph; cf. Syll3. 1259); Trapp 2003:
198–9. Cf. Jordan 2003: 32–3; Pébarthe 2007: 82–3; Eidinow and Taylor 2010, A1.
Outside: çæ N e ŒæÆ- |  eª åıæØŒ , | I Ð ÆØ b ˝ÆıÆØ | j ¨æÆıŒºBØ j
ŁıØHØ.
Inside:  Åæª | K غ E YŒØ | åÆæ ŒÆd ªØÆ  · | ŒÆd ÆPe oø
çÆ[Œ] [å ] | ªÆÆ, Y Ø º, | I  łÆØ j þÆ j ØçŁæÆ | ‰
Pº(Æ) ŒÆd c Øıæøa | ŒÆd ŒÆÆÆ : ıåe I ø.
Outside: Carry it into the pottery (district?). And give it to Nausias or Thrasykles or the
son.
Inside: Mnesiergos sends to the people at home greetings and good health; he said that
he too was like that. Send a covering, if you please, sheepskins or goatskins, the
cheapest possible and not shaped into cloaks, and shoe-soles; I will make a
return when I get the chance.
40. Letter (?) on lead, c.5 cm. long and 0.1 cm. thick, from the Pnix of Athens, 425–325 bc.
Raubitschek 1943: 10–11 (facsimile fig. 11); BE 1944, no. 90; Vinogradov 1998: 153 n. 4,
no. 16 (SEG 49, 325). Cf. Eidinow and Taylor 2010, A2. What follows is Raubitschek’s text:
¨
åÆæ(Ø) ŒÆd [ - - - ]
ˆ ÆŁøØ Ææ[ - - - ]
Å
æت ø[ _ ---]
5 ŒÆŁ Ø[ - - - ]
Ææ Ø[ - - - ]
ƪŒØ[ - - - ]
_ _ _ __
The nature of the text is not evident: the second word, åÆæØ , is typical of the opening of a
letter; on the other hand, the invocation ‘gods’, Ł, at the beginning is surprising. Eidinow
and Taylor 2010, 43 n. 62, speak of ‘interesting echo of a public document, if the reading
is correct’.
 Appendix 1 353
The possibilities are as follows:
1. J. and L. Robert, BE 1944, no. 90, proposed: Ł. åÆæ ŒÆd [ªØÆ  ] (as in the letter
of Mnesiergos). ‘Gods. Greetings and good health’ which they must have thought
could be followed by something along these lines: ‘For Gnathios, on the part of
[name] son of Arignotos [name] deposited.’
2. Alternatively, Vinogradov 1998: n. 4, no. 16 proposed, on the assumption of shorter
lines: ¨ª[ øØ] | åÆæ(Ø) ŒÆd | ˆ ÆŁøØ· ÆE[] | q
æت -, | ŒÆŁ Ø |
Iæ Ø- | Æ KŒØ[ - - ] (‘Greetings to Theognetos and Gnathios; the slave of
_ _ here, he deposited . . . ’). Here, while the beginning works well, it is
Arignotos was
difficult to see where the pais (‘slave’) of Arignotos comes from. If we stick to
Vinogradov’s idea of very short lines, something like Ææ|B
æª ø[ (‘Arignotos
came, he deposited . . . ’) might work better; however, this requires supposing that
the line end did not coincide with the word end, as seems to be the case in the rest of
this text.
The use of ŒÆ Ł at l. 5 shows that this concerns a commercial transaction; the tense
is interesting (epistolary past?). a ¼æ Æ (of which Ææ ØÆ possibly is a diminutive)
points to small tools, or alternatively, to sails or tackle of a ship.
41. Letter on a rolled up lead sheet (5
13.4 cm) from the agora in Athens, beginning
of fourth century bc.
Jordan 2000: 91–103 (with photographs and facsimile) (SEG 50, 276) cf. Harris 2004: 157–
70; Harvey 2007: 49–50; Eidinow and Taylor 2010, A3.
A ¸BØ {Ø} K غºØ ˛ ŒºE ŒÆd BØ Åæd ÅÆH æØØÐ
ÆPe I º   K HØ åƺŒøØ, Iººa æe e  Æ ÆPÐ KºŁÐ
ŒÆd K ıæŁÆØ Ø ºØ ÆPHØ.
Łæ øØ ªaæ ÆæÆÆØ  ı  ÅæHØ·
Æت   I ººıÆØ· ÆØ· æ źƌÇÆØ· Aºº A[º]º .
B Faint traces of letters on what would have been, once the sheet was rolled up, the outer
part, probably corresponding to the address.
Lesis sends to Xenokles and his mother by no means to overlook that he is perishing in
the foundry but to come to his masters and find something better for him. For I have
been handed over to a man thoroughly wicked. I am perishing from being whipped;
I am tied up; I am treated like dirt; more and more. (Trans. Jordan)
42. Private lead ‘letter’ from Athens (dimensions: c. 8.5
8 cm), present location
unknown, dated to before 370/369 bc.
Edited from a photograph by Jordan 2003: 23–39 (SEG 53, 256; Gauthier, BE 2004, no.
140); Sosin 2008: 105–8; cf. Eidinow and Taylor 2010, A4.
[—]Æø <˜>ØŒÆØæå K غº-
ø !Æıæø Æ Øøæ-
ÆŁÆØ ŒÆd ºŁÐ ŒÆ-
d ˝ØŒ æÆ e ˜ -
5 ø [] Iºçe ŒÆd
æ-
Ł Ø _ , ‰ Ææ Kb IØ-
_ Ð Ø ŒÆd K Ø ºØ,
Œ
ŒÆd ˆºÆıŒÅ ŒÆN `NÆ-
 øæ ŒÆd K Ø º-
10 Ø ŒÆd c æ æ
[ æ]ºŁB ÆØ c-
[ - -_ -_ _- c.17 - - - - - ]
354 Appendix 1
1: [—]Æø <˜>ØŒÆØæå Jordan; [—]Æø <˜>ØŒÆØæå<øØ> [vel sim.], or, less likely, —Æø Ø
˚``%"ˇ Sosin.
I, Pasion, son of <D>ikaiarchos, am sending by letter, to punish and prosecute Satyrion and
Nikostratos, Deinon’s brother, and Arethousios because they are wronging me and plotting
against me and Glauketes and Aiantodoros and they are plotting and that X not be paid
before . . .
The text above is problematic, since there is no addressee. Jordan had interpreted: ‘I, Pasion
son of Dikaiarchos, am sending a letter for Satyrion to punish and prosecute . . .’; but the
construction of K غºø + accusative and infinitive is unusual. Hence, Sosin’s proposal of
restoring an addressee in the dative in the first line, which would yield a translation as
follows:
I, Pasion, am sending by letter to <D>ikaiarchos, to punish and prosecute Satyrion and
Nikostratos, Deinon’s brother, and Arethousios because they are wronging me and plotting
against me and Glauketes and Aiantodoros and they are plotting and that X not be paid
before . . .
In Sosin’s translation, this is a letter without prescript, but with a sender and an addressee,
Pasion and Dikaiarchos. Alternatively, as Sosin himself suggests, one could assume that in l.
1 the writer had intended to write ‘—Æø Ø ˜ØŒÆØæå’ (the missing ˜ is almost easier to
understand this way); in this case, an unnamed sender (his name would have been on the
outside) is asking Pasion son of Dikaiarchos to deal with his affairs. The letter would have
been a kind of ‘safe-conduct’ for Pasion. The insistence on punishment on the one hand
and plotting on the other, in a letter written on lead, seems to me to speak for an
interpretation as a curse, or rather, in this specific case, as a prayer for justice (on epistolary
curses see above, ch. 2. 3). Although in curses the defigens usually avoids naming himself/
herself, this does not apply in the same measure to judicial prayers (but it is surprising to
find not only the name of the curser, but also his patronymic!) The first person and the use
of Øøæø find a parallel in an unpublished curse tablet from the sanctuary of Palaimon in
Athens, also dated to the fourth century bc (discussed in Versnel 2009: 311–12 and 317).

B. Documents whose letter status is uncertain

1. A graffito on a black glazed skyphos from Panskoje I, on the north shore of the Black
sea, dated c.300 bc. Two lines of writing were incised concentrically, one on the inner
and one on the outer part of the base; these are almost illegible. Better preserved is the
graffito on the inner part of the vase (remains of six lines are visible), inscribed in the
same hand.
Vinogradov 1997: 486 (SEG 42, 722); Jajlenko 2001: 223–32 (SEG 51, 984). Cf. Eidinow
and Taylor 2010, E16.
The document had been restored by Vinogradov 1997: 486 as an extract of the proxeny
decree from Olbia of 340–330 bc, with which the Olbiopolitai honoured the Athenians
Xanthippos son of Aristophon of the deme Erchia and Philopolis son of Philopolis of the
deme Deirades (I. E. Levi, Inscriptiones Olbiae (1917–1965) (Leningrad, 1968), no. 5).
According to Vinogradov, we should imagine the atmosphere of the symposion as the
context for the engraving of the extract (which concerns only Xanthippos):
[
]º[ØÆ] |
ªÆŁE å[Ø]· | [ O]º Ø ºEÆØ  [ø]- | [ŒÆ ] ˛Æ Ł ( øØ) HØ vacat | 5
[
æØ(çH )? ¯æ]åØE ? ŒÆd K [N]- | [æ ÅØ _AØ
_ ]E Æ[PF ?] - - - - - _ _
_ _
Exemption from taxation. To good fortune. The citizens of Olbia gave to Xanthippos the
(son of Aristophon? of] Erchia both in [peace for all] his [descendants?]
 Appendix 1 355
Jajlenko (2001: 223–32) has proposed to read in the signs of l. 4 numerals, making of this a
commercial letter:
[› E Æ K ]º [HØ E Ø] | IªÆŁE [åØ - - ] | [ O]º Ø ºEÆØ ¯"[ - - - ] |
[ - - ]`:˝ ƒ —[ - - ] | 5 [ - - - - - ] X! ŒÆØ ENE [ - - ] | [ - - - - - ]![ - - - -_]
[X to Z] sends, good fortune! [ - - ] The inhabitants of Olbia, 51 vases for a price of 5
drachmae and a quarter . . . stater
Chaniotis in SEG 51, 984 remarks that in a letter one would expect the present K غºØ
and not K ]º (this is, however, debatable, as it could be an epistolary past: compare the
letter of Mnesiergos, no. 35). More importantly, he notes that IªÆŁc [åØ does not make
much sense in a letter or an invoice.
2. A graffito on a black-glazed vase from Nikonion, dated to the late fifth–early fourth
century bc.
The first editor, V. Alekseyev (VDI 2004.3 70–3 no. 4, in Russian; cf. SEG 54, 691) thought
that these lines reflected a contest within a symposion. The text has been republished by
S. Yu. Saprykin (in ¯PåÆæØæØ - - Domanskogo (St. Petersburg, 2007), 68–71, in
Russian; cf. SEG 57, 717), who prefers to see here personal notes about the area by a sea-
captain or merchant from a Doric polis. A. Avram, BE 2008, no. 397 and 397bis, points out
that this text may be a private letter or an oracular enquiry; if the latter, it might have been
addressed to Achilles in Leuke. Below, Avram’s text:
:
 HæÆŒºÆ [ - - - - - - - ]
...

æd å æ- [ - - - - - - - - - ]

ÆY æ Ø d Œ¼º[ºøØ - - - - - - ]

3. Lead tablet from Pech-Maho, c.450–440 bc.

SEG 38, 1036; SEG 41, 891; Decourt 2004, no. 135.
Although categorized as letter by Dana 2007: 68, this document looks rather more like the
memorandum of a commercial transaction, and belongs with commercial agreements
rather than letters.
4. and 5. Two texts on lead from Lattes (near Montpellier) and Ruscino (near Perpignan).
M. Bats (2011: 203–4) mentions two texts on lead (I should like to record my thanks to M.
Dana for alerting me to the existence of these documents). One is an extremely damaged
document from Lattes, dated to c.430 bc by the archaeological context. It had been folded in
three, and is perforated at the base as if by nails; one word seems to be legible, ή (cf. Py
2009: 304–5, with photograph). The other comes from Ruscino, probably dates to the
fourth century bc, and is to be published by J. de Hoz. As long as they remain unpublished,
the status of these documents (letters? memorandums? see Decourt, BE 2012 no. 512)
remains uncertain.

C. A unique later document from Acra on the Cimmerian Bosporus

Official lead letter of Hellenistic or imperial times from Acra: lead plaque (7.5
6.5 cm)
folded in four, found in the sea; seven lines of text are preserved, from the upper left part of
the plaque, dated by the editors on the basis of the writing to between the second half of the
third century bc and the second century ad.
Saprykin and Fedoseev 2008: 72–9 (with illegible photograph and English summary at
79–80), also published (in English) in Saprykin and Fedoseev 2010b: 427–34 (SEG 58, 775).
356 Appendix 1
B æı › K d º[ø . . . ] | øØ åÆæØ · ƒæa  ıª[æÆ? - - ] | F b ‰<> øÅæÆ
 [ŒEŁÆØ ? –] |_ a  e æ[ ØF ? – – ] | –] ÆØ K  ±æı [ – – | – ] Iæc ŒÆd Y
O [æø
fi ? – – ] | [ – – ] ÆPc K o[ÆØ ? – –.
The letter, much later than any other lead letter, is for the time being a unicum. It is sent by
a magistrate in charge of the city to a (Phili)ppos (or Leoni)ppos.
Botrys may have been a royal official, appointed by one of the Bosporan kings, or a
prefect (see the comments of A. Avram, BE 2009, no. 384). The very fragmentary remaining
lines of the letter have been understood by the editors as concerning flooding risks; but as
Avram points out, ll. 2–4 can also be (better) restored as ƒæa Æ ı [æøØ K] | ı ø fi
!BæØ ÆP[ŒææØ] | , while the two final lines escape interpretation (at l. 5 Avram
proposes & <Œ> Ææı). The interesting point from our perspective is that if Botrys sends a
(lead) letter, although he is based in Acra and the topic discussed concerns the internal
administration of Acra, this means that for internal communications a letter was indeed the
expected format; the choice of the lead remains slightly surprising.
 APPENDIX 2

Ancient Traditions on the Invention of Writing

In what follows I have collected some of the main (and less accessible) Greek sources on the
invention of writing. For more detail, I must refer to the formidable collection of data in
Schneider 2004, as well as to Jeffery 1967. I have preferred to present the scholia as a whole,
rather than to extract the ancient authorities from them. Authorities are underlined;
inventors are in bold.

1. Scholia in Dionysii Thracis artem grammaticam, ed. Hilgard (Grammatici Graeci, i. 3

(Leipzig, 1901)).1
Commentarius sub auctore Melampode vel Diomede, schol. Dion. Thr. 32. 9–13:

Øb  çÆØ  f åÆæÆŒBæÆ H  Øå ø  f Ææ’ E  e  Eæ F K ç ØŒ  纺ø
fi
ª ªæÆ ı ŒÆÆ çŁBÆØ  E IŁæ Ø, Øe ŒÆd ç Ø Œ ØÆ ºª ÆØ a ªæÆÆ· ƒ ,
‹Ø  Ø Œø Kd oæ Ø· ƒ , ‹Ø › ÆØƪøªe  F åغºø ›  EØ Kç Fæ  ÆP.
Some say that the shapes of the elements which we use were transmitted to mankind by
Hermes, written on a palm-leaf, which is why the letters are called phoinikeia; others
however say that they are a discovery of the Phoenicians; yet others, that the paedagogue
of Achilles, Phoenix, invented them.
Scholia Vaticana, schol. Dion. Thr. 182. 15–183. 23:
— æd b B H ªæÆø æ ø ØÆçæø ƒ ƒ æØŒ d ƒæÅÆ· ƒ b ªaæ
—æ ÅŁÆ ºª ıØ  ø æ , ¼ºº Ø b  ØŒÆ e  F åغºø ÆØƪøª,
¼ºº Ø b e غØ  ˚ , ¼ºº Ø b c ŁÅA, ¼ºº Ø b K PæÆ F KææEçŁÆØ  E
IŁæ Ø æe Tçº ØÆ. oæÅÆØ b På ç’ e – ÆÆ· o æ  ªaæ K  ŁÅÆ [20]
a ÆÆ ŒÆd a Ø ºA· æd ªaæ ÆFÆ K Ø ÅŁBÆØ,  E IØ å Ø KåæH , x  N
MŁºÅÆ å ØBÆØ, K ı Œ ŒÆd Æ EÆ, N b Ł j ç, ºØ › ø a I  ØåÆ  a
Æ ØH, ‰ F ƒ   øÆE Ø· ŒÆd a Ø ºA , K z ªŒ ØÆØ. ¼ºº Ø b ºª ıØ, ‹Ø ŒÆd a
ÆŒæa  º ıÆE  K  ŁÅÆ, H æÆåø e æ æ   a ÆŒæA [25] æ ØŁ Å
¼øŁ  IÆ ºÅæ ø <ÆPH> c Ø· K ı ªaæ Id  F ø ŒÆd ¼ø<Ł > ÆŒæ,
ŒÆd Id  F Å ŒÆd ¼øŁ  ÆŒæ.
Øb b ç Ø Œ ØÆ KŒº Æ a ªæÆÆ, < ƒ  d
çø Œ ØÆ,> Ææa e B çøB NŒÆ r ÆØ a ªæÆÆ.
(183.)
H  Øå ø æ c ¼ºº Ø  ŒÆd  0Eç æ  K  ıæø fi ˚  çÆ · ƒ b På
æ , B b  Ø Œø æ ø æe A ØŒ æ  ª ª BŁÆØ, ‰ ŒÆd  Hæ   K
ÆE ƒ æ ÆØ ŒÆd æØ ºÅ ƒ æ E· çÆd ªaæ ‹Ø  ØŒ  b yæ  a [5]  Øå EÆ,
˚  b XªÆª  ÆPa N c  EººÆ. —ıŁøæ  b [‰] K fiH æd  Øå ø ŒÆd  ººØ
› ˜ºØ  K fiH æd åæø æe ˚ ı ˜ÆÆe  ÆŒ  ÆØ ÆP çÆØ· K ØÆæıæ FØ
  Ø ŒÆd ƒ غÅØÆŒ d ıªªæÆç E Æ Ææ  ŒÆd ˜Ø Ø  ŒÆd  EŒÆÆE , R ŒÆd
 ººøæ  K N H ŒÆƺªø fi ÆæÆ Ł ÆØ. [10] Ø Ø b  ıÆE  æ c ºª ıØ e
Å   ŒÆd ! æ Å ŒÆ’  OæçÆ ª   · ØŒº Å b › ŁÅÆE  `Nªı  Ø c
oæ Ø IÆ ŁÅØ· ˜ ØÅ b K ˚æfiÅ çÅd æ ŁBÆØ ÆP· `Nåº  b —æ ÅŁÆ

1
For the various collections (commentaries of Diomedes and Melampous, Scholia Vaticana, Scholia
Marciana, Scholia Londinensia) forming these scholia, and for their relationship see Lallot 1998;
Schneider 2004: 120. On the passage of Dionysius Scytobrachion (no. 3 below) see Corcella 1986; on
Ptolemy Pindarion and Pronapides, the teacher of Homer responsible for the introduction of the letters
at Athens, see Montanari 1995: 41–58, with further references.
358 Appendix 2
çÅd æÅŒÆØ K fiH ›øø fi æÆØ, !Å å æ  b K  ıæø fi  Oæ  Æ ŒÆd
¯PæØ Å e [15] —ƺÆÅ çÅd æÅŒÆØ, ÆÆ b  EæB, ¼ºº Ø b ¼ºº .
ØŁÆe b ŒÆa Æ   æ a ª ª BŁÆØ, Øe ŒÆd ¼ºº Ø Ææ’ ¼ºº Ø Nd åÆæÆŒBæ 
H  Øå ø· x  b ıd åæ ŁÆ, Nd " øØŒ , N ªŒÆ  æå  ı Ææ’ ŁÅÆ Ø
łçØÆ,  f ªæÆÆØ, Xª ı  f ØÆŒº ı, ÆØ  Ø c " øØŒc ªæÆÆØŒ,
Xª ı a [20] ªæÆÆ. ˜ØÆ b a  Øå EÆ ªæç ŁÆØ ‰ ªæç   F —æ Æ Å ›
ŁÅÆE · H ªaæ IæåÆ ø ƒ b  ØæÅe ªæÆç , ƒ b ºØŁÅ, ƒ b  ıæ çÅ,
ƒ b ŒØ Å [ æd z K  E B å   ÆŁ E].
The historians give different accounts of the invention of the letters: some say that Prometheus
discovered them, others Phoenix the paedagogue of Achilles, others Cadmus the Milesian,
others Athena, and others that they fell from the sky for men’s use. But they were not
discovered, all of them, by one person: for the aspirated and the doubles were discovered
later, and before they thought of these, they used the corresponding letters. Thus if they wanted
to express chi, they used k and the aspirate, if theta or phi, again equally the corresponding
signs with aspirates, as do the Romans; as for the doubles, they used those out of which they are
formed. Others, however, say that also the longs were found at the end, and that initially the
short vowels accomplished their purpose having a long added on top. For they used to write
instead of ø, with a macron above, and instead of Å, , also with a macron above.
And some called the letters phoinikeia, <as if phonikeia>, because the letters (grammata)
are the image of the voice (phone).2
But others, among whom Ephorus in his second book [FGrH 70 F 105], say that Cadmus was
the inventor of the letters; for others he was not the inventor, but the transmitter of the
Phoenicians’ invention to us, as both Herodotus in his Histories (5.58) and Aristotle [fr. 501
Rose] recount; for they say that the Phoenicians invented the letters, but Cadmus brought them
to Greece. For their part, Pythodorus in his book On Letters and Phillis of Delos in his book On
Times say that Danaus brought them over before Cadmus; their statements are supported by the
Milesian writers Anaximander [FGrH 9 F 3], Dionysius [FGrH 687 F 1] and Hecataeus [FGrH 1
F 20], who are also mentioned in this respect by Apollodorus in his Catalogue of the Ships [FGrH
244 F 165]. Other say that Musaeus, the son of Metion and Sterope, who lived at the time of
Orpheus, was their inventor; Anticlides the Athenian [FGrH 140 F 11a, cf. 11b] attributes their
invention to the Egyptians; Dosiadas [FGrH/BNJ 458 F 6] says that they were invented in Crete;
Aeschylus affirms in his homonymous play [v. 460] that Prometheus invented them, while
Stesichorus in the second book of his Oresteia [fr. 213 PMG] and Euripides [cf. fr. 578 Kannicht]
say that the inventor was Palamedes; Mnaseas [fr. 52 Cappelletto] proposes Hermes, and others
others. One must believe that there were inventors in each given place, since the shapes of the
letters are also different in the various localities; those which we use now are the Ionian, as a result
of Archinus’ introducing in Athens a decree, that the grammatistai, that is the teachers, teach the
Ionian grammatike, that is the letters. Pronapides the Athenian arranged that the letters be
written as we write now; for among the ancients, some wrote in coiled lines, others arranging
them as bricks in rectangles, others boustrophedon, other in columns.
Scholia Vaticana, schol. Dion. Thr. 184. 20–186. 4 Hilgard:
 Ø Œ ØÆ b a ªæÆÆ Kºª  , u çÅØ  0Eç æ  › ˚ıÆE  ŒÆd  Hæ  , K d
 ØŒ  yæ  ÆP· ¯PçæØ  , ‹Ø  ºø fi e æ æ  Kªæç  , ‹ KØ åæH Ø
ç ØØŒ F·  ¯  ø f <b> ŒÆd Ææ , K Øc K º Ø ç ØØŒ Ø Kªæç  · X, ‹ æ
Œæ E KØ N E, ‹Ø ç Ø  ÆØ  ’ ÆPH ›  F Xª ı ºÆ æ ÆØ· [25] @æø b
ŒÆd   ŒæÅ ›  OºŁØ  I e  Ø ŒÅ B ŒÆ ø  ŁıªÆæ·  ººØ  b ›  F
æåØ ı, K Øc ƒ I ªæÆç Ø I e ç ØŒ  º  r å  ŒÆd  ’ ÆP F ªæÆç · ˜ FæØ b
› !Ø  › ƒ æØŒe K OªfiÅ <H> ÆŒ  ØŒH I e  ØŒ   F åغºø æ ç F·

2
For the connection between sound and sign see above, 14–15, 21, 29, 65, 67 n. 27. Lucan, Phars. 3.
220–1, combines Phoenicians and voice: Phoenices primi, famae si creditur, ausi j mansuram rudibus
vocem signare figuris (‘The Phoenicians were the first, if rumor is to be believed, who dared to mark
with rough signs the voice, so as to make it last’).
 Appendix 2 359
ºÆæ  b ›   Ø  I e  ØŒ   F (185.) —æ  ı ŒÆd ¯Pæ Å, æ  ÆPa K
˚æfiÅ, n I Œ Ø   ÆÆŁı çŁ Æ.
æ Æd b H º Ø H [H] åÆæÆŒæø,
 ıØ H OŒ, x  H  ÆŒæH ŒÆd H æØH Ø ºH ŒÆd <H> æØH Æø,
[5] ź Ø !Øø Å b › ˚ E  H  ÆŒæH ŒÆd  F  ŒÆd  F ł, —ƺÆÅ b H
Æø ŒÆd  F Ç, X, u çÆ Ø ,  ¯ åÆæ  › !ıæÆŒ Ø .
# O Ø <b> c H ªæÆø oæ Ø !Øçø fi j —ƺÆfiÅ j  ØŒØ j —æ ÅŁ E
Kç  ıØ, j K  Ø ŒfiÅ I e PæÆ F øŒÆØ çÆ , j Ææ’ `Nªı  Ø æÅŒÆØ ¨Ł,
n  EæB æÅ  ıØ, [10] < PŒ OæŁH ºª ıØ>· ŒÆd ªaæ çØ  ŒÆ KÅØ æªÅ e
¼Łæø , KåÆæ Æ ÆPfiH  ØÆÅ K ØÅ ØÅÆ, u  åÆŁÆØ ÆFÆ a  Øå EÆ.
 ŒıÆØ b ÆFÆ ŒÆd æe H " ºØÆŒH ZÆ, ‰ Bº  KŒ H B ºº æ ç ı· çÅd ªaæ ›
ØÅc “ªæłÆ K ÆŒØ [15] ıŒfiH Łı çŁæÆ ºº·” › b B ºº æ çÅ æe  F

æøœŒ F º ı q, ŒÆd ªaæ   q  F ˆºÆŒ ı  F K fiH
æøœŒfiH ºø
fi Z . Iººa
ŒÆd K —ıŁ E q ºÅ K ت ªæÆ  “çØæø ’ IŁÅŒ ºÆg I e
ź  ø·” ŒÆd
K
çø fi B
fi ø
fi [20] “ %Œ Æ F ŁıªÅæ ŒÆd
ÅŁ  Nd ˝å ØÆ | ŒæÅ·
ź ÆØ ªæ 
’ TÆÆ· | çÆØ b æ åø º ıæ, ŁÅ EØ ’ ª Å· | ŁBŒ   — æºÆ,
ıƒe  ¯ ıƺ ı.”
 a b e K d ˜ ıŒÆº ø  ŒÆÆŒºıe P d H æغ ØçŁø [25]  Eººø
KçºÆ  ÆPH c Å, ºc H — ºÆªH H Iç’  Eºº  N Æææ ı
ºÆÅŁø, R ŒÆd › ØÅc  ı ŒÆº E, çŒø “ŒÆd ¸º ª  ŒÆd ˚ÆŒø  E 
— ºÆª ”· Ææ’ z ÆŁ  æH Ø  ØŒ  N  ‚ººÅÆ XªÆª , ª    Æææø
Z  ŒÆd ı å E K æ Æ Ø   Ø· ‹Ł  ŒÆd  Ø Œ ØÆ (186.) ŒÅØŒH O Ç ÆØ. ƒ
b ç Ø Œ ØÆ ºª ıØ ÆP, ƒ  d çø Œ ØÆ, K Øc çøB KªªæÆ ı ı Ø NŒ 
N ,  Æ ºB ª ª ı Æ  F ø N c Ø  çŁ ªª  ŒÆa c H B ØøH غ Œ ,
u æ IªŒÅ IªŒ Å.
They called the letters phoinikeia, ‘Phoenician’, as Ephorus of Cumae [70 F 105] and Herod-
otus [Hdt. 5. 58] say, because Phoenicians invented them; Euphronius however says, because
earlier they used to write with a red ochre called miltos [red lead: Plin NH 33, 115], which has a
reddish colour; and Eteoneus and Menander [BNJ 783 F 5], because they used to write on palm
leaves; or, a better explanation, because the mind is reddened by them, that is, is brightened;
but Andron [FGrH 10 F 9/ Bollansée, FGrHContIV 1055 F 5–6] and Menecrates from
Olynthus affirm that the name derived from that of Phoenice, the daughter of Actaeon;
Apollonius the son of Archibius [= Apollonius sophista], because the copying clerks had a
wooden instrument made of palm-wood and wrote with it; Duris the Samian historiographer
[FGrH 76 F 6] in the eight book of his Macedonica suggests that the name derives from
Phoenix the tutor of Achilles; Alexander the Rhodian from Phoenix the son of Pronapos and
Europe, who would have invented them in Crete; Rhadamanthus killed him out of envy.
Inventors of the remaining eight characters, that is the two long vowels and the three doubles
and the three aspirated, were for the two longs, the xi, and the psi, Simonides of Ceos, and for
the aspirated and the zeta, Palamedes or, as some say, Epicharmus the Syracusan.
Those who connect the invention of the letters to Sisyphus or Palamedes or Phoenix or
Prometheus, or who say that they fell from the sky in Phoenicia, or that Thoth, whom they
interpret as Hermes, invented them in Egypt, <are not correct>; for nature, when she created
man, also gave him an aptitude such as enabled the creation of these letters. And it appears that
they existed already before the Trojan war, as is clear from the stories concerning Bellerophontes;
for the poet says, ‘having written in a folded tablet many soul-destroying things’ [Il. 6. 169]; but
Bellerophontes lived before the Trojan war, as he was the grandfather of Glaucus who participated
in the Trojan war. Moreover, there was in Pytho a lebes inscribed with ‘Amphitryon dedicated me
as booty from the Teleboi’ [Hdt. 5. 59]; and in the island Thasos: ‘Daughter of Ocean and Tethys,
I am the source Nucheia; the Teleboi gave me this name; I pour water for brides-to-be, and
health for the mortals; Pterelas son of Enyalius dedicated me’ [Anth. Pal. 9. 684].
After the flood at the time of Deucalion, none of those Hellenes that survived preserved a
memory of them, but for the Pelasgians who had moved from Greece to the barbarians, those
that the poet calls ‘divine’, saying ‘and the Leleges and the Caucones and the divine Pelasgians’
(Il. 10. 429); and the Phoenicians, who lived close to the barbarians and dedicated themselves
assiduously to commerce, having learnt the letters from them first brought them to
360 Appendix 2
Greece; hence they are named phoinikeia. But others consider them to have been named
phoinikeia as if phonikeia, implying a change of the omega to the diphthong oi according to
the Boeotian dialect, as with angkone and angkoine, because they are potential images of the
written voice.
Scholia Vaticana, schol. Dion. Thr. 190. 19–35 (cf. 191. 19–192. 7) Hilgard:
@ººø N e ÆP.—
H  Øå ø › ˚  æ  KØ, u çÅØ  0Eç æ  ŒÆd
æØ ºÅ· ¼ºº Ø b ºª ıØ, ‹Ø  Ø Œø Nd æÆÆ, ˚  b ÆFÆ
Ø æŁ ı  N c  EººÆ· —ıŁøæ   çÅØ ‹Ø ŒÆd æe ˚ ı › ˜ÆÆe KŒ EŁ ,
X Ø I e  Ø ŒÅ, ÆPa   ŒØ · Ø Ø b  ıÆE  æ c ºª ıØ H  Øå ø e
Å   ŒÆd ! æ Å, ª    ŒÆa  f ŒÆØæ f [25]  F  Oæçø· ØŒº Å b
 E `Nªı  Ø c oæ Ø IÆ ŁÅØ· `Nåº  b —æ ÅŁÆ çÅd æÅŒÆØ ÆP·
˜ ØÅ b K ˚æfiÅ ºª Ø æ ŁBÆØ ÆP· !Å å æ  b —ƺÆÅ æ c ÆPH
Ø EÆØ, fiz ıçø E ŒÆd ¯PæØ Å, ÆÆ b < EæB. ØŁÆe b> ŒÆa Æ  
æ a ª ª BŁÆØ [30] H  Øå ø· <Øe ŒÆd ¼ºº Ø Ææ’ ¼ºº Ø Nd åÆæÆŒBæ  H
 Øå ø·> x  b  E åæ ŁÆ F, Nd " øØŒ , Øa e IæåÆØ æÆ r ÆØ c " Æ H
¼ººø ØƺŒø, ŒÆd ÆPc æÅ H ¼ººø  f  EººÅØŒ f æ E  ı. ˜ØÆ b
a  Øå EÆ ªæç ŁÆØ ‰ [35] ªæç   F —æ Æ Å › ŁÅÆE ·
Differently, on the same subject. Cadmus is the discoverer of the letters, as Ephorus and
Aristotle say; others however say that they are the discovery of the Phoenicians, and that
Cadmus brought them over to Greece; Pythodorus says that Danaus brought them here
from there, that is from Phoenicia, actually before Cadmus; some affirm that Musaeus, the
son of Metion and Sterope, is the discoverer of the elements, a contemporary of Orpheus;
Anticlides traces their invention back to the Egyptians; Aeschylus says that Prometheus
invented them; Dosiadas says that they were discovered in Crete; Stesichorus makes of
Palamedes their inventor, and Euripides agrees with him, while Mnaseas suggests Hermes.
It must be believed that there were discoverers of the letters in each place; for the shapes of
the elements are different in different places. Those which we use now are Ionian, because
the Ionian is the most ancient of the other dialects, and she first of all found the Greek
forms. But Pronapides the Athenian arranged that the elements be written, as we write now.
Scholia Vaticana, schol. Dion. Thr. 191. 29–192. 11 Hilgard:
"   b ‹Ø Æææø  NØ æÆÆ a Ø& 0 ªæÆÆ ÆFÆ, Æ  ª  Ø Œ º   æ
ı, !Øø  ı b  F ˚ ı oæ Æ a  ÆŒæ, e Å ŒÆd e ø, ŒÆd a  Ø ºA, e  ŒÆd e
ł, —ƺÆ ı b a æ Æ ÆÆ ŒÆd e Ç.
˚ƺ FÆØ b a  Øå EÆ ç Ø Œ ØÆ, K d › ˚   EØ J N  ‚ººÅÆ ÆFÆ
  ªŒ · j ‰ çø Œ Ø ØÆ ZÆ, X Ø çøB [35] KªªæÆ ı ı Ø[ ŒÆd] NŒ ,
 F ø  ÆºÅŁ  N c Ø  çŁ ªª  ŒÆa c B ØøH غ Œ , ‰ e IªŒÅ
IªŒ Å· c ªaæ IªŒºÅ, lØ KŒ  F IªŒH  IªŒÅ ºª ÆØ, IªŒ Å ƒ B Øø d [192]
ºª ıØ· oø s ŒÆd çø Œ ØÆ ç Ø Œ ØÆ, ŒÆa e r   e Ææªøª  e º ª  
ŒÅØŒe åÅÆØÇ Æ· j ç Ø Œ ØÆ, K d Øa  º ı, ‹ æ Kd åæHÆ ç ØØŒ F, æ æ 
Kªæç  j ‹Ø æ æ  K º Ø ç ØØŒ Ø Kªæç  · j ‹Ø K ıæ<æ> Ø ŒÅæfiH
Kı F · X, [5] ‹ æ Kd Œæ E , ‹Ø ç Ø  ÆØ  ’ ÆPH ›  F Xª ı ºÆ æ ÆØ.

Øb b ºª ıØ ŒÆa ƒ æ Æ, ‹Ø I e  Ø ŒÅ B ŒÆ ø  ŁıªÆæ· ¼ºº Ø I e
 ØŒ  < F> —æ  ı ŒÆd ¯Pæ Å. ºª ÆØ b  Øå E  j Ææa e  åø, e <K
 Ø> æ  ÆØ, j I e  F  å ı  F ÅÆ    c Ø, j I e ! å ı Ø, [10]
ŁÅÆ ı ªÅª  F, æ  ÆP, ‰ —ØÆæ ø çÅ · åÅÆ Ç ÆØ b Ææa e  Eå 
 Øå E .
It should be known that the following sixteen letters, Æ  ª  Ø Œ º   æ   ı, are the
invention of barbarians, the two longs, Å and ø, as well as the two doubles,  and ł, are the
discovery of Simonides of Ceos, while the three aspirates and the Ç of Palamedes. The
elements are called phoinikeia, because Cadmus, a Phoenician, brought them to Greece; or
because they are ‘phonic’, that is power and image of the written sound, with the ø having
 Appendix 2 361
changed to the diphthong Ø according to the Boeotian dialect, as ankone ankoine; for the
Boeotian call the ankale [‘bent arm’], which is the ankone, from ankon, the bend of the
elbow, ankoine. So then Phoenician from phonic, formed according to the derived figure
called possessive; or phoinikeia, because earlier, people wrote with miltos, which is a reddish
colour, or because earlier, people wrote on palm-leaves; or because they were impressed in
reddish wax; or, which seems the best argument, because through them the mind is
illuminated, that is, clarified. But some affirm that (they are called phoinikeia) according
to the story that they took the name from Phoenice the daughter of Actaeon; others, from
Phoenix the son of Pronapos and Europe. The stoichos [‘element’] takes its name either
from ‘to march’, when I advance in line, or from the element that indicates the rank, or from
a certain Stoichus, an Athenian born of the earth, who found them, as Pindarion [=
Ptolemy Pindarion, cf. RE Suppl. 9, Ptolemaios 78a] says; and from stoichos is formed
stoicheion.
Scholia Marciana, schol. Dion. Thr. 320. 17–30 Hilgard:
ˇh b ªæÆÆ r å  ƒ  ‚ººÅ  K IæåB, Iººa Øa  ØØŒ ø ªæÆø ªæÆç  a
ÆıH· ŒÆd ªaæ ÆPa a  Ø Œ ØÆ,  E OÆØ ‰ ŒÆd a  EæÆœŒ, ŒÆa  ÅØ H
 EæÆ ø [20]  E  ØØ æŁÅ. —ƺÆÅ ’ o æ  KºŁ, Iæ   I e  F ¼ºçÆ,
 ŒÆb Æ  E  ‚ººÅØ yæ  Øå EÆ, Æ  ª  Ø Œ º   æ   ı· æ ŁÅŒ b ÆP E
˚  › غØ  ªæÆÆ æ Æ, Ł ç å, Øe ŒÆd ººfiH fiH åæø fi  E  ŒÆ Æ KåæH ·
‹Ł  ƒ IæåÆE Ø c å   e ł c łÆº Æ ƺ Æ ªæÆç  ŒÆd º ª , Iººa ŒÆd ººa
ÞÆÆ ¼ººø K ç ı ŒÆd ªæÆç .  a [25] ÆFÆ !Øø Å › ˚ E  æg
æ ŁÅŒ  , Å ŒÆd ø,  ¯ åÆæ  b › !ıæÆŒ Ø  æ Æ, Ç  ł, ŒÆd oø K ºÅæŁÅÆ
a NŒ ØÆæÆ. # OŁ  ØÅ   › # OÅæ  e IæØŁe H NŒ Ø Øº ø H Ææ’
 EæÆ Ø, N e H NŒ Ø  Øå ø ı øŁÆ IæØŁ, ŒÆd ÆPe c N Æ ÅØ
B " ºØ  H NŒ Ø æø ÞÆłøfi ØH Øa H NŒ Ø æø  Øå ø K Å ·
oø b ŒÆd c  O ØÆ.
Nor did the Greeks possess the letters from the beginning, but they wrote their own through
the Phoenician letters; for even the Phoenician ones in their names, as the Hebrew letters,
were discovered by the Phoenicians in imitation of the Jews. Palamedes, later, beginning
with the Æ discovered for the Greeks sixteen stoicheia only, Æ  ª  Ø Œ º   æ   ı;
Cadmus of Miletus added to them three letters, ¨, , (, so that for a long time only these
nineteen signs were used; hence the ancients, not having the letter ), used to write and say
ƺ Æ for łÆº Æ; and they also pronounced and wrote in a different way many other
words. Afterwards, Simonides of Ceos found and added two letters, ˙ and %, and
Epicharmus of Syracuse three, Z, ˛, ), so that the number of 24 was reached. Hence
Homer, imitating the number of the twenty-two books of the Jews, a number formed on the
image of the twenty-two letters, himself oganized his own poem of the Iliad in twenty-four
rhapsodies, because of the twenty-four letters, and similarly for the Odyssey.

2. Ephorus of Cumae, FGrH 70 F 105a and b (v. supra, schol. Dion. Thr. 183, 1 and 184,
20 Hilgard); FGrH 70 F 105 c (Clem. Alex. Strom. 1. 75. 1):
˚  b  EØ q › H ªæÆø ῞¯ººÅØ æ , u çÅØ ῎¯ç æ , ‹Ł  ŒÆd
 ØØŒØÆ a ªæÆÆ ῾˙æ   Œ ŒºBŁÆØ ªæç Ø. ƒ b  ØŒÆ ŒÆd !æ ı ªæÆÆ
K Ø BÆØ æ ı ºª ıØ.
Cadmus, who invented the letters for the Greeks, was Phoenician, as Ephorus says, and for
this reason Herodotus writes that the letters were named Phoinikeia. Others say that the
Phoenicians or Assyrians first thought of the letters.

3. Dionysius Scytobrachion, FGrH/BNJ 32 F 8 (D. S. 3. 67. 1):

Åd  ı Ææ’  ‚ººÅØ æH  æ c ª ŁÆØ ¸   ÞıŁH ŒÆd º ı, Ø b ˚ ı
Œ  Æ  KŒ  Ø ŒÅ a ŒÆº  Æ ªæÆÆ æH  N c  EººÅØŒc  ÆŁ EÆØ
362 Appendix 2
غ Œ , ŒÆd a æ Ū æ Æ Œø fi ÆØ ŒÆd  f åÆæÆŒBæÆ ØÆı HÆØ. Œ Øfi B b
s a ªæÆÆ  ØØŒØÆ ŒºÅŁBÆØ Øa e Ææa  f  ‚ººÅÆ KŒ  Ø Œø    åŁBÆØ,
N fi Æ b H — ºÆªH æø åæÅÆø  E  Æ Ł EØ (20) åÆæÆŒBæØ — ºÆªØŒa
æ ƪ æ ıŁBÆØ. [(2) e b ¸   K d ØÅØŒfiB ŒÆd  ºø fi  fi Æ ŁÆıÆŁÆ ÆŁÅa å E
ºº  . . . (4) æd b  Oæçø  F æ  ı ÆŁÅ F <a> ŒÆa æ  Iƪæł  , ‹Æ a
æ Ø ÆP F Ø  ø .] e ’ s ¸   çÆd  E — ºÆªØŒ E ªæÆØ ıÆ  
a  F æ ı ˜Ø  ı æ Ø ŒÆd a ¼ººÆ ıŁ º ª Æ I ºØ E K  E  ÆØ.
(5) › ø b   Ø åæÆŁÆØ  E — ºÆªØŒ E ªæÆØ e  OæçÆ ŒÆd —æ Æ Å
e  ˇæ ı ،ƺ , PçıB ª ª Æ  º Ø· æe b   Ø ¨ı Å e
¨ı  ı  F ¸Æ   , ŒÆa c ºØŒ Æ ª ª Æ c  Oæçø, <n> ºÆÅŁBÆØ
ŒÆa ºº f  ı B NŒ ıÅ, ŒÆd ÆæÆƺ E B ¸ØÅ N c æe  æÆ
åæÆ [B NŒ ıÅ] ,ø TŒ Æ F, Ł ÆŁÆØ b ŒÆd c ˝FÆ, K fi w ıŁ º ª FØ ƒ
KªåæØ Ø [IæåÆE Ø] æÆçBÆØ e ˜Øı , ŒÆd a ŒÆa æ   F Ł F   ı æ Ø
ÆŁÆ Ææa H ˝ıÆø ıÆŁÆØ c æıª Æ O ÆÇ Å ÅØ, IæåÆØŒ E fiB
 ØƺŒø fi ŒÆd  E ªæÆØ åæÅ  .
He says then that among the Greeks Linus was the first to discover the rhythms and song,
and when Cadmus brought from Phoenicia the letters, as they are called, Linus was again
the first to transfer them into the Greek language, to give a name to each character, and to
fix its shape. Now the letters, as a group, are called ‘Phoenician’ because they were brought
to the Greeks from the Phoenicians, but as single letters the Pelasgians were the first to
make use of the transferred characters and so they were called ‘Pelasgic.’ [2. Linus also, who
was admired because of his poetry and singing, had many pupils . . . 4 About Orpheus, the
third pupil, we shall give a detailed account when we come to treat of his deeds.] Now Linus,
they say, composed an account in Pelasgic letters of the deeds of the first Dionysus and of
the other mythical legends and left them among his memoirs. 5 And in the same manner
use was made of these Pelasgic letters by Orpheus and Pronapides, who was the teacher of
Homer and a gifted writer of songs; and also by Thymoetes, the son of Thymoetes, the son
of Laomedon, who lived at the same time as Orpheus, wandered over many regions of the
inhabited world, and advanced into the western part of Africa as far as the ocean. He also
visited Nysa, where the ancient natives of the city relate that Dionysus was reared, and, after
he had learned from the Nysaeans of the deeds of this god in all detail, he composed the
‘Phrygian poem’, as it is called, in which he made use of both archaic language and letters.
(Trans. Oldfather slightly modified)

4. Diodorus Siculus 5. 57:

ƒ ’  HºØÆØ Øç æ Ø ª ÅŁ  H ¼ººø K ÆØ fi Æ Ø ªŒÆ ŒÆd ºØ’ K
Iæ º ª fi Æ . . . .  ø b ŒÆæ b N ¸  Iç Œ  , ˚ƺ  b N c ˚H·
Œd ’ N `Yªı   I æÆ ŒØ c  HºØ  ºØ O ÆÇ Å, I e  F Ææe
Ł   c æ Ū æ Æ· ƒ ’ `Nª Ø Ø ÆŁ  Ææ’ ÆP F a æd c Iæ º ª Æ
Ł øæÆÆ. (3) o æ  b Ææa  E  ‚ººÅØ ª   ı ŒÆÆŒºı F, ŒÆd Øa c
K æ Æ H º ø IŁæ ø I º ø, › ø   Ø ŒÆd a Øa H ªæÆø
 ÆÆ ıÅ çŁÆæBÆØ· (4) Ø’ m ÆN Æ ƒ `Nª Ø Ø ŒÆØæe hŁ   ºÆ 
KØØ ØÆ a æd B Iæ º ª Æ, ŒÆd H  Eººø Øa c ¼ª ØÆ ÅŒØ H
ªæÆø IØ Ø ıø K åı , ‰ ÆP d æH Ø c H ¼æø oæ Ø
K ØÆ . (5) › ø b ŒÆd ŁÅÆE Ø Œ Æ  K `Nª ø fi ºØ c O ÆÇ Å
!Ø, B › Æ ıå  Iª Æ Øa e ŒÆÆŒºı. Ø’ L ÆN Æ ººÆE o æ  ª  ÆE
˚  › ª æ  KŒ B  Ø ŒÅ æH   ºçŁÅ Œ  ÆØ ªæÆÆ N c  EººÆ·
ŒÆd I ’ KŒ  ı e º Ø e ƒ  ‚ººÅ   Æ I Ø æ  ıæ Œ Ø æd H ªæÆø,
Œ ØB Ø  Iª Æ ŒÆ å Å  f  ‚ººÅÆ.
The Heliadae, having shown themselves superior to all other men, likewise excelled in
learning and especially in astrology; . . . Of their number Macar came to Lesbos, and
Candalus to Cos; and Actis, sailing off to Egypt, founded there the city called Heliopolis,
naming it after his father; and it was from him that the Egyptians learned the laws of
 Appendix 2 363
astrology. 3 But when at a later time there came a flood among the Greeks and the majority
of mankind perished because of the abundance of rain, it happened that all written
monuments were also destroyed in the same manner as mankind; 4 and this is the reason
why the Egyptians, seizing the favourable occasion, appropriated to themselves the know-
ledge of astrology, and why, since the Greeks, because of their ignorance, no longer laid any
claim to writing, the belief prevailed that the Egyptians were the first men to effect the
discovery of the stars. 5 Likewise the Athenians, although they were the founders of the city
in Egypt men call Saïs, suffered from the same ignorance because of the flood. And it was
because of reasons such as these that many generations later men supposed that Cadmus,
the son of Agenor, had been the first to bring the letters from Phoenicia to Greece; and after
the time of Cadmus onwards the Greeks were believed to have kept making new discoveries
in the science of writing, since a sort of general ignorance of the facts possessed the Greeks.
(Trans. Oldfather slightly modified)

5. Diodorus Siculus 5. 74. 1:

ÆE b  ÆØ  ŁBÆØ Ææa  F Ææe c H ªæÆø oæ Ø ŒÆd c H K H
Ł Ø c æ ƪ æ ı Å ØÅØŒ. æe b  f ºª Æ ‹Ø !æ Ø b æ Æd H
ªæÆø N , Ææa b  ø  ØŒ  ÆŁ   E  ‚ººÅØ ÆæÆ ŒÆØ, y Ø ’
Nd ƒ  a ˚ ı º Æ  N c ¯Pæ Å, ŒÆd Øa  F  f  ‚ººÅÆ a
ªæÆÆ  Ø Œ ØÆ æ ƪ æ  Ø, çÆd  f  ØŒÆ PŒ K IæåB æ E, Iººa  f
 ı H ªæÆø  ÆŁ EÆØ  , ŒÆd Bfi  ªæÆçB fi ÆfiÅ  f º  ı H
IŁæ ø åæÆŁÆØ ŒÆd Øa  F ıå E B æ ØæÅÅ æ Ū æ Æ.
To the Muses it was given by their father (Zeus) to discover the letters and to combine
words in the way which is designated poetry. And in reply to those who say that the Syrians
are the discoverers of the letters, that the Phoenicians having learned them from the Syrians
passed them on to the Greeks, and that these Phoenicians are those who sailed to Europe
together with Cadmus and this is the reason why the Greeks call the letters ‘Phoenician’,
they say that the Phoenicians were not the first to make this discovery, but that they only
changed the forms of the letters, and the majority of mankind made use of that way
of writing them, and so the letters received the designation mentioned above. (Trans.
Oldfather modified)

6. Pliny, NH 7. 192–3:
Litteras semper arbitror Assyrias fuisse, sed alii apud Aegyptios a Mercurio, ut Gellius, alii
apud Syros repertas volunt, utrique in Graeciam attulisse e Phoenice Cadmum sedecim
numero, quibus Troiano bello Palameden adiecisse quattuor hac figura ZCX(, totidem
post eum Simoniden melicum CXOY, quarum omnium vis in nostris recognoscitur. Aris-
toteles decem et octo priscas fuisse et duas ab Epicharmo additas CZ quam a Palamede
mavult. 193 Anticlides in Aegypto invenisse quendam nomine Menen tradit, XV annorum
ante Phoronea, antiquissimum Graeciae regem, idque monumentis adprobare conatur. e
diverso Epigenes apud Babylonios DCCXX annorum observationes siderum coctilibus later-
culis inscriptas docet, gravis auctor in primis; qui minimum, Berosus et Critodemus,
CCCCXC, ex quo apparet aeternus litterarum usus. in Latium eas attulerunt Pelasgi.
I have always been of the opinion that the letters were of Assyrian origin, but other writers,
such as Gellius for instance, suppose that they were invented in Egypt by Mercury: others
will have it that they were discovered by the Syrians; and both that Cadmus brought from
Phoenicia sixteen letters into Greece. To these, Palamedes, it is said, at the time of the
Trojan war, added these four, Ł, , ç, and å. Simonides, the lyric poet, afterwards added the
same number, Ç, Å, ł, and ø; the sounds denoted by all of which are now received into our
alphabet. Aristotle, however, is rather of the opinion that there were originally eighteen
letters, Æ  ª  Ç Ø Œ º   æ   ı ç, and that two, Ł namely and å, were introduced by
Epicharmus, and not by Palamedes. Anticlides says that a certain Menes, in Egypt, invented
364 Appendix 2
the letters fifteen years before the reign of Phoroneus, the most ancient of all the kings of
Greece, and this he attempts to prove by the monuments. On the other hand, Epigenes, a
writer of very great authority, informs us that the Babylonians have a series of observations
on the stars, for a period of seven hundred and twenty thousand years, inscribed on baked
bricks. Berosus and Critodemus, who make the period the shortest, give it as four hundred
and ninety thousand years. From this statement, it would appear that letters have been in
use from all eternity. The Pelasgi were the first to introduce them into Latium.

7. Lucian, Consonantium lis, 5. 3:

ŒÆd ‹ ª æH  E  f  ı   ı ØÆı Æ, Y ˚  › ÅØÅ Y
—ƺÆÅ › ˝Æı º ı—ŒÆd !Øø fiÅ b Ø Ø æ   ıØ c æ Ł ØÆ ÆÅ— P
B
fi  Ø  , ŒÆŁ’ m ƃ æ æ ÆØ  ÆØ FÆØ, ØæØÆ,  æH  ÆØ j   æ , Iººa
ŒÆd ØÅÆ, L ,ŒÆ  H å Ø, ŒÆd ı Ø ı E .
Now the first who shaped for us these laws, either Cadmus the islander or Palamedes, son of
Nauplius, or also Simonides, whom some credit with the measure, did not merely determine
our order of precedence in the alphabet, which should be first or second; they also had an
eye to our individual qualities and faculties.
 APPENDIX 3

Official Letters Sent by Greek Poleis or Koina

and Inscribed on Stone, in Chronological Order

The list stops at the beginning of our era. Note that covering letters have normal (non-
highlighted) numbers; letters transmitting information or a request are numbered in italics;
letters that in and of themselves conveyed a decision are numbered in bold; the five Cretan
‘letter-decrees’ are grouped as nos. 70–4. I have underlined the numbers corresponding to
those letters whose prescript or conclusion is fragmentary or lost, since this implies
uncertainty as to the exact formal presentation of the document. Sender and place of
inscription are in bold, as is the date. A list of seventy-one letters sent by Roman magistrates
in the same period of time is given separately; I have not added references to mentions of
letters in inscriptions. For both Greek and Roman letters, bibliographical references have
been kept to a minimum.
1. Letter of the magistrates and city of Istron to the Coans concerning the asylia,
inscribed in Cos, 242 bc.
Rigsby and Hallof 2001: 335–8, no. 2 (SEG 51, 1056) = IG xii 4,1 214, A ll. 1–13.
ll. 1–3: the kosmoi and the city of Istron greet the council (or, the magistrates? the text is
restored; see below, no. 2) and the people of Cos; the greeting is followed by a narrative,
with first person plural, ‘you have sent to us’ ([
 ]æø[ø ƒ Œ  Ø ŒÆd ± ºØ ˚ø ø
AØ øºAØ ŒÆd HØ] |  øØ åÆæØ· I [ºÆ Ææ’ ± b Łøæf ]- | æØ , ˜øÆ,
—ºøÆ, [Q K ªªغÆ . . . ); the decision comes at ll. 6–7 in the forms typical of a decree:
_
IªÆŁAØ åÆØ·] |  åŁÆØ AØ ºØ A[Ø
 æøø åŁÆØ (and other infinitives); at ll.
11–12 the prescription to inscribe the decision in the prytaneum of Istron labels the
document with psephisma (e b ł[çØ Æ   IƪæłÆØ K HØ] | æıÆøØ , cf.
no. 2 below); there are no final greetings (the text closes on the indemnity to be given to the
theoroi, l. 13). Another letter (no. 2) is inscribed directly underneath.
2. Letter of the kosmoi and the city of Phaistos to the Coans concerning the asylia,
inscribed in Cos, 242 bc. Inscribed under the preceding letter (and followed by a
decree of Hierapytna).
Rigsby and Hallof 2001: 335–8, no. 2 (SEG 51, 1056) = IG xii 4,1 214, A: ll. 14–18, B: ll.
19–24.
ll. 14–16: the kosmoi and the city of Phaistos greet the magistrates (or the council? The text
is restored; see above, no. 1) and the city of Cos; the greeting is followed by the motivation,
introduced with K  (ÆØø ƒ Œ  Ø ŒÆd [± ºØ ˚ø ø E ¼æåıØ] | ŒÆd AØ ºØ
åÆæØ · K [d ˚HØØ I ºÆ Łø]- | [æf]   Iıº[Æ . . . ) The decision comes
on the back _ of the stele, B ll. 2–3, with invocation of good fortune, followed by infinitives
(IªÆŁAØ åÆØ· q[ ] . . . ). At ll. 4–5, in the context of the decision to inscribe the document
(in Phaistos, in _ the Pythion), the text is defined psephisma (e b łçØ Æ | [ 
IƪæłÆØ K fiH ƒæHØ ø  º]ºø H —ıŁø·). There is no final greeting, the text
closes on the indemnity for the theoroi. On B l. 7 begins a decree of the Hierapytnians, on
the same topic, expressed in the traditional format: [  IæÆ ıø E Œ  ]Ø ŒÆd
fiH  øØ· K [Ø]- | [c ˚HØØ çºØ Z H  ]ø H  IæÆ ıø<> . . . , B ll. 7–8. _
__ _
366 Appendix 3
3. Letter of the Thessalonikeis to the Delians, followed by decree, inscribed in Delos,
c.230 bc.
IG xi 4, 1053 = IG x 2, 1, 1028 (cf. Rhodes with Lewis 1997: 191).
ll. 1–19: Delian decree for the Macedonian Admetus; ll. 20–45: another Delian decree for
Admetus, adding the exhortation to send an envoy to the Thessalonikeis to have the
honours registered there as well (respectively IG xi 4 664 and 665); ll. 46–54 letter,
beginning with:  ºØ ¨ƺØŒø ˜Åºø BØ ıºBØ ŒÆd | HØ  øØ åÆæØ; genitive
absolutes describe the arrival of the envoy and the accomplishment of his duties, then an
aorist and a perfect mark the acceptance and the sending of a copy of the decree to the
Delians, so that they may know that their requests have been accepted ( æ ŁÆ ŒÆd
(...)  çÆ   E e IªæÆç ‹ ø NB); this is directly followed, at ll. 55–77, by
the decree of the Thessalonikeis (no concluding formula for the letter).
4 . Letter (beginning fragmentary; but at l. 12 PıåE) transmitting a decision of
Lappa (Crete) to Tenos, inscribed in Tenos, second half of the third century bc.
IG xii, 5, 868A (IC 2 xvi 2 = Rigsby 1996, no. 59; cf. Rhodes with Lewis 1997: 308).
At l. 5 Rigsby proposes to restore a  åŁÆØ, to signify the decision taken; his restorations
imply a line-length considerably longer than that supposed by Hiller. That the letter
conveyed a decision is clear, but it is impossible to say how exactly the decision was
expressed.
5. Letter of the kosmoi and city of Knossos to the Coans concerning the doctor Hermias
son of Emmenidas, inscribed in Cos, 221–219 bc.
IC 1 viii 7, Syll.3 528, Samama 2003: 127 (with date 219–217 bc); IG xii 4,1 247 (Rhodes with
Lewis 1997: 300).
˚øø ƒ Œ  Ø ŒÆd ± ºØ ˚Øø AØ øºAØ ŒÆd HØ - |  øØ åÆæØ. K Øc
æıø ˆæıø . . . The first twenty lines survive, but the text breaks off when
still recounting the motivation.
6. Letter of Gortyna to the Coans concerning the same doctor, inscribed in Cos, 218 bc.
IC 4 168 = Laurenzi 1941, no. 3; Samama 2003: 126 (with date 219–217); IG xii 4, 1 248
(Rhodes and Lewis 1997: 302).
ˆæıø ƒ Œ  Ø ŒÆd ± ºØ ˚Øø AØ øºAØ ŒÆ[d ]- | HØ  øØ åÆæØ· K Ø . . .
(a long motivation clause follows). At ll. 20–1 comes the decision, with øº Ø ÆPfiH P- |
[åÆæØ]B, o  ± E K ÆØÆØ; the Coans are also praised at ll. 22–3 for having sent
Hermias . . . The end is lost.
7. Letter of Argos to Magnesia on the Maeander, inscribed in Magnesia, 208 bc.
I. Magnesia 40 (Rigsby 1996, no. 90; cf. Rhodes with Lewis 1997: 69).
Beginning partly lost. [ - - - !ƪ]ø AØ ıº[AØ] | [ŒÆd HØ  øØ åÆæØ· ƒ Ææ
] H
æıÆd ŒÆd | ŁÆæd Ææƪ_ Ø ]غŒ _ —ıŁÆª æı, ˚ ø | [˜Øıı,_
¸ ] —ıŁÆª æı   łçØ Æ | 5 [I øŒÆ HØ ] øØ ŒÆd ÆPd غª IŒº-
| [Łø E K HØ] łÆç ÆØ ŒÆÆŒåøæØ Ø -_ | [æd F IªH] []H
¸ıŒçæıÅH·  ªªæçÆ  s |  E e IªæÆç F łÆç Æ F åŁ |
HØ  øØ. ææøŁ. The letter, addressed_ _ _ _ to the boule and demos of Magnesia on the
Maeander, concerns the acceptance of the Leucophryena and honours for the Magnesian
theoroi; it is followed (ll. 10–19) by a decree: æıA Ææƪ ø Ææa F  ı
H | !ƪø H æe HØ !ÆØæøØ غŒı F | —ıŁÆª æı, ˚ ø F
˜Øıı, ¸Æ ı F | —ıŁÆª æı ŒÆd I  ø _ e łçØ Æ ŒÆd Øƺª- | H
IŒºŁø E ªªæÆ Ø  HØ - | 15 øØ ıºE a ŁıÆ ŒØAØ a H
 Appendix 3 367
åÆØH, | L ıºE › B  › H !ƪø H æe HØ !ÆØ- | æøØ AØ æ ØØ
AØ ¸ıŒçæıÅAØ, ŁB  b | Ææa H Æ ØA ŒÆd _ KŒåØæÆ E K ƪªººØ | ‹
ŒÆd E a ˝ Æ K ƪªººØ ÆØ. Interestingly, the (unusually concise) decree has
no prescript: it may have been left out by the Magnesians, but it is then odd that they chose
to inscribe the covering letter. A dating formula was possibly at the beginning of the letter,
so that the covering letter was incorporated in the dossier.
8. Letter of Knossos to Magnesia on the Maeander (beginning lost), accompanying a
honorific decree by the Cnossians awarding proxenia and related honours to two
citizens of Magnesia (ll. 9–19), inscribed in Magnesia, 208? bc.
I. Magnesia 67; IC 1 viii 10 (Rhodes with Lewis 1997: 300, with date third–second century bc).
[ - - Œæ] I [ - - ] | [ . . . . . . c.19 . . . . . . H] Ææ’ ± d Ø ø | [ . . . . . . c.22 . . .
. . . ․․] ± d ØÆÆçÅÆ[]- | [ - - ]ø·  ªª[æ]çÆ  b H łÆç Æ- | 5 [ e
I]ªæç· s s Ø IƪæłÆ- | [ K ]a øØa ŒÆd  · IªæłÆ  |
ªa[æ Œ]Æd ± b K ºÆØ ºØŁÆØ ŒÆd KŁŒÆ  | K HØ ƒÆæHØ H  ººø H ˜ºçØø· |
[ææø]Ł.| The decree began at l. 9 (K d A æåØÆ Œ Ø ø H f Zø- | 10 æåøØ
 ˚ø[ø] []E Œ [ Ø ] ŒÆd | AØ ºØ· ¨ØªÆ # ø,
 ç[HÆ
 ]- | çH
!ªÅÆ I e !ÆØæø [æ]ı | q  ŒÆd PæªÆ ), and went on until l. 21, where
the stone breaks off, so that _ the_ details of the request for inscription cannot be known.

9–10. Letter of the magistrates and synhedroi of the Aitolians to the boule and demos
of Xanthos (ll. 79–88), and letter of the Kytenians to the boule and demos of Xanthos
(ll. 88–110). Before the letters, decree of Xanthos (ll. 1–72), taken as an answer to the
requests (and mentioning the letters with the requests at ll. 10, 12–13, 67–9), and decree
of the Aitolians (ll. 73–9). Inscribed in Xanthos, 206/205 bc.
SEG 38, 1476, C and D (cf. Rhodes with Lewis 1997: 141, 153, 441).
Letter of the Aitolians, ll. 79–88: ªºÆ , —Æƺ- | ø, ! º ŒÆd ƒ æØ H
`NøºH ˛ÆŁø AØ ı- | ºAØ ŒÆd HØ  øØ åÆæØ. ¸[Æ] æÆ , `Y , Ū , ƒ
I - | øŒ   E a K غ, Kd b ˜øæØE . . . . ŒÆºH s Ø . . .
ææøŁ, and letter of the Kytenians, ll. 88–91: ˜øæØø H I e | !Ææ ºØ ƒ
ºØ ˚ıØ NŒ ˛ÆŁø AØ ı- | ºAØ ŒÆd HØ  øØ åÆæØ· I ºŒÆ 
Ł
 b æ- | Ø . . . ; ll. 99–102: IØÇ  s  b | ÆŁÆ A ıªªÆ A
 ÆæåÆ ± E | Ł’  b c æØØE a ªÆ A K AØ !Ææ º[Ø ]- | ºØ
˚ıØ KƺØçŁEÆ, Iººa ÆŁBÆØ . . . ; ll. 108–10: ªØŒ b P  ± E
PåÆæØB K  Iººa ŒÆd | []E `NøºE ŒÆd E ¼ººØ ˜øæØØ AØ ŒÆd ºØÆ
ÆغE | — º ÆøØ Øa e ıªªB ± H r  ŒÆa f ÆغE . There is no closing
___
greeting formula.
11. Letter of Sybrita (Crete) to the Teians, inscribed in Teos, 201 bc.
IC 2 xxvi 1 (Rigsby 1996, no. 141; cf. Rhodes with Lewis 1997: 307, with date 204–202 bc).
#ıæØø ± ºØ ŒÆd ƒ Œ  Ø %Åø AØ øºAØ ŒÆd HØ |  øØ åÆæØ (they go on to
mention first Perdiccas, the envoy of Philip V, and then the Tean ambassadors; the central
part is very fragmentary, and at l. 21 the decisions taken are referred to with Ææa e ªæÆçb
 ª Æ); the text closes with PıåE (l. 27).
12. Letter of Polyrrhenia to the Teians for asylia, inscribed in Teos, 201 bc.
IC 2 xxiii 3 (Rigsby 1996, no. 137; cf. Rhodes with Lewis 1997: 307, with date 204–202 bc).
—ºıææÅø | —ºıææÅø ƒ Œ  Ø ŒÆd ± ºØ %Åø HØ  øØ | ŒÆd AØ øºAØ
åÆæØ· Then, a sentence with verb in first person plural; decision:  - | åŁÆØ
—ºıææÅø E Œ  Ø ŒÆd AØ ºØ· I Œæ- | ŁÆØ (ll. 7–9); conclusion: ææøŁ
(l. 13).
368 Appendix 3
13. Letter of Kydonia to the Teians, inscribed in Teos, 201 bc.
IC 2 x 2 (Rigsby 1996, no. 139; cf. Rhodes with Lewis 1997: 301, with date 204–202 bc).
˚ıøØÆA. | ˚ıøØÆA ± ºØ ŒÆd ƒ ¼æå %Åø AØ øº[AØ] | ŒÆd HØ  øØ
åÆæØ. K Øc [%Ø]Ø çºØ ŒÆd ıªªE |  æå Øa æª ø . . . (motivation
clause) . . . I ŒæÆŁÆØ Ø Ø (l. 16) . . . (decision) . . . ææøŁ (l. 27).
14. Letter (? beginning missing; at the end, l. 11, ææøŁ) of Hierapytna to the Teians,
inscribed in Teos, 201 bc.
IC 3 iii 2 (Rigsby 1996, no. 144).
15. Letter of the Thebans to Polyrrhenia, inscribed in Polyrrhenia, end third–
beginning second century bc.
IC 2 xxiii 1 (cf. Rhodes with Lewis 1997: 119, 307).
¨ÅÆø | [ƒ] ¨ÅÆø º ÆæåØ ŒÆd ƒ æØ —[ºı]- | [æÅ]ø E Œ  Ø Œ[Æ]d
B
fi ºØ åÆæ[Ø·] | [F] Ææ
 E łÅç Æ E I Æ[ºE]- | [Ø ]Ææ
 H
 ªªæçÆ   E e <IªæÆç·> . . . The beginning of the Theban decree follows.
16. Letter of Axos in Crete to the Aetolians, inscribed in Delphi, 200–170 bc, with the
corresponding decree of the Aetolians.
IC 2 v 19 = Syll.3 622, B (Rhodes with Lewis 1997: 153, 300).
ϝÆø ƒ Œ  Ø ŒÆd ± ºØ `NºH[ ıæØ ] ŒÆd HØ æÆ- | ƪHØ ŒÆd HØ ƒ æåÆØ
åÆæØ. ªØŒ . . . (information concerning Epikles, followed by request) . . . ll. 10–13:
ŒÆºH t ØÅE çæ- |  ‹ ÆØ . . . There is no concluding formula. This is
neither a covering letter, nor a letter-decree: the Axians are informing the Aetolians of the
situation of their citizen Epikles, requesting that if someone tries to harm him and his
family he may be prevented, on the common and on the private level (koina and idia), and
that the agreement of common citizenship may remain forever (± b ŒØ ºØ[Æ ] IØÆ
 æåfiÅ I[ƪæÆç]).
17. Letter of an unknown city (Cnidus?) to Bargylia, probably from Bargylia, beginning
second century bc.
I. Iasos 606 (Rigsby 1996, no. 174; cf. Rhodes with Lewis 1997: 327, 329).
[ - - ¼æå] BÆæªıºØÆH AØ ºØ [åÆæØ·] | [ - - I ]ºŒÆ  Ł
  [ - -]
| . . . (the following text, fragmentary, concerns ambassadors, asylia, an agon, thearodokoi,
and inscription on stone) . . . l. 17 ææøŁ. After the letter on the stone is a decree, either by
Bargylia or by another city, ll. 18–29; but as Rigsby 1996: 355 points out, the letter is not an
accompanying cover for the decree that follows it.
18–19. Group of four documents, of which two are letters, inscribed in Magnesia, 194/
193 bc.
I. Magnesia 91 = IG ix2 1 187, 13 ff. (Syll.3 598; McCabe Magnesia 119 and 120; Rhodes with
Lewis 1997: 153).
(a) Honorific decree of the Amphictions for Sosikles son of Diokles.
(b) Letter of the Delphians, accompanying and explaining the preceding decree (ll. 1–2:
[ƒ ¼æå ˜]ºçH ŒÆd ± ºØ !ƪø AØ [ıºAØ ŒÆd HØ  øØ] | [åÆæØ·
ª]ØŒ), cf. ll. 7–8: [ŒÆd  E K]- | [ŒæÆ  ]  E ªæłÆØ æd ÆPF.
[ææøŁ].
(c) Honorific decree of the Aetolians for Sosikles and Aristodamos sons of Diokles of
Magnesia.
 Appendix 3 369
(d) Letter of the strategos of the Aetolians Dikaiarchos to the boule and demos of
Magnesia on the Maeander, probably again praising Diokles (—Ææa F æÆŪF
H `N[øºH]. | ˜ØŒÆ]Ææå !ƪ[]ø AØ ıºAØ ŒÆd HØ  øØ åÆæØ· [ . . . ),
closing with a restored [ææø]- | [Ł].
20. Letter of the archons, synhedroi, and koinon of the Epeirotai to the Rhodian
commander of the League of the Nesiotai in Tenos (?), followed by decree, inscribed
in Tenos, c.200 (192?) bc.
SEG 37, 709; SEG 40, 690 (cf. Rhodes with Lewis 1997: 183).
 ØæøA ƒ ¼æå ŒÆd ƒ æØ ŒÆd e ŒØe ˇ . . . ˝ . . . #`'- | [.]øØ åÆæØ·
æł › I ƺd IæåØŁøæ (followed by four lines, after which _ the text, concerning
the Dodonean Naia, becomes illegible). It is impossible to tell whether there was a final
salutation. At ll. 9–10 the decree resulting from Charops’ embassy was mentioned
([ ªªæ]- | [çÆ]  b  E ŒÆd e łçØ Æ n K[łÅçÆ?]; it is thus clear that this
was (mainly) a covering letter. The decree (basically lost) followed after a vacat, from l. 14
onwards.
21. Letter of Chyretiai in Perrhaebia to Oloosson, also part of the koinon of the
Perrhaebii (ll. 1–11), followed by a honorific decree of Chyretiai (ll. 11–43), 190 bc.
Archaiologike Ephemeris, 1917, p. 10 n. 304 (cf. Rhodes with Lewis 1997: 166).
[ıæØø] [ƒ ]ƪ[d Œ]Æ[d  ] ºØ
Oº[]- | [ø E ƪ]E ŒÆ[d] B fi ºØ
åÆ[æ]Ø· I [º]- | [ŒÆ ] [æ]e  A æ[]ı[] . . . (names) . . . | 5 f
I [Æ] [ E], L [e ł]çØ[ Æ ]- | [åØ ]Ø [a]
¯ ø fi #[ø]
[]æı . . . ææøŁ.
22. Letter of Amphissa to Skarphaea (in Opuntian Locris), inscribed in Amphissa, 200–
150 bc, followed by a decree.
IG ix 12 3, 750 (cf. Rhodes with Lewis 1997: 146 and 148).
After three fragmentary lines (the end of another document) we read: vac IªÆŁAØ åÆØ vac
| 5  çØ ]ø [ƒ ¼æ]å ŒÆd ± ºØ #ŒÆæçø E Iæå [Ø ] | [ŒÆd AØ ıºAØ] ŒÆd
AØ ºØ åÆæØ· H  ø Ø ø  e | [A ] [ºØ][ ] ± H !ÅçøØ
æ Øæı !ÆŒ Ø  (æŒÆø fi | [e I]ªæÆç KÆ ºŒÆ  d a  æÆ
ºØ, | [ŒÆŁ æ] ŒÆd ÆPe › !Å çÆ ± b Æ挺. Then, the text moves to a
decree, the only marker being a dating formula, and the term nomographon to indicate the
decision: Åe  [H]- | 10 [] [)]ŒÆ fi K ’ NŒØ.  ªæçø· K Øc !Å çÆ
æ Ø[]- | [æı] !ÆŒg  (æŒØ NÆæe Æ  çŁd  e A ºØ[ ] . . . The
resolution formula appears at l. 24:  åŁÆØ fiH  ø fi · K ÆØÆØ  ÆPe . . . Then, there
are instructions for sending copies of the text to Skarphea and Opous, and from ll. 29 to 33 a
short summary of the honours awarded to Menophantos. There is no concluding formula.
23. Letter of the prytanis and the demos of Byllis (Illyria) to Sparta, inscribed in Sparta,
beginning second century bc.
IG v i 28 (cf. Rhodes with Lewis 1997: 81 and 185).
BıººØ ø › æÆ- | [Ø Œ(Æd) › ]B [ ] ¸ÆŒÆ[Ø]- | [ø]  øØ åÆæØ· [I ]- |
[ºŒ]Æ [] Ø ø ø[] | [Ææ
_ H _ IªæÆç]
_ _ . . . The rest_ is lost.
__ _ _
_ _ _
24. Letter of the kosmoi and the polis of Allaria in Crete to the boule and the demos of Paros,
together with a Parian decree, inscribed in Allaria, beginning second century bc.
IC 2 i 2 B (cf. Rhodes with Lewis 1997: 299).
The story begins with messengers from Paros bringing a decree to Allaria (A); the letter
follows: ººÆæØøA ƒ Œ  Ø ŒÆd ± ºØ —Ææø AØ ıºAØ | ŒÆd HØ  øØ åÆæ .
370 Appendix 3
Ææƪ ø H æı- | A ’ I , with a narrative in genitive absolutes that ends
in the resolution formula: IªÆŁAØ åÆØ  åŁÆØ ººÆæØøA | E Œ  Ø ŒÆd AØ ºØ
(ll. 9–10); this is followed by decisions in the infinitive (which include inscribing the text in
the two cities, if there is agreement, 19–25: Ka <b> | ıŒE ÆFÆ HØ  øØ HØ
—Ææø, IƪæÆ- | łø ƃ ºØ I ç æÆØ K ºÆ ºØŁÆ | ŒÆd IŁø —æØØ b
K e ƒæe A ˜ Ææ , | ººÆæØHÆØ b K e ƒæe H {Ø}  ººø . ÆF- | Æ b rÆØ
Kç’ ªØÆØ ŒÆd øÅæÆØ A ºø | I çæA.) The letter closes at ll. 25–6: Ka  Ø
çÆÅÆØ  E æŁEÆØ | j IçºÆØ, PåÆæØH  .
0EææøŁ.
25. Letter of the Metropolitai to the magistrates and the polis of Hypata (ll. 1–8),
followed by a honorific decree for a doctor, inscribed in Hypata, 182–181 (IG ix)
or 160–159 (Corrigenda) or 179–146 bc. (Wilhelm).
IG ix 2, 11; Wilhelm 1909: 146, 132; Samama 2003: 77 (differently Rhodes with Lewis 1997:
178, who do not follow Wilhelm’s restoration and consider this a letter from an unknown
city).
[!]Ææ [ºØH ƒ ]ƪd ŒÆd  [ ]ºØ | [ ( ÆÆø ]E[ ¼]æåı[Ø] ŒÆd [B]Ø - | º[Ø
_ _ _ _ _[H]
åÆæØ]. _  ø Ø- | ø  e B ºø  H | 5 ˆºÆŒøØ ¯Pæı HØ  -
| [æø]Ø ºÅØ I ºŒÆ  |  [E e] IªæÆç, ¥ Æ N- | B[]. There are no closing
greetings; the decree follows directly on the stone.
26. Letter of Aptera in Crete to the Teians, inscribed in Teos, 170 bc(?).
IC 2 iii 2 (Rigsby 1996, no. 154; cf. Rhodes with Lewis 1997: 299)
Long letter (58 lines), referring to a previous decree ( ª Æ, l. 14 and 54), of which the letter
is an IÆøØ , a renewal (ll. 54–5).  æÆø |  æÆø ƒ Œ  Ø ŒÆd ± ºØ %Å- | ø
AØ øºAØ ŒÆd HØ A øØ åÆæØ· . . . 25 Øe ŒÆd  åŁÆØ HØ  øØ | I ŒæÆŁÆØ . . .
ææøŁ (l. 58). The earlier text is Rigsby 1996, no. 145 = IC 2 iii 1, dated to 201 bc;
fragmentary both at the beginning and at the end, this text, known only from an eight-
eenth-century bad copy, is in part written in the first person plural (in l. 2: [I ÆÇ] ŁÆ
ŒÆd [K] ÆØH ; l. 4: [)]Œ z ŒÆd Ææ’ ±[ H; l. 6: Øæ[] ); the text we have ends
with KłçØ[ÆØ K d Œ  ø H a] | BæŁø, Åe ˜[Ø]Œ[ı]Æø [ - - ], ll. 12–13. This
might have been a letter too.
27. Letter of the Athenian strategoi to the epimeletes of Delos Charmides, followed by a
senatus consultum, inscribed in Delos, post 164 bc.
I. Délos iv, 1510.
ƒ æÆŪd Ææ Ø K Ø ºÅ- | E ˜ºı åÆæØ· ª ø | ºØ ø º ªø K E
ıºE | æd ( . . . ),  ( . . . ) · 10 ªæłÆØ b ŒÆd æ  ( . . . ) · 11–14  -| åÆ  
Ø ŒÆd F K-| åŁ 
ÆPF  ª Æ | e IªæÆç. (Followed by the senatus
consultum.)
28. Letter of the grammatophylax of Sparta to Amphissa, followed by a decree of the
polis, inscribed in Amphissa, mid-second century bc.
Rousset 2002a: 83–90 (SEG 52, 541).
IªÆŁAØ åÆØ | ¸ÆŒÆØ ø ªæÆ Æçº<Æ> — ººØ
¯ Øæı | › K d
Oæ ı
 çØø ¼æåıØ ŒÆd AØ ºØ åÆæØ· | H Zø Ææ
± b ªæÆ ø K HØ Æ øØ
 - | 5 ªªæÆçÆ  E e IªæÆç. v  ªæçø ˜Æ Œæ- | Å ¸Æçæı· vv ˜Å æØ
! ı  çØB æ  | r  ŒÆd PæªÆ A º ÆPe_ ŒÆd_ KŒª ı · |  æåØ
b ÆPHØ ŒÆd ªA ŒÆd NŒÆ ŒÅØ ŒÆd IºØÆ[] | ÆPHØ ŒÆd KŒª Ø ŒÆd IçºØÆ ŒÆd
_
IıºÆ ŒÆd a ºØ- | 10 a  ØÆ ‹Æ ŒÆd E ¼ººØ æØ ŒÆd PæªÆØ A | º
 æåØ· .
 Appendix 3 371
29. Letter of archons and polis of Eretria to the boule and demos of Cos, accompanying
a honorific decree for Coan judges, inscribed in Cos, mid-second century bc.
IG xii 4, 1 169; Crowther 1999: 293 (SEG 49, 1116).
[æı?] F æı غ çæ F %Ø æåı | [F ˜Æ ]ŒºF —Ææ Œı F
_ crowns | [
¯ææØø] ƒ ¼æå_ _ ŒÆd
%Ø Æ | three _  ºØ ˚Øø
_ BØ | [ıºBØ ŒÆd] HØ
 øØ åÆæØ· H KłÅçØ - | 5 [ø ç
 ]H Ø H HØ   øØ  H ŒÆd | [E
I ]ƺEØ ØŒÆÆE I ºŒÆ- | [   E I]ªæÆç çæƪØ Ø BØ Å - |
[ÆØ çæƪE]Ø ¥ Æ ÆæÆŒºıŁB. ææøŁ. (ll. 1–8). The motivation clause of the decree
follows immediately after a vacat (ll. 9–10: [K Øc  ]łø  H łçØ Æ ŒÆd ØŒÆ- |
[ƪøªf Œ]Æd æı . . . ), without any dating formula or other introduction.
30. Very fragmentary letter of the Kerkyreis to the Ambrakiotai, inscribed in Kerkyra,
mid-second century bc (B, ll. 1–16), followed by the description of the boundaries
of the Ambrakiotai (ll. 17–24) and of the Athamanes (25–33), and preceded, on the
left side of the same stone, by a letter of P. Cornelius Blasio to the Kerkyreis with
annexed senatus consultum (A, ll. 1–20, then break). The dossier concerned a
territorial dispute between the Ambrakiotai and Atamanes.
IG ix 12 4, 796B; SEG 47, 604; 49, 591bis.
Œæ Æ e ª[  æd åæÆ ŁÆ AØ ŒÆd  æÆŒØÆØ ·] | ˚æŒıæÆø [ƒ ¼æå - -
 æÆŒØøA E ¼æåıØ - - åÆ]- | æØ· #ıæ æ[] [åı?, - - I ƺ ] |
Ææ
 H I øŒÆ ± E ŒÆd e Ææ
 H łçØ Æ. . . . At l. 17 Hallof proposes to recognize
the beginning of the boundary description, and to restore a greeting formula before it: - - ] |
Œ ‹Ø ŒÆa f —`'`[ - - - ææøŁ.] | 17  æÆŒØøA æتÅØ [ - - ].
31. Letter(?) of the Pharians to the Parians.
SEG 23, 489 (cf. Rhodes with Lewis 1997: 185 and 262).
Only the final part of the document is extant, without a concluding formula: ll. 1–2 [› EÆ
F E , › EÆ] ZźÆ I ºÅ- | [Æ a ªæÆ Æø - - ]ı F #ŒØıæı, but its
position makes it likely that it is indeed a covering letter. It is followed by a decree of the
Pharians (ll. 3–21), one of the Parians (ll. 22–40 and B, ll. 1–22) and by an oracle of the god,
all inscribed in Pharos, mid-second century bc.
32. Letter of the archons of Elatea to the Delphians, accompanying a manumission,
Delphi, c.150–130 bc.
Fouilles de Delphes III 2: 120.

¯[ºÆø ƒ ¼æå] ˜º[çH] []E Iæå []Ø Œ[Æd] | AØ ºØ åÆæØ. ˆØŒ a
çæƪØ [ ]- | Æ HØ K غøØ I ºıŁæÆ, IƪªæÆ Æ[] | K HØ Ææ’ ± b
ƒæHØ <H>Ø K ŒºÆ ØøØ.  E s s Ø[]- | , çæÆ ‹ ø ŒÆd Ææ
 b
IƪæÆçB fi ± ÆP- | a I ºıŁæøØ K HØ ƒæHØ F  ººø | F —ıŁı. The
manumission (ll. 8–26) and the names of the witnesses (ll. 26–8) followed .
33–35. Three letters, two from Sicyon, one from Kerynia, in a group of eleven as yet
unpublished documents, from Argos, post 146 bc.
Rhodes with Lewis 1997: 69–70, 76, and 97.
The dossier comprises an Argive decree, followed by three letters of a consul (L. Mummius)
to the Argives; then, a letter of the synhedroi of Sicyon to the Argives (#ØŒıø æØ
. . . æªø E ¼æåıØ åÆæØ), followed by a letter of L. Mummius to the Sicyonians
and by a letter of the grammateus of the synhedrion of the Sicyonians to the Argive
damiorgoi (#ØŒıø ªæÆ Æf ıæø ( . . . ) æªø Æ ØæªE å[Æ]æØ); then, a
letter of Kerynia in Achaia to Sicyon (˚Ææıø Æ Øæªd ƒ e æ  ( . . . )
372 Appendix 3
#ØŒıø [E ¼æ]åıØ ŒÆd ıæØ åÆæØ), one of Q. Fabius Maximus to Dyme, and
at the end the decisions taken.
36. Letter of the archons and polis of Delphi to the Athenians acknowledging reception
of a psephisma and honouring the Athenian demos (ll. 1–17), inscribed in Delphi,
shortly after the mid-second century bc.
Fouilles de Delphes III 2: 94 (cf. Rhodes with Lewis 1997: 131).
[ƒ ¼]æå ˜ºçH ŒÆd ± ºØ ŁÅ[Æ]ø AØ [ıºAØ ŒÆd HØ]  øØ åÆæØ. This is
followed by the content and then the resolution: ƒ Ææ
 [H] Ææƪ Ø Ł
± b
æıÆ ( . . . ) I øŒÆ ( . . . ), غªÅÆ ( . . . ), K ØÆ ( . . . ) · K ÆØE  ( . . . ) ·
 åŁÆØ s ( . . . ). There is no concluding formula.
37–38. Letters of the Gortynioi (ll. 115–21) and of the Hierapytnioi (ll. 125–30), quoted
in the context of the inscription preserving the decision of the judges of Magnesia on the
Maeander (ll. 1–11) in the dispute opposing Hierapytna and Itanos. (The two parties,
after an intervention of Ptolemy in favour of the Itanioi, asked the Romans, who
suggested Magnesia as an arbiter.) The dossier is known through two copies, one very
fragmentary from Magnesia, the other one from Itanos in Crete; it was inscribed in 112/
111 bc, but the two letters are earlier (c.145 bc?).
IC 3 iv 9 (cf. the copy in Kern I. Magnesia 105).
l. 107: ªø ] ªaæ c F Æغø —º Æı æÆÆ ŒÆd [ŒŒıæø ]Å Ææa F
ŒØ[F H ˚æÅÆØø] | [ Ææ
z] ºÆ P ŒÅØ ŒÆŁ Ø e ÆæÆŁb  E تæÆ [ Æ
æØ]Eå,
 Æ[][Ø ˆæØØ I ]- | []ºÆ (K) غc ØçÅÆ ‹Ø K d c
B ÆPH c [ŒÆºı] []Å ¸Œ[Å —æÆØØ ººıØ | ( . . . ) 111 [ˆ]æıø b
ı Ææ ø K d B Œæø  IæÆ ı[]Ø , [ ÆæŁ]  E
 [ØØ I Æ]- |
[ºEÆ] æe *Æı[]f æ æ  e ˆæıø K غ, Ø
w[ ] [ŒÅ]º[] –[ ÆØ
Kª ‹Ø | [K Ø ºH ] æ Ø ˆæØØ F ŒÆa
 Æı ı çæ K[ ]ı
ÆP[E æd —æÆ]Ø[ø ‹Ø] | [ æd c] B ÆPH c ¸ŒÅ ªÆØ ‰ ŒıæØ .
[ . . . .]   K غ[c ŒÆd IªæçÅ, ª]- | 115 [ Å ] 
Pe IغªÆ , KŒæÆ 
b ŒÆd ÆPB I[ªæÆç] ŒÆÆåøæÆØ [e  ªªæÆ ·] | (Here the letter begins, ll.
116–21) [ˆ]æıø ƒ Œ æ Ø ŒÆd ± ºØ
 Æø E Œ æ Ø ŒÆd []AØ º[Ø] åÆæ·
 Ø [Ø_  ’ Iæ Ø ] | [n ]ºøŒ ‹Ø ƒ —æÆØØ NŒ ÆØ æd A
¸ŒÆ ‰ [K Ø]æÆÆ ª [Æ , ŒæÆBÆØ ÆP]- | [A , ] KŒæÆ   E K Ø ºø
I EºÆØ·  b i  ŒÆºH [ÆØ ․] K fiH åøæø fi [- - - - ] | [․․] [․․]Å Ł <>Ø
ÆææH< > _N_ åæÆ å K e åøæ[] K[ ØÆø]· ªªæç[Æ  s  E, PŒ
Z]- | 120 ø ø çºø fiH  ÆغE ŒÆd ÆPE  E, [Ø]a [ Æ]e K[ Ø ]º Ø
Œ[Æ]d ø[º Ø Id - - ] | [․][․]ÆØ fiH  ÆغE ŒÆd E H Æغø çºØ . (The
allusion to Ptolemy Philometor dates the letter to before 145, year of the king’s death; the
Gortynioi had been favourable to the Itanioi, even though in 112 they sided with the
Hierapytnioi). The stone continues with the second letter (ll. 125–30), sent by the Hier-
apytnioi: . . . 125 ÆPd
 ÆØ c  ªªæÆ Å K غc·  IæÆ ıø ƒ [Œ  Ø
Œ]Æd ± ºØ
 Æ[ø E ] | [Œ ] Ø ŒÆd AØ ºØ åÆæ· ºØ  º  ªæłÆØ  E
‹ ø N Øa [e ]ı ÅŒe - - | 129 [ . . . ] › ø b ŒÆd ±  , Y  ŒÆ ı ÆfiÅ Ød  H
ØF K fi A ± [fi A ÅŁd] K Øæł . [ææøŁ].
39. Letter of the prytaneis and of the ‘chosen for the phylake’ of Miletos to the Eleans
(ll. 29–40), in the context of a long document (seventy lines on two columns)
concerning the arbitration between Sparta and the Messenians, inscribed in
Olympia on the base dedicated by the Messenians, c.138 bc.
I.v. Olympia 52, ll. 29–40 = Syll.3 683 (cf. Rhodes with Lewis 1997: 94 and 375).
 Appendix 3 373
!غÅø ƒ æıØ ŒÆd ƒ ØæÅ Ø K d BØ çıºÆŒBØ |  Hºø E ¼æåıØ ŒÆd E
ıæØ åÆæØ· ÆæÆ- | ª ø æe  A æıH Ææa !Åø | !Åæı
F ˜Øıı, غı F ˚æÆı, ŒÆd | Ææƌƺø FÆØ ÆPE IªæÆç æe
- | A B ªªÅ Å Œæø !ÅØ  ŒÆd | ¸ÆŒÆØ Ø ŒÆa e  ª Æ B
ıªŒºı, ŒÆd B  | ıºB ŒÆd F  ı ıªåøæÅø a æź[ø] - | Æ ŒÆd
K ØÆø  E FÆØ ÆPE cª ŒæØ,  - | []Æ ÆPc B fi K غB
fi Kή 
E æı- | []ÆE , ‹ ø ØÆŒ øØ ÆPc æe  A KçæƪØ - | [Å] B fi
fi çæƪEØ. Cf. also ll. 11–12, where it is stated that the decision and the letters
[Å ]Æ
have been brought to Olympia by ambassadors of the Messenians, who asked that the
documents be inscribed.
40–41. Letter of the strategos and the grammateus of the synhedrion of the Magnetes to
the demiourgoi and the demos of Kleitor (the closing formula is lost, but was probably
present), followed by a letter of the strategoi and nomophylakes of Demetrias to the
damiourgoi and demos of Kleitor. The entire dossier was inscribed in Kleitor, probably
c.130 bc.
Peloponnesos, IG v 2, 367 = IPArk 19 (cf. Rhodes with Lewis 1997: 167 and 169, with
reference to SEG 11, 1113 and date to c.160–150)
[Zøº] [æ]Æ[Å]ª[e] !ƪ[]ø ŒÆd !Ø[][Æ ] ªæÆ [ ]Æ[]f ıæı
˚ºØæø []E_ [Å ØæªE ŒÆd HØ  øØ _ åÆæØ. ˜ø]- | [æ Ł] ( . . . four lines of
text . . . ) [I ]ºŒÆ[  ]b  E[] ŒÆd F łÅç[ Æ] [e IªæÆç e
 ªªæÆ . ææøŁ] (ll. 1–6); it accompanied a decree of the Magnetes, ll. 7–23 (at
l. 20 of the decree it was specified that Zoilos and Meidias should write to the cities of
Kleitor, Patrae, and Demetrias); the letter of the strategoi and nomophylakes of Demetrias to
Kleitor followed, at ll. 24–9 (˜Å[ ÅæØø] ƒ []æÆŪd ŒÆd ƒ  çº[ÆŒ ]
˚ºØæ]ø E [Å Øæ]ªE Œ[Æd HØ  øØ åÆæØ. e Ææ’  H] | 25 [I]< >
[ƺ]Æ ØŒÆc ªØŒ ( . ._. the text continue with praise of the judge and request of
inscribing_ the honours voted for him on stone . . . ) I [ºŒÆ ] b  [E ŒÆd F
łÅç Æ e IªæÆç e  ªªæÆ . ææøŁ] ), accompanying a decree of
Demetrias, ll. 30–49 (here too at l. 46 it is specified that strategoi and nomophylakes will
have to write to Kleitor and Patrae). Both the decrees of the Magnetes and of Demetrias
allude to a letter sent previously (and not inscribed) with which they had asked for judges.
42. Letter of the koinon of the Amphictions to the Athenians concerning the Dionysiac
technitai (ll. 40–51), 134 or 130 bc, preceded by a decree of the Amphictions of 278
bc (ll. 2–39), and followed by a new decree of the Amphictions (ll. 52–94) of 130 bc.
Inscribed in Athens, 134 or 130 bc, in the theatre of Dionysos. The letter is,
however, missing from the copy of the same decree inscribed in Delphi, and was
probably also missing from the other copy inscribed in Athens, in the temenos of the
Dionysiac technitai.
IG ii2 1132, ll. 40–51 = Le Guen 2001: TE 7 = CID 4 115. Note the archival docket ‘from the
metroon’: see Thomas 1989: 77 and n. 20; Sickinger 1994: 289–92.
KŒ F ÅæØı· K d ˜Å æı [¼æå Åe BÅ]- | æ ØH · e ŒØe H
 çØŒØ [ø ŁÅÆø E] | ıºE ŒÆd HØ  øØ åÆæØ· æı[ø æe ]- |
A . . . 50–51  ªªæçÆ   E F ª[ªÅ ı] | ç
 H  ª Æ e IªæÆç.
43. Letter of Tyre to Delphi, inscribed in Delphi, 125 bc.; probably brought there by
ambassadors who came at the time when Tyre became independent to ask for recogni-
tion of asylia from the Delphians for their city (or possibly better, following Rigsby 1996:
483–4, to inform the Delphians of some event).
Pomtow, Klio 1918, 26, no. 49; SEG 2, 330; Curty 1995, no. 12; cf. Rigsby 1996: 483–4.
374 Appendix 3
ll. 1–4: ¨ . %å[Å IªÆŁ]. | %æı B ƒæA ŒÆd Iºı  [ıºc ŒÆd › B  ˜ºçH
BØ] | ıºBØ ŒÆd HØ  øØ E ıªª[Ø åÆæØ· ( . . . ) K _Øc K]- | ıº ŁÆ ( . . . ).
The text becomes more and more fragmentary, and after the twelfth line is altogether lost;
at l. 10 we find an aorist, K ØÆ , but that does not necessarily mean that we have
already reached the decisions.
44. Letter of the grammateus of a Macedonian or Achaean polis (Phenea?) to Sparta,
Sparta, 123 or 121 bc.
IG v i 30 (cf. Rhodes with Lewis 1997: 81 and 90, with the restoration []ÆA, and open
date).
[]ÆA | [› EÆ ªæÆ] Æf ı- | [æø - - - ] › e   ŒÆd N- | [Œe 
¸Æ]ŒÆØ ø Kç æØ | [ŒÆd AØ ºØ åÆ]æØ· H Zø _Ææ’[ K]- | [ d ŒØ ø K] HØ
IæåøØ K ıºøØ | [łÅçØ ø ª]ªæÆçÆ  E e I- | [ªæÆç· K d - - ]ŒæÅ
#ç[ı] | [ - - - - s ØH] ØÆ[ ºE - -]
45. Letter (very fragmentary) of the koinon of the Amphictions to the Athenians
concerning the technitai, following a slightly earlier decree of the Athenians,
inscribed in Athens, on the acropolis, in 117–116 or 112 bc.
IG ii2 1134, ll. 77–103 = Le Guen 2001, TE 12, E = CID 4, 120.
l. 77: [e ŒØe H  çØŒØ ø ŁÅÆø BØ ıºBØ ŒÆd HØ]  øØ åÆæØ . . . ; with this
letter, the Delphians confirm an earlier decision (ll. 96–8: KŒæÆ [ a ŒåæÅ Æ]-| Ø Æ
K BØ Ł [øæØBØ ıºÆÆØ K ]d ¼æå K ˜ºçE ¯[PŒºı] | ŒæØÆ rÆØ ŒÆd ÆØÆ);
there is no closing formula.
46. Letter of the koinon of the Amphictions inserted in an Athenian decree (?), from the
agora in Athens, 117/116 bc.
SEG 26, 117 (S. V. Tracy, Hesperia 45 (1976), 287–8 n. 3).
We have only ll. 1–5 (very fragmentary): ‘under the archonship of Menoitos, tribe Anti-
ochis, [?] son of Attalos Berenikides was secretary’, 3 ff.: æØ ƒÆ ı æÅØ | [B
æıÆÆ · ıºc K ıºı]ÅæøØ· H ææø  łçØ- | [Ç - - c.14 - - ŒÆd
ı æ ]æØ vac. 3–4 e ŒØ[e H  çØ]- | [ŒØ ø ŁÅÆø BØ ıºBØ ŒÆd] HØ
 øØ åÆæØ. [vac. c.4 - c.5 - ]
_
47. Letter (very fragmentary) of the grammateus of the synhedrion Xenon to the
strategoi (ll. 17–21), preceded by a letter of a Roman magistrate(?) to the Karystians
or Euboians, and followed by a hypomnema, from Karystos, second century bc.
IG xii 9, 5.
l. 2: [ - - ] K BØ K Ø[]ºE[ - - ] ll. 14–15: [ - - ]e  Å Æ e [æd ø - - ] | [ - -
ŒÆÆ] æ Æ N a Ææ
 [E Å ØÆ ªæ ÆÆ] . . . ll. 17–19: [ - - ]˛[]ø
ªæÆ _Æf _ _ _ _ ı[æı - - ] | [ - - æÆÅ]ª[E] åÆæ[Ø]. H Zø [Ææ - - ] | [ E -
-  ]E [e I]ªæÆç [  çÆ ] . . .
48. Letter of an unknown city to Demetrias (ll. 13–19), followed by a decree (of which
there are scant remains at l. 20) for judges sent by Demetrias; the whole was prefaced
by a decree from Demetrias (ll. 1–12) in which it was decided to have the honours
inscribed (so Wilhelm 1909). Demetrias, II bc.
Wilhelm, Hermes 40 (1909), 53–4 (revision of IG ix 2, 1106, where the text was interpreted
as a decree of the Magnetes followed by a letter of the Magnetes to Demetria; thus Rhodes
with Lewis 1997: 169).
 Appendix 3 375
ll. 13 ff. [ - - - - ]Æ ˜ ø - - - | [ - ˜Å Å]æØø E [æÆŪE ŒÆd] [ |çºÆØ ŒÆd HØ
 øØ åÆæØ. ]H Ø [H H  ø] [ ]Æ- | [æa - - ZغøØ
¯ غ]
æ ØæøØ . . .
49. Letter of the ephors and the city of Sparta to the tagoi and demos of Larisa, Larisa,
second century bc.
IG ix 2, 518.
[¸]ÆŒÆØ ø ç[æ]Ø ŒÆ[d ±] ºØ[ ] | [¸]ÆæØ[Æ]ø E ƪE Œ[Æd] HØ []- |
[ øØ_ åÆæ. I]< >ºŒÆ< _ > [e]Ł
_ ’  b | [ æØ
_ _ _ æØ] ØŒ ¯PÆ _ æ[ı] | 5 [ - - -
_ _
]ˇ#¸' [ - ]˜[ - - - ] | [ - - - ]ˇ˚[ - ]ˇ(`[ - ] | [ - - - ]+˝%ˆ+! ¸[ - - - ] The
_ _ _
mention of envoys from Sparta being sent makes it likely that this is a covering letter.
50. Letter of an unknown city to Sparta (very fragmentary), Sparta, second century bc.
IG v i, 8 (cf. Rhodes with Lewis 1997: 81).
[ - - ø ƒ ]æÆƪd ŒÆd ± ºØ ¸ÆŒÆØ- | [ ø Kç æØ Œ]Æd AØ ºØ åÆæØ· Ł
| [ ØÅÆ ø_ ]  a ıº[a ŒÆd a] 
ŒŒºÅ- | [Æ ± H - - ]F ¸ ,
`Nåæø [F] `Yåæø- | [ H  æø] _ºØA . . . Unclear whether this is a letter
conveying a decision or simply a request.
51–52. Decision of the Amphictions (ll. 1–12) and letter of Hypata to the Amphictions
concerning the designation of the hieromnamon of the Pythian years for Euboea
(ll. 13–25), followed by an abstract of the earlier decisions (ll. 1–9, with at l. 5 the
mention of a second letter sent by the Amphictions to the people of Hypata to ask them
whether they consider that they should revise their decision), and by a new letter of
Hypata to the Amphictions (ll. 9–28), inscribed in Delphi, 110 bc?
CID 4 121 and 122 = Fouilles de Delphes III 1: 578, resp. col. II, 14 ff. and col. I, 9 ff.
A prominent heading (K[ غc  ( ÆÆ]ø, l. 13) and a standard epistolary opening
are followed by the motivation clause ( (_ _ ÆÆø[] ƒ ¼æå[ HØ ŒØHØ H
 çØŒØ ø åÆæØ· K Øc ( . . . ), l. 14); the rest is very fragmentary. The second letter
is quoted within the second record, without a separate heading ( ºØ K[Œ]ÅÆ ƺŒØE
ŒÆŁ [Ø] ŒÆd  ªæÆçEÆ  e [H  ( Æ]-| [Æø Iæå ø æØåØ æe] f
 çØŒÆ K غ· [ ( ÆÆ]ø[] [ƒ ¼æå] |10 [ŒÆd ± ºØ fiH ŒØfiH H]
 çØŒØ ø åÆæØ· KŒ Ø Ł  K غa[] Kfi v [ØÆ]-| [çÆ a ªª Æ ŒÆd
ÆæŒÆº]Æ ( . . . ), ll. 9–12). Here too, the central part and end of the letter are lost.
53. Letter of someone to the magistrates and the people of Kyparissos, Kyparissos,
second–first century bc.
Messenia, SEG 11, 1025; Sherk 1969, no. 46 (see list of letters by Roman magistrates, R25 bis).
[IªÆŁBØ] åÅØ. | [› EÆ - - - H ˚]ı ÆæØø ¼æåıØ ŒÆd B fi | [ ºØ] åÆæØ. | [ - - - -
I] ººåŁÆØ  ’ K F c | [ - - - ]ø K ØÅ Å H | [ - - - - ]ı ƺÆı F . . . Because
of the first person reference, it is unlikely to be a letter from a city; but the text is very
uncertain.
54. Letter of the Delphians to the Athenians, inscribed in Athens, 106–105 bc.
IG ii2 1136, with addenda p. 672 = Syll.3 711K.
Only the two last lines of the letter are preserved (Dittenberger restores three other lines before),
accompanying a honorific decree for the priestess of Athena Chrysis (ll. 2–32). ll. 1–2: [ . . . c.7
. . . Ł]ÅÆøØ H[Ø æ]øØ [e IªæÆç] | [‹ ø] NB· vv ææøŁ. v K Øc Œº.
55. Letter of an unknown polis (Athens has been suggested) to the Delphians, inscribed
in Delphi, end second century bc.
Fouilles de Delphes III 4: 33; SEG 3, 381.
376 Appendix 3
Five small fragments of a document that went on for some 30 lines; it might be an Attic
document sent to Delphi, over which someone might have added, in local dialect, [å]Æ
[IªÆŁ] | [K d - - - ¼æå , K] d B[ - - - æıÆÆ - - - ] | [ - - - , ˜ºçH BØ ]ºØ
åÆ[æØ - - -] | [ - - Ææƪ ø æ]ıH [ll. 1–4 ]. _

56. In the dossier concerning a dispute between Thronion and Skarphaea, inscribed in
Delphi, end second–beginning first century bc, letter with which the judges
(Athenians? this part is in Attic dialect) transmit their decision and the entire
dossier to the Amphictions.
CID 4 123–5 = Fouilles de Delphes III 4: 38, 3 = Klio XVI, no. 130, col. II; cf. Ager 1996, no.
167–I.
(ll. 17 ff.: [ŒæØa IŒÅŒ Æ ‹Ø Kª]  çøØ æe Iºººı , ŒÆd ،Ō Æ[ ] | [f
¨æØÆ Œ]Æd [Nº]Åç Æ łçı ŒÆ KÆ, f b #Œ[Ææ]- | [çÆ NºÅç ]Æ
łçı []. [ˆ]ªæçÆ  s  E, ¥ Æ NB·  ª[]- | [ªæÆ ÆØ b]  E ŒÆd e
IªæÆç B IتæÆçB B I ŁÅ[ ] | [ E  e H] KÆ ƺø
æıH Ææa B ºø B ¨æ- | [Øø, › ]ø b ŒÆd B ŁÅ IتæÆçB
 e H KÆ ƺø[] | [ æı]H Ææa B ºø H #ŒÆæçø). The dossier
contains Thronion’s version of events, in the form of a kind of letter written in the first
person, with epistolary deictics, and even an apostrophe to the other polis, but lacking
introductory and final formulas (ll. 7–16: I çغºªØ ± ºØ H ¨æØø | æd a
ƒæ Æ Æ d a ºØ H #ŒÆæçø· “K غºØ Ø A H[] | [
¯ ØŒ]Æ Øø
¸ŒæH I çØŒØÆ e æ æ , ŒÆŁ æ ŒÆd a æ ÆÆ K[] | [a Łı]Æ ŒÆd a
¼ººÆ a  ØÇ Æ æ æ K f  çØŒÆ ı[ ]- | [º]Å ÆØ, ŒÆd ŒÆŁ
n ŒŒæØ ÆØ
æ æ æd ø K  çÆØ ŒÆa e[] | [I ]çØŒØØŒe  · ŒÆd E ŒæÆE K b
ŒÆd ŒıæØØ F K d K Ø[º]- | [º] æ , ŒÆd e ƒæ  Æ ŒÆŁÆŁÆØ Ææ
K b
ŒÆd  - | ŁÆØ 
K F, K  ŒÆ K d ŒÆŁŒfiÅ ± ƒæ Æ Æ. %f , ºØ #ŒÆæç- | ø,
IØ ØEÆØ ŒÆŒ æƪ ø ŒÆÆçØÇ Æ, ŒÆd KØØÇŁÆ[Ø Ł]- | ºØ IŒø e K d
K غº æ A ƒæ Æ Æ ”). The dossier continues with Skarphaea’s version,
written in an indirect, impersonal and traditional style (ll. 17–22).
57. Letter of the strategoi of Lampsakos to the magistrates of Thasos (ll. 2–8), inscribed
in the agora of Thasos, first third of the first century bc.
I. Lampsakos, no. 7 (cf. Rhodes with Lewis 1997: 412).
¸Æ łÆŒÅH | ¸Æ łÆŒÅH æÆŪd ¨Æø ¼æåıØ | åÆæØ· N æøŁ (sic), e  i
YÅ, ŒÆd  E | 
ªØÆ . H ªª ø | çغÆŁæ ø Ææ
 E HØ ºfiÅ  H |
˜ØıæøØ — ı | KÆ ºŒÆ  a IªæÆçÆ |  Æ  e c K غ.
The letter rather surprisingly presents a formula valetudinis, but lacks a final greetings; it is
followed by two proxeny decrees for a Thasian (ll. 9–46).
58. Letter(?) of the Rhodians to the Thasians (ll. 1–20), followed by an annex decree of
the Rhodians for Dionysodoros of Thasos (ll. 21–7), from the agora of Thasos, first
third of the first century bc.
Dunant-Pouilloux 1958, 172 (cf. Rhodes with Lewis 1997: 267, with a date around 60 bc).
The beginning of the letter (as also the end of the decree) is lost. There is no closing
epistolary formula. It is possible to classify the text as a letter because of ll. 16–19
( ªªæçÆ[ ] | [b Œ]Æd F łÆç Æ , ŒÆŁ
n ø[ŒÆ ± ı]- | [ºa ŒÆd] › A 
a æÆ, e I[ªæÆç] | [¥ Æ . . . ) _

59 . Letter of a city or of a Roman proconsul to Sparta, from Sparta, first century bc.
IG v i, 9: litterae vel urbis cuiusdam vel proconsulis Romani.
[ - - - ¸ÆŒÆØ ]ø Kç æØ ŒÆd | [AØ ºØ åÆæØ· - - ]øØ  H ŒÆd HØ | [ - - - ]HØ
ªæÆ ÆE ÆP|[H - - - - æ]Æ ÆP[]- . . . The rest is lost.
 Appendix 3 377
60. Letter of strategoi and boule of an unknown city to the boule and demos of Tegea,
Tegea, first century bc.
IG v 2, 21 (cf. Rhodes with Lewis 1997: 91, Crowther 1999: 299–300) = IPArk 6.
The beginning is partly preserved, but after l. 3 the text is lost ( - - ]ø æÆŪd ŒÆd 
[ıºc %ªÆH BØ ıºBØ ŒÆd] | [fiH  øØ å]ÆæØ. %H Kł[ÅçØ ø Ø H E Ææ

 H ] | [I ƺE]Ø ØŒÆ[ÆE I ºŒÆ   E e IªæÆç,] | [¥ Æ NB - - - ).
_
61. Letter of the Delphians to the Dionysiac technitai of Athens, inscribed on the wall of
the treasury of the Athenians in Delphi, 98/97 bc, followed by a decree of the
Delphians (ll. 3–61: it closes with the instruction to send a copy of the decree to the
boule and demos of the Athenians, and to the synodos of the Dionysiac artists in
Athens as well).
Fouilles de Delphes III 2:48 = Le Guen 2001, TE 14.
ll. 1–2: ˜ºçH ƒ ¼æå ŒÆd ± ºØ E æd e ˜Ø ı åÆØ åÆæØ· [A ]
ŁÆ ç
± H | I ŒæØ  ªªæçÆ   E e IªæÆç, ‹ [ø NB].
ææøŁ.
62–63. Letter of the Gephyraioi to Delphi (preceded by a dating formula), and answer of
the Delphians to the Gephyraioi, from the agora of Athens, 37–36 bc. An oracle of
Apollo, now lost, closed the dosssier.
IG ii2 1096; SEG 30, 85, 6; (cf. Follet, Topoi 8 (1998), 260 n. 67).
¨[] | ªÆŁB fi åfiÅ.
¯ d ¨ Łı ¼[æå , F b ª]- | ı F ˆçıæÆø
غø[ı ¼æå (Follet prefers a patronymic) —ÆØ]- | [Æ]Øø #ŒØæçæØH [ - - -
- - ] | 5 [ K غc Ææa F ªı æe ˜ºçf ] | [ˆçıæÆø e ª ˜ºçH BØ ºØ
åÆæØ. v çØ]- | [ºÆ I ][ºŒÆ  ƺÆØA N IÆøØ ¨ çغ] | ˜Øæı AºÆØÆ
__
[ŒÆd —Æ Å Zø !ÆæÆŁØ] | - (ll. 9–14, request of the Gephyraioi, without final
greeting) - | 15 K غc Ææa ˜ºçH[ æ]e e ª [ - - - ] | ˜ºçH ƒ ¼æå ŒÆd <±>
[ ]ºØ HØ ªØ HØ ˆ[çıæÆø] | åÆæØ v ˆØŒ f [I ]ƺ< >ı ç’ [ H
K d] _ | a ÆÆ ŒÆd <K> æÆ[Ø ] bæ F BıǪı Œ[Æ]d ƒ[æø ] | [˜]Øe K
—ƺºÆøØ ˜Ø[ ı] F ˜Øæı AºÆØø [¨]- | çغ] ˜Øæı A[ºÆØÆ_ _ _ŒÆd]
—Æ Å Zø !Æ[æÆ]- | 20 [ŁØ I ]  [øŒ Æ ± ]E a Ææ
 H  çŁ[Æ]
| [ªæ ÆÆ æd A Æ]Æ ([łÅç ÆÆ _ŒÆd _ _ a K غ]a Follet) ŒÆd IÆø ı
a | [ æåıÆ   a] ºØ ± H ŒÆd e Łe N- | [ŒØ ÆÆ H ˆçıæÆø]
_
(Follet prefers Wilhelm’s NŒØ ÆÆ ŒÆd çغÆ) ŒÆd ŒŒÆººØæÅŒ Æ ŒÆd | 25
[K æøÆŒ Æ e ) ? –]ªØ. (Follet: K æøÆŒ Æ e Æ]E) a s K ØæÆØ
| [ŒÆd e åæÅ e F ŁF I ]º ŁÆ []Ł
 b | [çæƪØ Ø AØ Æ ÆØ
çæÆ]ªEØ.
64–68: Dossier inscribed in Delphi, 30–25 bc, on the monument of Diodoros, with
letters and city decrees.
64. Letter of Sparta to the Delphians, 30–25 bc.
Fouilles de Delphes III 1:487b = IG v i 1566 (with restoration at ll. 3–4: F ªª[  ç’
± H  ª Æ ] | [˜ØæøØ] ˜øæŁı HØ  æ[øØ ºÆØ I ºº]-);  ª Æ also
in Rhodes with Lewis 1997: 81, with a date to c.29 bc.
[K d] ¼æå —º æ[åı]. | [¸]ÆŒÆØ ø çæØ ŒÆd ± ºØ[ ˜ºçH E ¼æåıØ]
| [ŒÆd AØ ]ºØ åÆæØ· F ªª[  ç’ ± H łÆç Æ ] | [˜Øæø]Ø ˜øæŁı HØ
 æø[Ø ºÆØ I ºŒÆ]- | [ ]  E e IªæÆç· K d ˜Ø [øæ ˜øæŁı
˜ºçe ] . . . Followed by a decree, missing its end.
65. Letter of the strategos and synhedroi of the koinon of the Thessalians to Delphi.
378 Appendix 3
Fouilles de Delphes III 1: 488.
Extremely fragmentary; Bourguet restores K d ¼æå [—º æåı.] | ºØ 
æ[ÆŪe F ŒØF F K ¨ƺ]ÆØ ŒÆd ƒ | æØ ˜[ºçH E ¼æåıØ ŒÆd BØ
ıºBØ å]ÆæØ.
66. Letter of an unidentified polis to the Delphians (text very similar to the following
one).
Fouilles de Delphes III 1: 489.
67. Letter of the doriarcheon of the koinon of the Dorieis of the metropolis Charigenes
to the archons and polis of Delphi, accompanying the annex decree.
Fouilles de Delphes III 1: 490 (IG v i 1566, cf. Rhodes with Lewis 1997: 141).
K d [¼æå —º æåı] | ÆæتÅ › ˜æØÆ[æåø F ŒØF H ˜øæØø H] | Kª
!Ææ ºø ˜[]º[ç]H[ E ¼æåıØ ŒÆd AØ] | ºØ åÆæØ. %H  ø [Ø ø  e
F ŒØ] | F H ˜øæØø ˜ØæøØ ˜øæ[Łı HØ  æøØ] | ºÆØ I ºŒÆ []
 E I[ªæÆç e  ª]- | ªæÆ ·. The decree follows.
68. Letter of an unidentified polis (possibly letter and decree from Hyampolis) to the
Delphians.
Fouilles de Delphes III 1: 495.
69. Letter of an unidentified polis or king/emperor/consul to Sparta, very fragmentary,
Sparta, date? The stone might actually contains remains of a sequence of two letters:
a cover letter by a secretary, transmitting an official letter addressed to the
Lacedaemonians.
IG v i, 10 (ll. 1–3 restored by Wilamowitz, with the annotation ‘Scriba alius doricae civitatis,
rogatus a Lacedaemoniis, exemplum mittit decreti in tabulis suis adservatis’); but note SEG
47, 355: ‘Letter from a Hellenistic or Roman ruler?’
[ - - - - - Æ ]- | []ƺ ÆØ [ - - ] | HØ K
K F [ - - Æ]- | ªæÆç æ[ - - ] | ¸ÆŒÆØ [ø
Kç æØ ŒÆd] | AØ _ [ºØ åÆæØ· - - ]. _

Texts Halfway Between a Letter and a Decree

70. Decree—but with closing epistolary greeting—of Axos, inscribed in Teos, 201 bc.
IC 2 v 17 = Rigsby 1996, no. 140 (called ‘decreto-epistola’ in Ghinatti 2004: 125).
ϝÆø. |  ϝÆø E Œ  Ø ŒÆd AØ ºØ łÆçØÆ Ø | ŒÆa e  · K Øc
(. . .) | 13.  åŁÆØ ϝÆıÆø E Œ  Ø ŒÆd AØ ºØ K ÆØÆØ b (. . .) I ŒæÆŁÆØ b
‹Ø (. . .) 24. ææøŁ.
71. Decree—but with closing epistolary greeting—of the Arcadians, inscribed in Teos,
201 bc.
IC 1 v 52 = Rigsby 1996, no. 150.
æŒø.  æŒø E Œ  Ø ŒÆd AØ ºØ· K Øc Œº ( . . . ) 20–1 I ŒæÆŁÆØ
E æتıÆE ‹Ø ± ºØ ( . . . ) 29–32  |  ŒÆŁØæøØ HØ ˜ØøØ A  ºØ |
ŒÆd A åæÆ  ø ( . . . )   45 ææøŁ.
72. Decree—but with closing epistolary greeting—of the Arcadians, inscribed in Teos,
170 bc? Renewal of asylia.
IC 1 v 53 = Rigsby 1996, no. 159.
`
’ æŒø.  æŒø E Œ  Ø ŒÆd AØ ºØ· K Øc ( . . . ) 17 e æ æ Łb
 E  ª Æ ( . . . ) 18  åŁÆØ I ŒæÆŁÆØ ( . . . ) 50 ææøŁ.
 Appendix 3 379
73. Decree—but with closing epistolary greeting—of Hyrtacina, inscribed in Teos, 170
bc? Renewal of asylia.
IC 2 xv 2 = Rigsby 1996, no. 160.
A fragmentary beginning is followed by a series of infinitives (IƪæłÆØ, K ÆØÆØ); there
are no pronouns of first and second person; conclusion with ææøŁ.
74. Decree—but with closing epistolary greeting—of Priansos, inscribed in Teos, 170 bc?
IC 1 xxiv 1, cf. Rigsby 1996: 289.
—æØÆø. |  —æØÆø E Œ  Ø ŒÆd AØ ºØ. K Øc  Hæ  !<Å> ı
ŒÆd !ŒºB ˜Øı-| ø KÆ ƺ 檪ıÆd æd ± b Æ-| 5 æa %Åø P 
Iæ[ç] < æ> < >ø< > K AØ | ºØ ŒÆd [غª] æd A [ . . . . . . . ][ . . ]Æ
Iººa | ŒÆd K Æ !ŒºB a ŒØŁæÆ   %Ø- | Łı ŒÆd —ºıı ŒÆd H ± H
ƺÆØH ØÅ- | A ŒÆºH ŒÆd æ ø , N<>ªŒ b ŒŒº | 10 ƒæÅ Æ  bæ
˚æBÆ ŒÆ[d ]H K [˚æ]ÆØ ª- | ª ø ŁH  ŒÆd æø, [ Ø]Å [ ]a |
ıƪøªa KŒ ººH ØÅA[] ŒÆd ƒæØƪæ|çø· Øe  åŁÆØ AØ ºØ K ÆØÆØ %Å
‹Ø | ºE º ª ØHÆØ æd ÆØÆ , K ÆØ-| 15 ÆØ b ŒÆd  Hæ  ŒÆd !ŒºB
‹Ø ŒÆºa | ŒÆd æ Æ  ÅÆØ a Ææ ØÅ Æ | K AØ ºØ ± H· ØÆÆçBÆ 
ÆFÆ ŒÆd %Å-| Ø ‹< >ø K تØŒØ. ææøŁ.

Official Letters Of Roman Magistrates Inscribed In Stone

R1. Letter of T. Quinctius Flamininus to the Chyretians, from Chyretiae in Thessaly,
197–194 bc.
Sherk 1969, no. 33; Rhodes with Lewis 1997: 166. Cf. Hofmann forthcoming.
R2. Letter of M. Valerius Messala, the demarchoi and the synkletos (heading: ῾'ø Æø) to
the Teians, from the temple of Dionysos in Teos, 193 bc.
Sherk 1969, no. 34. Cf. SEG 30, 1377; Hofmann forthcoming.
R3. Letter of M. Acilius Glabrio to the Delphians, from Delphi, early in 190 bc.
Sherk 1969, no. 37, Rousset 2002b, no. 41.
R4. Letter of the consul L. Cornelius Scipio and his brother to Heraclea on the Latmos,
189 bc.
Sherk 1969, no. 35; Robert 1978: 501; SEG 37, 860; Ma 1999, no. 45. Cf. Hofmann
forthcoming.
R5. Letter of L. Cornelius Scipio and his brother to the Colophonians, Colophon,
190–189 bc.
Sherk 1969, no. 36; Ma 1999, no. 46.
R6–R7. Letters of the praetor Spurius Postumus to the Delphians and to the Amphictions,
from Delphi (same stone), 189 bc.
Sherk 1969, no. 1 A and B; Fouilles de Delphes III 34: 353; Rousset 2002b, no. 42; and for the
second letter CID 4 104; cf. Ferrary 2009: 130.
R8. Letter of C. Livius Salinator to the Delphians, from Delphi, 188 bc.
Sherk 1969, no. 38 = Syll.3 611 (Bagnall-Derow 2004, no. 41).
R9. Letter of a Licinius to the Amphictions, from Delphi, 186 bc?
Sherk 1969, no. 39; CID 4 105.
380 Appendix 3
R10. Letter of a Roman magistrate or legate on King Perseus to the Amphictions or the
Delphians, from Delphi, 171–170 bc.
Sherk 1969, no. 40 = Syll.³ 613 B (cf. SEG 31, 542; 45, 481).
R11. Very fragmentary letter of the Kerkyreis to the Ambrakiotai, inscribed in Kerkyra,
mid-second century bc, preceded by a letter of P. Cornelius Blasio to the Kerkyreis
with attached senatus consultum.
IG ix 12 796B; Sherk 1969, no. 4; SEG 47, 604.
R12. Letter of M. Aemilius to the Mylasaeans, from Magnesia, accompanying a senatus
consultum on the dispute between the Magnesians and the Prieneans, mid-second
century bc.
Sherk 1969, no. 7; Ager 1996, no. 120.
R13–R16. Three letters of the consul L. Mummius to the Argives, and one from him to the
Sicyonians, all inscribed in Argos, post 146 bc.
Rhodes with Lewis 1997: 69–70, 76, and 97.
R17: A letter (unpublished) of Q. Fabius Maximus to Dyme in Achaia, from Argos, part of
a dossier including the preceding letters, 144 bc.
Rhodes with Lewis 1997: 97.
R18. Letter of Q. Fabius Maximus to the Dymaeans, from Dyme.
Rhodes with Lewis 1997: 97, Kallet-Marx 1995: 129–53, date 144 bc? (Sherk 1969, no. 43,
115 bc?)
R19–R20. Two letters of Roman magistrates to the associations of Dionysiac artists, very
fragmentary, from Thebes, second half of the second century bc.
Sherk 1969, no. 44, IG vii 2413 and 2414; Le Guen 2001, nos. 34 and 51.
R21. Letter of a Roman magistrate to the Amphictions (very fragmentary), from Delphi,
c.120–115 or 117–116 bc.
Sherk 1969, no. 42; CID 4 119 A (cf. Ferrary in Rousset 2002b).
R22. Letter of L. Calpurnius Piso to the Itanians, from Itanos in Crete, accompanying a
senatus consultum on the dispute between Itanos and Hierapytna, 112 bc.
Sherk 1969, no. 14; Ager 1996, no. 158 (cf. SEG 2, 511).
R23. Letter (fragmentary) of the grammateus of the synhedrion Xenon to the strategoi
(ll. 17–21), preceded by a letter of a Roman magistrate (?) to the Carystians or
Euboians, and followed by a hypomnema, from Carystos, second century bc.
IG xii 9, 5.
R24. Letter of P. Sextilius to the Triccaeans, from Tricca in Thessaly, accompanying a
senatus consultum, second century bc?
Sherk 1969, no. 8.
R25. Letter of someone (a Roman magistrate?) to the Triccaeans, from Tricca in Thessaly,
concerning an arbitration, second century bc.
Sherk 1969, no. 45 (IG ix 2, 301).
R25bis. Letter of someone (a Roman magistrate?) to the magistrates and the people of
Kyparissos, Kyparissos, second–first century bc.
Messenia, SEG 11, 1025; Sherk 1969, no. 46. Cf. Greek list above, no. 53.
R26–R27. Two letters of Q. Mucius Scaevola, respectively to the Sardianoi and to the
Ephesians, concerning (and accompanying) the treaty between Sardis and Ephesus,
from Pergamum, 98–97 or 94–93 bc.
Sherk 1969, no. 47; SEG 38, 1267; Ager 1996, no. 170.
 Appendix 3 381
R28. Letter of C. Cassius to the Nysaeans concerning Chairemon (followed by two letters
of Mithridates to his satrap Leonippos), from Akça (Nysa) in Caria, 88 bc. (A strange
letter, obviously translated from Latin; no final formula.)
Sherk 1969, no. 49 (Syll.3 741).
R29–R30. Letters of Sulla and of the quaestor propraetore Lucullus confirming inviolability
to the temple of Isis and Sarapis at Mopsuestia, from Mopsuestia, probably shortly
after 85 bc and in 86 bc respectively.
SEG 44, 1227; Rigsby 1996, no. 217; cf. SEG 56, 1801.
R31. Letter of the proconsul Q. Oppius to the Plaraseis and Aphrodisieis, 85 bc (but
inscribed in a dossier dated to the second century AD).
Reynolds 1982, no. 3 (InsAph 8.2).
R31bis. Letter of an unidentified city or of a proconsul to Sparta, from Sparta, first century bc.
IG v i, 9: litterae vel urbis cuiusdam vel proconsulis Romani (cf. Greek list above, no. 59).
R32–R33. Letters of L. Cornelius Sulla to the Coans concerning the Dionysiac artists, and
to the Dionysiac artists (followed by a decree of the Senate now lost), from Cos,
respectively c.84 and 81 bc.
Sherk 1969, no. 49; le Guen 2001, no. 56.
R34–R35. Two letters of Sulla to the Stratoniceans, followed by a senatus consultum, from
the temple of Hekate at Lagina in Caria, 81 bc.
Sherk 1969, no. 18; I. Stratonikeia 505.
R36. Letter of L. Cornelius Sulla to the Thasians, accompanying a senatus consultum, from
Thasos, 80 bc.
Sherk 1969, no. 20; SEG 18, 349.
R37–R38. Two letters of Cn. Cornelius Dolabella to the Thasians, discussing the above-
mentioned senatus consultum and referring to other letters sent to the Peparethians
and Skiathians, from Thasos; same stone as R48 below, 80 bc.
Sherk 1969, no. 21.
R39. Letter of the consuls M. Terentius Varro Lucullus and C. Cassius Longinus to the
Oropians, with summary of documents and senatus consultum, from Oropos, 73 bc.
Sherk 1969 no. 23; Epigr. tou Oropou 308, IG vii 413.
R40. Letter of a Roman magistrate (the governor of Asia) to the Mytilenaeans (very
fragmentary), from Mytilene, post 55 bc.
Sherk 1969, no. 51.
R41. Letter of a Roman magistrate to the Milesians, the Ephesians, the Trallianoi, the
Alabandeis, the Mylaseis, the Smyrnaioi, the Pergamenoi, the Sardianoi, and the
Adramyttenoi, from Miletos and Priene (two copies), 51–50 bc.
Sherk 1969, no. 52; SEG 50, 1178; cf. Ferrary 2009: 135–8.
R42. Letter (very fragmentary) of a Roman magistrate to Ilion, from Ilion, first century bc.
Sherk 1969, no. 53; I. Ilion 77.
R43–R44–R45. Three letters of C. Iulius Caesar to the Mytilenaeans, from Mytilene, 48 to
45 bc. (Part of a dossier including also two senatus consulta passed in 25 bc, and a treaty).
Sherk 1969, no. 26.
R46. Letter of C. Iulius Caesar to the Pergamenoi, from Smyrna (part of a dossier which
must have been engraved in many cities), accompanying a decision, post 48 bc
(Pharsalos).
Sherk 1969, no. 54; Rigsby 1996, no. 180.
382 Appendix 3
R47. Letter of P. Servilius Isauricus to the Pergamenoi, concerning the asylia of the
Asklepieion, from Pergamon, 46–44 bc.
Sherk 1969, no. 55; Rigsby 1996, no. 181.
R48. Letter of L. Sestius Quirinalis to the Thasians, from Thasos (part of a dossier inscribed
on the same stone as nos. R37 and R38 above), 44–42 bc.
Sherk 1969, no. 56.
R49. Covering letter of M. Antonius (entirely erased after Actium) accompanying
a ºªæçÅ Æ of Caesar granting the right of asylum to the temple of Artemis in
Sardis, Sardis, 44/43 bc.
SEG 39, 1290; Rigsby 1996, no. 214
R50. Letter of M. Antonius to the koinon Asiae, from Tralles (but probably inscribed in
Smyrna), 42–41 or 33–32 bc.
Sherk 1969, no. 57.
R51. Letter of Octavian to Plarasa/Aphrodisias, from Aphrodisias, 39–38 bc.
Reynolds 1982, no. 6 (Sherk 1969, no. 28A; InsAph 8. 25).
R52–R53–R54. Letters of Octavian to the inhabitants of Rhosos concerning Seleukos, from
Rhosos in Syria, the first one (probably 42 bc) accompanying a decree, the two others
(31–30 bc) independent.
Sherk 1969, no. 58; SEG 54, 1625.
R55. Letter of a Roman magistrate (fragmentary) to the Mylaseis, from Mylasa, post 39 bc.
Sherk 1969, no. 59; I. Mylasa 601.
R56. Letter of Octavian to Stephanos, from Aphrodisias, 39–38 bc.
Reynolds 1982, no. 10 (InsAph 8. 29).
R57. Letter of Stephanos to Plarasa / Aphrodisias, 39-38 bc.
Reynolds 1982, no. 11 (InsAph 8. 30).
R58. Letter of Octavian to Ephesus, from Aphrodisias, 39-38 bc.
Reynolds 1982 no. 12 (InsAph 8. 31).
R59. Letter of Octavian to the Mylaseis, from Mylasa, 31 bc.
Sherk 1969, no. 60; I. Mylasa 602.
R60–R61–R62: group of letters, most of them very fragmentary and of uncertain date, from
Ephesos. Note (a) SEG 43, 768: extremely fragmentary letter (it might actually be the
beginning of the letter SEG 43, 757), addressed to the gerousia of Ephesos, dated by
the letter-shape to the second half of the first century bc; (b) SEG 43, 757: second
century ad copy of a letter of Caesar or Octavian, extremely fragmentary; (c) SEG 43,
758: second century ad copy of letter of Octavian addressed to the council and the
people of Ephesos, 29 bc. Three other letters, of which almost nothing remains, and
whose date is uncertain, were part of this archive (SEG 43, 769–71), as well as a further
eight letters, of imperial date.
R63. Letter (in Latin and with fragmentary Greek translation) of a Vinicius, preceded by a
iussum of Augustus and Agrippa, to the Cumaeans, from Kyme in Asia, 27 bc.
Sherk 1969, no. 61.
R64. Letter of M. Agrippa to the Argive gerousia (the stone has also an inscription in
honour of Alexander of Sicyon), from Argos, 17–16 bc.
Sherk 1969, no. 63.
 Appendix 3 383
R65. Letter of Augustus (very fragmentary; traces of another letter as well?) to the Eresians,
from Eresos, post 15 bc.
Sherk 1969, no. 64.
R66. Letter of C. Norbanus Flaccus to Aizanoi, confirming earlier decisions mentioned in a
letter of Augustus, Aizanoi, between 19 and 10 bc.
MAMA IX 13; Wörrle, Chiron 41 (2011), 357–76.
R67. Letter of P. Fabius Maximus to the koinon of the Greeks of Asia (preserved in Greek
and Latin, both versions very fragmentary), followed by two decrees of the koinon,
from Apameia, Priene, Eumeneia, Dorylaion, and Maeonia, c.9 bc.
Sherk 1969, no. 65, Rhodes with Lewis 1997: 396.
R68. Letter of P. Cornelius Scipio to the Thyatirenoi (possibly followed by a local decree),
from Thyatira, 10–6 bc.
Sherk 1969, no. 66.
R69. Letter of Augustus to the Cnidians (followed by a letter of Hadrian to the Astypa-
laeans) concerning a murder affair, from Astypalaea, second half of 6 bc.
Sherk 1969, no. 67.
R70. Letter of Augustus to the Sardianoi, answering a decree of the city (all part of a dossier
of twelve documents related to a citizen of Sardis called Menogenes), from Sardis,
5 bc.
Sherk 1969, no. 68.
R71. Letter of Cn. Cornelius Lentulus to the Nysaeans, inscribed in the context of the
restoration of all the sacred writings by the strategos of Nysa Artemidoros and part of
a huge dossier, from Nysa, 1 bc.
Sherk 1969, no. 69.
 Bibliography

Achterberg, W., Best, J., Enzler, K., Rietveld, L., and Woudhuizen, F. (2004), The Phaistos
Disc: A Luwian Letter to Nestor (Amsterdam).
Ager, S. (1996), Interstate Arbitrations in the Greek World, 337–90 b.c. (Berkeley).
Ahl, F. M. (1967), ‘Cadmus and the Palm-Leaf Tablets’, AJPh 88, 188–94.
Almagro Basch, M. (1952), Las inscripciones ampuritanas griegas, ibéricas y latinas
(Barcelona).
Alster, B. (1987), ‘A Note on the Uriah Letter in the Sumerian Sargon Legend’, ZA 77,
169–73.
Altman, J. G. (1982), Epistolarity: Approaches to a Form (Columbus, OH).
Amadasi, M. G. (1998), ‘Sulla formazione e diffusione dell’alfabeto’, in G. Bagnasco Gianni
and F. Cordano (eds.), Scritture mediterranee fra IX e VIII sec. a.C. (Milan), 27–51.
Ando, C. (1999), ‘Was Rome a Polis?’, Classical Antiquity 18, 5–34.
Arena, R. (1994), Iscrizioni greche arcaiche di Sicilia e Magna Grecia, iii: Iscrizioni delle
colonie euboiche (Pisa).
Arnott, G. W. (1996a), Alexis: The Fragments (Cambridge).
——(1996b), Menander, ii (Cambridge, MA).
Asheri, D. (1988), Erodoto: Le storie, i: La Lidia e la Persia (Milan).
Asheri, D., and Corcella, A. (2003), Erodoto: Le storie, viii: La vittoria di Temistocle (Milan).
——, and Medaglia, S. (1990), Erodoto: Le storie, iii: La Persia (Milan).
Austin, C., and Olson, S. D. (2004), Aristophanes: Thesmophoriazusae (Oxford).
Austin, M. M. (2006), The Hellenistic World from Alexander to the Roman Conquest, 2nd
edn (Cambridge).
Avezzù, G. (1982) (ed.), Alcidamante: Orazioni e frammenti (Rome).
Awianowicz, B. (2009), ‘A Hellenistic Ostracon from Nikonion’, ZPE 168, 196–8.
——(2011), ‘A New Hellenistic Ostracon from Nikonion’, ZPE 178, 237–9.
Badian, E. (1979), ‘The Name of the Runner’, AJAH 4, 163–6.
——(1987), ‘The Peace of Callias’, JHS 107, 1–39; repr. in From Plataea to Potidaea: Studies
in the History and Historiography of the Pentekontaetia (Baltimore, 1993), 1–72.
——(1991), ‘The King’s Peace’, in M. Flower and M. Toher (eds.), Georgica: Greek Studies
in Honour of George Cawkwell, BICS suppl. 58, 25–48.
Bagnall, R. S., and Cribiore, R. (2006), Women’s Letters from Ancient Egypt, 300 b.c.–a.d.
800 (Ann Arbor).
——, and Derow, P. (2004), The Hellenistic Period: Historical Sources in Translation
(Oxford).
Bakker, E. J. (1993), ‘Activation and Preservation: The Interdependence of Text and
Performance in an Oral Tradition’, Oral Tradition 8, 5–20.
——(2006), ‘Contract and Design: Thucydides’ Writing’, in A. Rengakos and A. Tsakmakis
(eds.), Brill’s Companion to Thucydides (Leiden), 109–29.
Bakker, M. P. de (2007), Speech and Authority in Herodotus’ Histories, diss. Amsterdam.
Barnabò, L. (1977/78), ‘Oracoli come messaggio. Erodoto testimone di una dimensione
orale dei responsi oracolari’, Bollettino dell’Istituto di Filologia Greca Padova 4, 157–74.
Barrett, J. (2002), Staged Narrative: Poetics and the Messenger in Greek Tragedy (Berkeley).
Barrett, W. S. (1964), Euripides: Hippolytos (Oxford).
Bartonek, A., and Buchner, G. (1995), ‘Die ältesten griechischen Inschriften von Pithe-
koussai (2. Hälfte des VIII. bis 1. Hälfte des VI. Jh.)’, Die Sprache 37/2, 129–231.
Bassi, K. (1997), ‘Orality, Masculinity, and the Greek Epic’, Arethusa 30, 315–40.
——(1998), Acting Like Men: Gender, Drama, and Nostalgia in Ancient Greece (Ann Arbor).
386 Bibliography
Bats, M. (1988), ‘La Logique de l’écriture d’une société à l’autre en Gaule méridionale
protohistorique’, Revue archéologique de Narbonnaise 21, 121–48.
——(2010), ‘Une lettre sur plomb à Lattes (Hérault)’, in T. Janin (ed.), Premières données
sur le Ve s. av. n.è. dans la ville de Lattara, Lattara 21 (Lattes), 749–56.
——(2011), ‘Emmêlements de langue et de systèmes graphiques en Gaule méridionale’, in
C. Ruiz Darasse and E. R. Luján (eds.), Contacts linquistiques dans l’Occident méditer-
ranéen antique (Madrid), 197–226.
Bearzot, C. (2003), ‘L’uso dei documenti in Tucidide’, in A. M. Biraschi et al. (eds.), L’uso
dei documenti nella storia antica (Naples), 265–313.
Beck, H. (2013), ‘P. Cornelius Scipio Africanus Maior (232)’, in BNJ <http://referenceworks.
brillonline.com/entries/brill-s-new-jacoby/p-cornelius-scipio-africanus-maior-232-a232>.
Belousov, A. V. and Saprykin, S. Yu. (2013), ‘A letter of Kledikos from Hermonassa’, ZPE
185, 153–60.
Bencivenni, A. (2010), ‘Il re scrive, la città iscrive. La pubblicazione su pietra delle epistole
regie nell’Asia ellenistica’, Studi ellenistici 24, 149–78.
——(2011), ‘ “Massima considerazione”. Forma dell’ordine e immagini del potere nella
corrispondenza di Seleuco IV’, ZPE 176, 139–53.
Bennet, J. (2001), ‘Agency and Bureaucracy: Thoughts on the Nature and Extent of
Administration in Bronze Age Pylos’, in S. Voutsaki and J. Killen (eds.), Economy and
Politics in the Mycenaean Palace States (Cambridge), 25–35.
——(2008), ‘Now You See It; Now You Don’t! The Disappearance of the Linear A Script
on Crete’, in J. Baines, J. Bennet, and S. Houston (eds.), The Disappearance of Writing
Systems: Perspectives on Literacy and Communication (London), 1–29.
Berger-Doer, G. (1990), ‘Iobates’, in LIMC, v (Zurich), 677–81.
Bernabé, A., and Jiménez San Cristóbal, A. I. (2008), Instructions for the Netherworld: The
Orphic Gold Tablets (Leiden).
Bers, V. (1997), Speech in Speech: Studies in Incorporated Oratio Recta in Attic Drama and
Oratory (Lanham, MD).
Bertrand, J.-M. (1985), ‘Formes de discours politique: décrets des cités grecques et
correspondance des rois hellénistiques’, Revue historique de droit français et étranger
63, 469–81; repr. in Du pouvoir dans l’antiquité: mots et réalités, Cahiers du Centre
Gustave Glotz 1 (Geneva, 1990), 101–15.
——(1992), Inscriptions historiques grecques (Paris).
——(1999), De l’écriture à l’oralité: Lectures des ‘Lois’ de Platon (Paris).
——(2006), ‘Reflexions sur les modalités de la correspondance dans les administrations
hellénistiques’, in L. Capdetrey (ed.), La Circulation de l’information dans les états
antiques (Bordeaux), 89–104.
——, and Gruenais, M.-P. (1981), ‘Sur les termes reputés techniques dans les sources
anciennes: l’exemple de topos’, Langage et société 16, 67–82.
Berve, H. (1926), Das Alexanderreich auf prosopographischer Grundlage (Munich).
Bessios, M., Tzifopoulos, Y., and Kotsonas A. (2012),
ŁÅ —ØæÆ I: ¯تæÆç ,
åÆæ ª Æ Æ ŒÆØ   æØŒ   ºÆ  Å ªø  æØŒ ŒÆØ ÆæåÆœŒ ŒæÆ ØŒ Æ
‘ªØ ’ Å
ŁÅ —ØæÆ  Å
ÆŒ Æ (Thessalonica).
Bétant, A.-E. (1843–7), Lexicon Thucydideum, 2 vols. (Geneva).
Bikerman, E. (1930), ‘Beiträge zur antiken Urkundengeschichte III.  0E ıØ and
 Å Æ’, APF 9, 155–82.
Biltcliffe, D. A. W. (1969), ‘P. Oxy. No. 2330 and its Importance for the Study of Nicolaus of
Damascus’, RhM 112, 85–93.
Binder, C. (2008), Plutarchs Vita des Artaxerxes: Ein historisches Kommentar (Berlin).
Bingen, J. (1975), ‘Préambule et promoteur dans le décret attique’, in J. Bingen,
G. Cambrier, and G. Nachtergael, Le Monde grec: Pensée, littérature, histoire, documents.
Mélanges Claire Préaux (Brussels), 470–9.
Black, J. A. (1998), Reading Sumerian Poetry (London).
 Bibliography 387
Black, J. A., Cunningham, G., Robson, E., and Zólyomi, G. (2004), The Literature of Ancient
Sumer (Oxford).
Blass, F. (1887–98), Die attische Beredsamkeit, 3 vols. (Leipzig).
Boardman, J. (1968), Archaic Greek Gems (London).
Boardman, J. (1992), ‘Linos’, in LIMC, vi.1 (Zurich), 290.
Boffo, L. (1995), ‘Ancora una volta sugli “archivi” nel mondo greco: conservazione
e “pubblicazione” epigrafica’, Athenaeum 83, 91–130.
——(2003), ‘Per una storia dell’archiviazione pubblica nel mondo greco’, Dike 6, 5–85.
Bollansée, J. (1998), ‘Andron of Ephesos, FGrHCont 1005’, in Die Fragmente der grie-
chischen Historiker Continued, 4: Biography and Antiquarian Literature, 4A: Biography,
i: The Pre-Hellenistic Period (Leiden), 122–67.
Bonner, R. J. (1905), Evidence in Athenian Courts (Chicago).
Boring, T. A. (1979), Literacy in Ancient Sparta (Leiden).
Bowie, A. M. (2007), Herodotus: Histories, Book VIII (Cambridge).
Bowie, E. (1999), ‘Kleobuline’, in DNP, vi, 576.
Bowman, A. K. (1998), Life and Letters on the Roman Frontier: Vindolanda and its People
(London).
——, and Thomas, J. D. (1983), Vindolanda: The Latin Writing-Tablets (London).
——, and Thomas, J. D. (1994 and 2003), The Vindolanda Writing-Tablets (Tabulae
Vindolandenses II–III) (London).
Bradbury, S. (2004), ‘Libanius’ Letters as Evidence for Travel and Epistolary Networks
among Greek Elites in the Fourth Century’, in L. Ellis and F. L. Kidner (eds.), Travel,
Communication and Geography in Late Antiquity: Sacred and Profane (Aldershot), 73–80.
Bravo, B. (1974), ‘Une lettre sur plomb de Berezan: colonisation et modes de contact dans le
Pont’, DHA 1, 111–87.
——(1977), ‘Remarques sur les assises sociales, les formes d’organisation et la terminologie
du commerce maritime grec à l’époque archaïque’, DHA 3, 1–59.
——(1987), ‘Une tablette magique d’Olbie Pontique, les morts, les héros et les demons’, in
Poikilia: Études offertes à J.-P. Vernant, Paris, 185–218.
——(2000–1), ‘Deux ostraca magiques d’Olbia Pontique et quelques données nouvelles sur
les procédés de la magie destructive’, Talanta 22–3, 149–64.
——(2001), ‘Luoghi di culto nella chora di Olbia Pontica’, in Problemi della chora coloniale
dall’ Occidente al Mar Nero. Atti del quarantesimo convegno di studi sulla Magna Grecia
(Taranto), 221–66.
—— (2007), ‘Testi iniziatici da Olbia Pontica (VI e V sec. a. C.) e osservazioni su orfismo
e religione civica’, Palamedes 2, 55–92.
—— (2011), ‘Tre lettere di mercanti del Mar Nero settentrionale rivisitate. Sui modi
di organizzazione e gli attori del commercio e sulla “giustizia privata” in poleis e emporia’,
Palamedes 6, 37–91.
Bremmer, J. N. (1999), ‘Linos’, in DNP, vii (Stuttgart), 252–3.
Bresson, A., Cocula, A.-M., and Pébarthe, C. (2005) (eds.), L’Écriture publique du pouvoir (Paris).
Briant, P. (1996), Histoire de l’empire perse: De Cyrus à Alexandre (Paris).
——(2003), ‘Histoire et archéologie d’un texte: la Lettre de Darius à Gadatas entre Perses,
Grecs et Romains’, in M. Giorgieri et al. (eds.), Licia e Lidia prima dell’ellenizzazione: Atti
del convegno internazionale, Roma, 11–12 ottobre 1999 (Rome), 107–44.
Brown, T. S. (1981), ‘Aeneas Tacticus, Herodotus and the Ionian Revolt’, Historia 30, 385–93.
Bruce, I. A. F (1967), An Historical Commentary on the Hellenica Oxyrhynchia (Cambridge).
Brugnone, A. (1984), Lilibeo: Testimonianze archeologiche dal IV sec. a.C. al V. d.C.
(Marsala), 158–64.
Buchheim, T. (1989), Gorgias von Leontinoi: Reden, Fragmente und Testimonien (Hamburg).
Burkert, W. (1983), ‘Oriental Myth and Literature in the Iliad’, in R. Hägg (ed.), The Greek
Renaissance of the Eighth Century bc: Tradition and Innovation (Stockholm), 51–6.
388 Bibliography
Calame, C. (1993), ‘Rythme, voix et mémoire de l’écriture en Grèce classique’, in
R. Pretagostini (ed.), Tradizione e innovazione nella cultura greca da Omero all’età
ellenistica: Scritti in onore di Bruno Gentili, 3 vols. (Rome), ii, 785–99.
Calame, C. (1995), The Craft of Poetic Speech in Ancient Greece (Ithaca, NY) (rev. and expanded
edn of Le Récit en Grèce ancienne: Énonciations et représentations de poètes (Paris, 1986)).
——(2000), ‘Orphik, Orphische Dichtung’, in DNP, ix (Stuttgart), 58–70.
——(2005), Masks of Authority: Fiction and Pragmatics in Ancient Greek Poetics (Ithaca,
NY).
——(2008), ‘Les Lamelles funéraires d’or: textes pseudo-orphiques et pratiques rituelles’,
Kernos 21, 299–311.
——(2009), Poetic and Performative Memory in Ancient Greece: Heroic Reference and
Ritual Gestures in Time and Space (Washington, DC) (English translation of Pratiques
poétiques de la mémoire: Représentations de l’espace-temps en Grèce ancienne (Paris
2006)).
Canali de Rossi, F. (2002), Iscrizioni storiche ellenistiche, iii: Decreti per ambasciatori greci al
senato (Rome).
Canfora, L. (1990), ‘Trattati in Tucidide’, in L. Canfora, M. Liverani, and C. Zaccagnini
(eds.), I trattati nel mondo antico: Forma ideologia funzione (Rome), 193–216.
——(1991), Demostene: Le Filippiche e altri discorsi, repr. with a supplementary bibliog-
raphy (Milan).
Cappelletto, P. (2003), I frammenti di Mnasea: Introduzione testo e commento (Milan).
Carastro, M. (2006), La Cité des mages: Penser la magie en Grèce ancienne (Grenoble).
Carey, C. (1994), ‘ “Artless Proofs” in Aristotle and the Orators’, BICS 39, 95–106; repr. in
E. Carawan (ed.), Oxford Readings in the Attic Orators (Oxford, 2007), 229–46.
——(2000), Aeschines (Austin, TX).
——(2007), Lysiae Orationes cum fragmentis (Oxford).
Caroli, M. (2012), ‘Il bibliagraphos di Cratino tra libri e decreti assembleari’, ZPE 182,
95–108.
Carraro, F. (2007), ‘The “Speaking Objects” of Archaic Greece: Writing and Speech in the
First Complete Alphabetic Documents’, in K. Lomas et al. (ed.), Literacy and the State in
the Ancient Mediterranean (London), 65–80.
Cartledge, P. (1978), ‘Literacy and the Spartan Oligarchy’, JHS 98, 25–37.
Cassio, A. C. (1998), ‘Epica greca e scrittura tra VIII e VII secolo a.C.: madrepatria e colonie
d’Occidente’, in G. Bagnasco Gianni and F. Cordano (eds.), Scritture mediterranee tra il
IX e il VII secolo a.C. (Milan), 67–84.
Cawkwell, G. L. (1969), ‘The Crowning of Demosthenes’, CQ, ns 19, 163–80.
Ceccarelli, P. (2002), ‘Message épistolaire et message oral au Proche Orient ancien et en Grèce
archaïque’, in E. Gavoille and L. Nadjo (eds.), Epistulae antiquae, ii (Louvain), 11–26.
——(2004), ‘Écriture féminine, écriture épistolaire, parole des rhétheurs: à propos du fr.
194 K.–A. de la Sappho d’Antiphane’, in L. Nadjo and E. Gavoille (eds.), Epistulae
antiquae, iii (Louvain), 11–32.
——(2005a), ‘Forme di comunicazione e ideologia della polis: discorso in assemblea,
decreto, ed epistola ufficiale’, in U. Bultrighini (ed.), Democrazia e antidemocrazia nel
mondo greco (Alessandria), 345–69.
——(2005b), ‘Messaggio scritto e messaggio orale: strategie narrative erodotee’, in
M. Giangiulio (ed.), Erodoto e il ‘modello erodoteo’: Formazione e trasmissione delle
tradizioni storiche in Grecia (Trento), 13–60.
——(2005c), ‘L’Agamemnon d’Eschyle et la vengeance d’Harpage. A propos d’Hérodote (I
107–129)’, in A. Kolde, A. Lukinovich, and A.-L. Rey (eds.), Koruphaioi andri: Mélanges
offerts à André Hurst (Geneva), 69–78.
——(2010), ‘Changing Contexts: Tragedy in the Civic and Cultural Life of Hellenistic
City-States’, in I. Gildenhard and M. Revermann (eds.), Beyond the Fifth Century:
 Bibliography 389
Interactions with Greek Tragedy from the Fourth Century bce to the Middle Ages
(Berlin), 99–150.
Ceccarelli, P. (forthcoming a), ‘Image and Communication in the Seleucid Kingdom. The
King, the Court, the Cities’, in A. Erskine, L. Llewellyn-Jones, and S. Wallace (eds.), The
Hellenistic Court (Swansea).
——(forthcoming b), ‘Letters and Decrees, or Political Styles in the Ancient World’, in
P. Ceccarelli, L. Doering, T. Foegen, and I. Gildenhard (eds.), Configuring
Communities.
——(forthcoming (c)), ‘Charon of Lampsakos (262)’, in BNJ <http://referenceworks.
brillonline.com/entries/brill-s-new-jacoby/charon-of-lampsakos-262-a262>.
Ceka, N. (1987), ‘Le Koinon des Bylliones’, in P. Cabanes (ed.), L’Illyrie méridionale et
l’Épire dans l’Antiquité: Actes du Colloque international de Clermont-Ferrand, 22–25
octobre 1984 (Clermont-Ferrand), 135–49.
Celentano, M. S. (1991), ‘L’epistola laconica: dalla concisione esemplare all’esiguità iperbo-
lica’, in A. Pennacini (ed.), Retorica della comunicazione nelle letterature classiche
(Bologna), 109–29.
Champion, C. B. (2013a), ‘Timaios (566)’, in BNJ <http://referenceworks.brillonline.com/
entries/brill-s-new-jacoby/timaios-566-a566>.
——(2013b), ‘Timonides of Leukas (561)’, in BNJ <http://referenceworks.brillonline.com/
entries/brill-s-new-jacoby/timonides-of-leukas-561-a561>.
Chaniotis, A. (1999), ‘Empfängerformular und Urkundenfälschung: Bemerkungen zum Urkun-
dendossier von Magnesia am Maeander’, in R. G. Khoury (ed.), Urkunden und Urkundenfor-
mulare im klassischen Altertum und in den orientalischen Kulturen (Heidelberg), 51–69.
Chantraine, P. (1972), ‘À propos du nom des Phéniciens et des noms de la pourpre’, Studii
Clasice 14, 7–15.
——(1999), Dictionnaire étymologique de la langue grecque: Histoire des mots, new edn.
with supplement (Paris).
Chantry, M. (1994), Scholia vetera in Aristophanis Plutum (Groningen).
Cipolla, P. (2003), Poeti minori del dramma satiresco: Testo critico, traduzione e commento
(Amsterdam).
Clavaud, R. (1987), Démosthène: Lettres et fragments (Paris).
Clay, J. C. (2011), Homer’s Trojan Theater: Space, Vision, and Memory in the Iliad (Cambridge).
Collard, C. (1975), Euripides: Supplices, 2 vols. (Groningen).
——(1991), Euripides: Hecuba (Warminster).
——, and Cropp, M. (2008), Euripides: Fragments, i: Aegeus–Meleager, ii: Oedipus–Chry-
sippus; Other Fragments, 2 vols. (Cambridge, MA).
Collins, R. F. (2010), ‘A Significant Decade: The Trajectory of the Hellenistic Epistolary
Thanksgiving’, in S. E. Porter and S. A. Adams (eds.), Paul and the Ancient Letter Form,
(Leiden), 159–84.
Connor, R. W. (1984), Thucydides (Princeton).
Conti Bizzarro, F. (1999), Poetica e critica letteraria nei frammenti dei poeti comici greci
(Naples).
Cooper, J. S. (1993), ‘Paradigm and Propaganda. The Dynasty of Akkade in the 21st
Century’, in M. Liverani (ed.), Akkad, The First World Empire: Structure, Ideology,
Traditions (Padua), 11–23.
——(2004), ‘Babylonian Beginnings: The Origin of the Cuneiform Writing System in
Comparative Perspective’, in S. D. Houston (ed.), The First Writing: Script Invention as
History and Process (Cambridge), 71–99.
——, and Heimpel, W. (1983), ‘The Sumerian Sargon Legend’, in J. M. Sasson (ed.), Studies in
Literature from the Ancient Near East . . . dedicated to Samuel Noah Kramer, JAOS 103/1, 67–82.
Corcella, A. (1985), ‘Erodoto VII 239: una “interpolazione d’autore”?’, ASNP, 3rd ser., 15/2,
313–491.
390 Bibliography
Corcella, A. (1986), ‘Dionisio Skytobrachion, i phoinikeia, e l’«alfabeto pelasgico»: per una
corretta interpretazione di Diodoro III.67.1 (FgrHist 32 F 8)’, AAT 120, 41–82.
——(1996a), Tucidide: La disfatta a Siracusa (Storie VI–VII) (Venice).
——(1996b), ‘Ecateo di Mileto così dice’, Quaderni di storia 43, 295–301.
Cortassa, G., and Culasso Gastaldi, E. (1990), Le lettere di Temistocle, i: Edizione critica,
traduzione, note testuali e indici; ii: Il problema storico. Il testimone e la tradizione, 2 vols.
(Padua).
Costa, C. D. N. (2002), Greek Fictional Letters: A Selection with Introduction, Translation,
and Commentary (Oxford).
Coupry, J. (1968), ‘Olbia de Ligurie’, Revue d’études ligures 34, 237–46.
——(1971), ‘Apports d’Olbia de Provence à l’épigraphie grecque d’Occident’, Acts of the
Fifth International Congress of Greek and Latin Epigraphy (Oxford), 141–7.
Crane, G. (1996), The Blinded Eye: Thucydides and the New Written Word (Lanham, MD).
Cribiore, R. (2001), ‘Windows on a Woman’s World. Some Letters from Roman Egypt’, in
A. Lardinois and L. McClure (eds.), Making Silence Speak: Women’s Voices in Greek
Literature and Society (Princeton), 223–39.
Cropp, M. J. (2000) (ed.), Euripides: Iphigenia in Tauris (Warminster).
Crowther, C. V. (1999), ‘Aus der Arbeit der Inscriptiones Graecae. 4, Koan Decrees for
Foreign Judges’, Chiron 29, 251–319.
Csapo, E., Johnston, A. W., and Geagan, D. (2000), ‘The Iron Age Inscriptions’, in
J. W. Shaw and M. C. Shaw (eds.), Kommos IV: The Greek Sanctuary (Princeton),
101–34.
Cugusi, P. (1983), Evoluzione e forme dell’epistolografia latina nella tarda repubblica e nei
primi due secoli dell’impero (Rome).
Curbera, J. (1999), ‘Defixiones’, in Sicilia Epigraphica, ASNP ser. 4, Quaderni 1, 159–85.
Curty, O. (1995), Les Parentés légendaires entre cités grecques (Geneva).
Dalby, A. (2000), ‘Lynceus and the Anecdotists’, in D. Braund and J. Wilkins (eds.),
Athenaeus and his World: Reading Greek Culture in the Roman Empire (Exeter), 372–94.
D’Alessio, G. B. (2004), ‘Past Future and Present Past: Temporal Deixis in Greek Archaic
Lyric’, Arethusa 37, 267–94.
Dana, M. (2004), ‘Lettre sur plomb d’Apatorios à Leanax. Un document archaique d’Olbia
du Pont’, ZPE 148, 1–14.
——(2007), ‘Lettres grecques dialectales nord-pontiques (sauf IGDOP 23–26)’, REA 109/1,
67–97.
——(2009), ‘Alphabets et exercises scolaires dans deux cités du Pont ouest: Istros et Tyras’,
ZPE 171, 71–82.
D’Angour, A. J. (1999), ‘Archinus, Eucleides and the Reform of the Athenian Alphabet’,
BICS 43: 109–30.
Darbo-Peschanski, C. (1987), Le Discours du particulier: Essai sur l’enquête hérodotéenne
(Paris).
Daviault, A. (1981), Comoedia togata: Fragments (Paris).
Davies, J. (2005), ‘The Origins of the Inscribed Greek Stela’, in P. Bienkowski, C. B. Mee,
and E. A. Slater (eds.), Writing and Ancient Near East Society: Essays in Honor of Alan
Millard (New York), 283–300.
Davies, M. (1991), Sophocles: Trachiniae (Oxford).
Day, J. W. (2010), Archaic Greek Epigram and Dedication: Representation and Reperfor-
mance (Cambridge).
Decourt, J.-C. (1993), ‘ÆEæ ŒÆd ªÆØ: à propos de quelques inscriptions lyonnaises ,
RPh 67, 237–50.
——(2004), Inscriptions grecques de la France (IGF ) (Lyons).
Deffner, A. (1933), Die Rede bei Herodot und ihre Weiterbildung bei Thukydides, diss.
Munich.
 Bibliography 391
Denniston, J. D., and Page, D. (1957), Aeschylus: Agamemnon (Oxford).
De Puma, R. D. (2001), ‘An Etruscan Mirror with the Prophesying Head of Orpheus’,
Record of the Art Museum, Princeton University 60, 18–29.
De Sanctis, G. (1893), ‘Contributi alla storia ateniese dalla guerra lamiaca alla guerra
cremonidea’, in G. Beloch (ed.), Studi di storia antica, ii (Rome), 3–62; repr. in G. De
Sanctis, Scritti minori, i (Rome, 1966), 248–302.
Detienne, M. (1988a) (ed.), Les Savoirs de l’écriture: En Grèce ancienne (Lille).
——(1988b), ‘L’Espace de la publicité: ses opérateurs intellectuels dans la cité’, in
M. Detienne (ed.), Les Savoirs de l’écriture: En Grèce ancienne (Lille), 73–81.
——(1989), L’Écriture d’Orphée (Paris) (the chapter ‘Une écriture inventive. La voix
d’Orphée, les jeux de Palamède’ was first published as an article in Critique 475
(1986), 1125–34).
Dettori, E. (1999), ‘Æıƪ- / ÆıŪ-, una iscrizione e alcune forme di ¼ªı Ø’, in A. Cassio
(ed.), Kata diálekton: Atti del terzo congresso internazionale di dialettologia greca
(Naples), 279–317.
Dewald, C. (1999), ‘The Figured Stage: Focalizing the Initial Narratives of Herodotus and
Thucydides’, in Th. M. Falkner, N. Felson, and D. Konstan (eds.), Contextualizing
Classics: Ideology, Performance, Dialogue (Lanham, MD), 221–52.
Dickey, E. (1996), Greek Forms of Address: From Herodotus to Lucian (Oxford).
Dickie, M. W. (1995), ‘The Dionysiac Mysteries in Pella’, ZPE 109, 81–6.
Diggle, J. (2004), Theophrastus: Characters, ed. with Introduction, Translation, and Com-
mentary (Cambridge).
Dillery, J. (1995), Xenophon and the History of his Time (London).
Di Marco, M. (2009), ‘Un epigramma-indovinello su tavoletta cerata (T. British Museum
inv. 29527 = P.Lond. Lit. 63 = Pack2 1765 = LDAB 5055)’, ZPE 169, 84–6.
Dobrov, G. (1993), ‘The Tragic and the Comic Tereus’, AJPh 114, 189–234.
Dorati, M. (1995), ‘Ctesia falsario?’, QS 41, 33–52.
Dornseiff, F. (1925), Das Alphabet in Mystik und Magie, 2nd edn (Leipzig).
Dover, K. J. (1968), Aristophanes: Clouds (Oxford).
——(1981), ‘The Language of Classical Attic Documentary Inscriptions’, TPhS 79/1, 1–14.
——(1993), Aristophanes: Frogs (Oxford).
Dow, S. (1936), ‘New Kinds of Evidence for Dating Polyeuktos’, AJA 40, 57–70.
Draaisma, D. (2000), Metaphors of Memory: A History of Ideas about the Mind
(Cambridge).
Drews, R. (1970), ‘Herodotus’ Other Logoi’, AJPh 91, 181–91.
—— (1973), The Greek Accounts of Eastern History (Cambridge, MA).
Dubois, L. (1989), Inscriptions grecques dialectales de Sicile (Rome).
——(1996), Inscriptions grecques dialectales d’Olbia du Pont (Geneva).
Duhoux, Y. (2000), ‘How Not to Decipher the Phaistos Disc’, AJA 104, 597–600.
Dunant, C. and Pouilloux, J. (1958), Recherches sur l’histoire et les cultes de Thasos, ii (Paris).
Dunbar, N. (1995), Aristophanes: Birds (Oxford).
Dunst, G. (1969), ‘Ein griechisches Tontäfelchen von der Küste bei Ampurias’, Madrider
Mitteilungen, 10, 146–54.
Easterling, P. (1985), ‘Anachronism in Greek Tragedy’, JHS 195, 1–10.
Edmonds, J. M. (1959), The Fragments of Attic Comedy, ii (Leiden).
Edmunds, L. (2008), ‘Deixis in Ancient Greek and Latin Literature: Historical Introduction and
State of the Question’, in Philologia antiqua: An International Journal of Classics 1, 67–98.
Edwards, R. B. (1979), Kadmos, the Phoenician (Amsterdam).
Eidinow, E. (2007), Oracles, Curses, and Risk among the Ancient Greeks (Oxford).
——, and Taylor, C. (2010), ‘Lead-Letter Days: Writing, Communication and Crisis in the
Ancient World’, CQ, ns 60, 30–62.
392 Bibliography
Eisenberg, J. M. (2008), ‘The Phaistos Disk: One Hundred Year Old Hoax?’, Minerva 19/4,
9–24.
Elvers, K.-L. (1994), ‘Der “Eid der Berenike und ihrer Söhne”: Eine Edition von IBulg III 2,
1731’, Chiron 24, 241–66.
Engel, D. (2000), review of A. W. Nightingale, Genres in Dialogue: Plato and the Construct
of Philosophy, AJPh 121, 316–20.
Englund, R. K. (1998), ‘Texts from the Late Uruk Period’, in J. Bauer, R. Englund, and
M. Krebernik (eds.), Mesopotamien: Späturuk-Zeit und frühdynastische Zeit (Freiburg),
15–233.
Erbse, H. (1992), Studien zum Verständnis Herodots (Berlin).
Exler, F. J. (1923), The Form of the Ancient Greek Letter of the Epistolary Papyri (Washing-
ton, DC).
Fain, G. L. (2006), ‘Apostrophe and SF˙GS in the Theognidean Sylloge’, CQ, ns 56/1,
301–4.
Falcetto, R. (2002), Il Palamede di Euripide: Edizione e commento dei frammenti
(Alessandria).
Fantuzzi, M. (1984), ‘Gli Iºº ªÆ ªæ Æ Æ di Crizia’, QS 19, 221–7.
Faraguna, M. (2006), ‘Gli archivi e la polis (problemi nuovi e vecchi alla luce di alcuni
recenti documenti)’, in L. Capdetrey and J. Nelis-Clément (eds.), La Circulation de
l’information dans les états antiques (Bordeaux), 53–71.
——(2007), ‘Tra oralità e scrittura: diritto e forme della comunicazione dai poemi omerici
a Teofrasto’, Etica & Politica 9, 75–111.
—— (2009), Oralità e scrittura nella prassi giudiziaria ateniese tra V e IV sec. a. C.’, in E. M.
Harris and G. Thür (eds.), Symposion 2007 (Vienna), 63–82.
——(2011), ‘Legislazione e scrittura nella Grecia arcaica e classica’, ZPE 171, 1–20.
Faraone, C. A. (1991), ‘The Agonistic Context of Early Greek Binding Spells’, in
C. A. Faraone and D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion
(Oxford), 3–32.
Fehling, D. (1989), Herodotus and his ‘Sources’: Citation, Invention and Narrative Art
(Leeds) (rev. trans. of Die Quellenangaben bei Herodot: Studien zur Erzählkunst Herodots
(Berlin, 1971)).
Ferlauto, F. (1990a), ‘Prometeo e le grammaton suntheseis’, BollClass 11, 164–77.
——(1990b), ‘L’invenzione dei ªæ Æ Æ Iºº ªÆ secondo Critias (Athen. I 28 B–C)’,
BollClass 11, 178–83.
Ferrari, F. (1988), Euripide: Ifigenia in Tauride. Ifigenia in Aulide (Milan).
——(2007), review of F. Graf and S. Iles Johnston, Ritual Texts for the Afterlife: Orpheus
and the Bacchic Gold Tablets (London, 2007), in BMCR 10, 15.
——(2010), ‘Oral Bricolage and Ritual Context in the Golden Tablets’, in A. Lardinois,
J. Blok, and M. G. M. van der Poel (eds.), Sacred Words: Orality, Literacy and Religion
(Leiden), 205–16.
——, and Prauscello, L. (2007), ‘Demeter Chthonia and the Mountain Mother in a New
Gold Tablet from Magoula Mati’, ZPE 162, 193–202.
Ferrary, J.-L. (2009), ‘After the Embassy to Rome: Publication and Implementation’, in
C. Eilers (ed.), Diplomats and Diplomacy in the Roman World (Leiden).
Finogenova, S. I. (2003), ‘Hermonassa’, in V. Grammenos and E. K. Petropoulos (eds.),
Ancient Greek Colonies in the Black Sea, ii (Thessalonica), 1007–45.
Fitzpatrick, D. (2001), ‘Sophocles’ Tereus’, CQ, ns 51, 90–101.
Flower, H. I. (1991), ‘Herodotus and Delphic Traditions about Croesus’, in M. A. Flower and
M. Toher (eds.), Georgica: Greek Studies in Honour of George Cawkwell (London), 57–77.
Flower, M. A. (1997), Theopompus of Chios: History and Rhetoric in the Fourth Century bc,
2nd rev. edn with addenda (Oxford).
——, and Marincola, J. (2002), Herodotus: Histories, Book IX (Cambridge).
 Bibliography 393
Fögen, T. (2004), ‘Gender-Specific Communication in Graeco-Roman Antiquity: With a
Research Bibliography’, Historiographia Linguistica 31, 199–276.
Foley, H. (1985), Ritual Irony: Poetry and Sacrifice in Euripides (Ithaca, NY).
Ford, A. (2002), The Origins of Criticism: Literary Culture and Poetic Theory in Classical
Greece (Princeton).
Fowler, H. N. (1888), Commentary on Thucydides Book 5 (Boston, MA).
Fowler, R. (2001), ‘Early History and Literacy’, in N. Luraghi (ed.), The Historian’s Craft in
the Age of Herodotus (Oxford), 95–115.
Fraenkel, E. (1950), Aeschylus: Agamemnon, 3 vols. (Oxford).
——(1955), ‘Ein Motiv aus Euripides in einer Szene der Neuen Komödie’, Studi in onore di
Ugo Enrico Paoli (Florence), 293–304.
Fried, L. S. (2004), The Priest and the Great King: Temple-Palace Relations in the Persian
Empire (Winona Lake, IN).
Friis-Johansen, H., and Whittle, E. W. (1980), Aeschylus: The Suppliants, 3 vols. (Copenhagen).
Fritz, K. von (1967), Die griechische Geschichtsschreibung, i (Berlin).
Frost, F. J. (1979), ‘The Dubious Origins of the “Marathon” ’, AJAH 4, 159–63.
Funck, B. (1996), ‘ “König Perserfreund”. Die Seleukiden in der Sicht ihrer Nachbarn’, in
B. Funck (ed.), Hellenismus: Beiträge zur Erforschung von Akkulturation und politischer
Ordnung in den Staaten des hellenistischen Zeitalters (Tübingen), 195–215.
Gagarin, M. (2009), ‘Response to Michele Faraguna’, in E. M. Harris and G. Thür (eds.),
Symposion 2007 (Vienna), 83–6.
Gager, J. G. (1992), Curse Tablets and Binding Spells from the Ancient World (New York).
Gagné, R. (2013), ‘Dancing Letters: the Alphabetic Tragedy of Callias’, in R. Gagné and
M. Hopman (eds.), Choral Mediations in Greek Tragedy (Cambridge), 297–316.
Gantz, T. (1993), Early Greek Myth: A Guide to Literary and Artistic Sources (Baltimore).
Garcea, A. (2002), ‘César et l’alphabet: un fragment du De Analogia (frg. 4 p. 148 Funaioli =
5 p. 179 s. Klotz)’, Histoire Épistémologie Langage 24/2, 147–64.
Garezou, M.-X. (1994), ‘Orpheus’, in LIMC, vii.1 (Zurich), 81–105.
——(2009), ‘Orpheus’, in LIMC Supplementum, i (Düsseldorf), 399–405.
Gärtner, H. A. (2001), ‘Rätsel’, in DNP, x, 754–8.
Garvie, A. F. (2009), Aeschylus: Persae. With Introduction and Commentary (Oxford).
Gasparri, C. (1986), ‘Dionysos’, in LIMC, iii.1 (Zurich), 414–514.
Gauger, J.-D. (2000), Authentizität und Methode: Untersuchungen zum historischen Wert
des persisch-griechischen Herrscherbriefs in literarischer Tradition (Hamburg).
Gauthier, Ph. (1985), Les Cités grecques et leurs bienfaiteurs (IV–I siècle av. J.-C): Contribu-
tion à l’histoire des institutions (Paris).
Gauthier, P. (1986), ‘Histoire grecque et monarchie’, in Les Monarchies (Paris), 43–62; repr.
in Ph. Gauthier, Études d’histoire grecque et d’institutions grecques: Choix d’écrits
(Geneva, 2011), 273–94.
Gavoille, L. (2000), ‘Epistula et litterae: étude de synonymie’, in L. Nadjo and É. Gavoille
(eds.), Epistulae antiquae, i (Louvain), 15–38.
Gavrilov, A. K. (1997), ‘Techniques of Reading in Classical Antiquity’, CQ, ns 47/1, 56–73.
Gazzano, F. (2002), ‘La diplomazia nelle Storie di Erodoto. Figure, temi, problemi’, in
L. R. Cresci, F. Gazzano, and D. P. Orsi, La retorica della diplomazia nella Grecia antica e
a Bisanzio (Rome), 7–67.
Gera, D. L. (1997), Warrior Women: The Anonymous Tractatus de mulieribus (Leiden).
——(2003), Ancient Greek Ideas on Speech, Language, and Civilization (Oxford).
Gerhard, G. A. (1905), ‘Untersuchungen zur Geschichte des griechischen Briefes I’, Philo-
logus 64, 27–65.
Ghinatti, F. (2004), ‘Ancora sulle origini dell’alfabeto greco’, Simblos: Scritti di storia antica
4, 25–48.
Giannattasio Andria, R. (2003), ‘Novità testuali su POxy 2330 (Ctesias Cnidius, FGrHist
688 F 8b)’, ZPE 144, 15–18.
394 Bibliography
Gibson, R. K., and Morrison, A. D. (2007), ‘Introduction: What is a Letter?’, in R. Morello
and A. D. Morrison (eds.), Ancient Letters: Classical and Late Antique Epistolography
(Oxford), 1–16.
Gigon, O. (1980), ‘Der Brief in der griechischen Philosophie’, Didactica Classica Gandensia
20, 117–32.
Glassner, J.-J. (2003), The Invention of Cuneiform, Baltimore (first published as: Écrire à
Sumer: L’Invention du cunéiforme (Paris, 2000)).
Godley, A. D. (1920–4) (trans.), Herodotus, 4 vols. (Cambridge, MA).
Goff, B. E. (1990), The Noose of Words: Readings of Desire, Violence and Language in
Euripides’ Hippolytus (Cambridge).
Goldhill, S. (1984), Language, Sexuality, Narrative: The Oresteia (Cambridge).
——(1987), ‘The Great Dionysia and Civic Ideology’, JHS 107, 58–76.
——(2000), ‘Civic Ideology and the Problem of Difference: The Politics of Aeschylean
Tragedy, Once Again’, JHS 120, 34–56.
——(2002), The Invention of Prose (Oxford).
Goldstein, J. A. (1968), The Letters of Demosthenes (New York).
Gomme, A. W. (1956), A Historical Commentary on Thucydides, iii: Books IV–V 24
(Oxford).
Gomme, A. W. (1959), A Historical Commentary on Thucydides, i: Book I (Oxford)
(corrected reprint of 1944 edn).
——, Andrewes, A., and Dover, K. J. (1970), A Historical Commentary on Thucydides, iv:
Books V 25–VII (Oxford).
—— —— and ——(1981), A Historical Commentary on Thucydides, v: Book VIII
(Oxford).
Goody, J., and Watt, I. (1963), ‘The Consequences of Literacy’, Comparative Studies in
Society and History 5, 304–45; repr. in J. Goody (ed.), Literacy in Traditional Societies
(Cambridge, 1968), 27–68 [from which I quote].
Gordon, R. (1999), ‘ “What’s in a List?” Listing in Greek and Graeco-Roman Malign
Magical Texts’, in D. R. Jordan, H. Montgomery, and E. Thomassen (eds.), The World
of Ancient Magic (Bergen), 239–77.
Görgemanns, H. (1997a), ‘Epistel’, in DNP, iii, 1161–4.
——(1997b), ‘Epistolographie’, in DNP, iii, 1166–9.
Grabbe, L. L. (2004), A History of the Jews and Judaism in the Second Temple Period,
i (London).
Graf, F. (1997), Magic in the Ancient World (Cambridge, MA) (English translation of La
Magie dans l’antiquité gréco-romaine: idéologie et pratique (Paris, 1994)).
——, and Johnston, S. I. (2007), Ritual Texts for the Afterlife: Orpheus and the Bacchic Gold
Tablets (London).
Grassl, H. (1972), ‘Herodot und die griechische Schrift’, Hermes 100, 169–75.
Grayson, A. K. (1975), Assyrian and Babylonian Chronicles, Texts from Cuneiform Sources
5 (New York).
Graziosi, B., and Haubold, J. (2010) (eds.), Homer: Iliad, Book vi (Cambridge).
Green, J. R. (1994), Theatre in Ancient Greek Society (London).
——(1995), ‘Oral Tragedies? A Question from St. Petersburg’, QUCC, ns 51/3, 77–86.
Green, K. (1992), ‘Deixis and the Poetic Persona’, Language and Literature 1/2,
121–33.
Green, M. W. (1981), ‘The Construction and Implementation of the Cuneiform Writing
System’, Visible Language 15/4, 345–72.
Grethlein, J. (2001), ‘Aesch. Suppl. 946–949’, Lexis 19, 175–82.
Griffin, J. (1998), ‘The Social Function of Attic Tragedy’, CQ, ns 48, 39–61.
Griffiths, E. M. (2007), ‘Fighting the Future: Euripidean Letters and Thucydides’ Athens’, in
C. Cooper (ed.), Politics of Orality (Leiden), 277–91.
 Bibliography 395
Grotta, C. (2008), ‘Le iscrizioni’, in N. Allegro (ed.), HIMERA v. L’abitato. Isolato ii. I
blocchi 1-4 della zona 1 (Palermo, 2008), 255–72.
Gruen, E. S. (2010), Rethinking the Other in Antiquity (Princeton).
Guarducci, M. (1969), Epigrafia greca, ii (Rome).
——(1974), Epigrafia greca iii (Rome).
Hall, E. (1996), Aeschylus: Persians, Edited with an Introduction, Translation and Commen-
tary (Warminster).
——(2006), The Theatrical Cast of Athens: Interactions between Ancient Greek Drama and
Society (Oxford).
Hall, J. (1997), Ethnic Identity in Greek Antiquity (Cambridge).
——(2002), Hellenicity: Between Ethnicity and Culture (Chicago).
Halla-aho, H. (2009), The Non-Literary Latin Letters: A Study of their Syntax and Pragmat-
ics (Helsinki).
Halleran, M. R. (1995), Euripides: Hippolytus (Warminster).
Hamon, P. (2005), ‘Le Conseil et la participation des citoyens: les mutations de la basse
époque hellénistique’, in P. Fröhlich and C. Müller (eds.), Citoyenneté et participation à
la basse époque hellénistique (Geneva), 121–44.
Hansen, M. H. (1989), The Athenian Ecclesia, ii (Copenhagen).
Harding, P. (2006), Didymos: On Demosthenes (Oxford).
Harris, E. M. (1995), Aeschines and Athenian Politics (New York).
——(2004), ‘Notes on a Lead Letter from the Athenian Agora’, HSCP 102, 157–70.
——(2006), ‘Antigone the Lawyer, or the Ambiguities of Nomos’, in Democracy and the
Rule of Law in Classical Athens: Essays on Law, Society and Politics (Cambridge), 41–80.
Harris, R. (1986), The Origin of Writing (London).
Harris, W. V. (1989), Ancient Literacy (Cambridge, MA).
Harrison, A. R. W. (1971), The Law of Athens, ii: Procedure (Oxford).
Hartog, F. (1980), Le Miroir d’Hérodote: Essai sur la représentation de l’autre (Paris).
Harvey, F. D. (1985), ‘Dona Ferentes: Some Aspects of Bribery in Greek Politics’, in P. A. Cartledge
and F. D. Harvey (eds.), Crux: Essays Presented to G. E. M. de Ste-Croix (London), 76–113.
——(2007), ‘ “Help! I’m Dying here”: A Letter from a Slave’, ZPE 163, 49–50.
Hatzopoulos, M. B. (2006), La Macédoine: géographie historique, langue, cultes et croyances,
institutions (Paris).
Havelock, E. A. (1982), The Literate Revolution in Greece and its Cultural Consequences
(Princeton).
——(1985), ‘The Oral Composition of Greek Drama’, in B. Gentili and G. Paioni (eds.),
Oralità, cultura, letteratura, discorso: Atti del convegno internazionale, Urbino 1980
(Rome), 199–227.
Head, P. M. (2009a), ‘Named Letter-Carriers among the Oxyrhynchus Papyri’, Journal for
the Study of the New Testament 31, 279–99.
——(2009b), ‘Letter-Carriers in the Ancient Jewish Epistolary Material’, in C. A. Evans and
H. D. Zacharias (eds.), Jewish and Christian Scripture as Artifact and Canon (London), 203–19.
Hedrick, Ch. W., Jr. (1994), ‘Reading, Writing, and Democracy’, in S. Hornblower and
R. Osborne (eds.), Ritual, Finance, Politics: Athenian Democratic Accounts presented to
David Lewis (Oxford), 157–74.
——(1999), ‘Democracy and the Athenian Epigraphical Habit’, Hesperia 68/3, 387–439.
——(2000), ‘For Anyone who Wishes to See’, in R. Frakes and D. Toye (eds.), The Dance of
Hippokleides: A Festschrift for Frank J. Frost, Ancient World 31, 127–35.
Heinimann, F. (1945), Nomos und Physis (Basle).
——(1975), ‘Mass—Gewicht—Zahl’, MH 32, 183–96.
Heinze, T. (1999), ‘Musaios’, in DNP, vii (Stuttgart), 501–3.
Henderson, J. (1998–2008) (ed. and trans.), Aristophanes, i: Acharnians, Knights (1998);
ii: Clouds, Wasps, Peace (1998); iii: Birds, Lysistrata, Women at the Thesmophoria
(2000); iv: Frogs, Assemblywomen, Wealth (2002); v: Fragments (2008) (Cambridge, MA).
Heni, R. (1977), Die Gespräche bei Herodot (Heilbronn).
396 Bibliography
Henrichs, A. (2003a), ‘Writing Religion: Inscribed Texts, Ritual Authority, and the Reli-
gious Discourse of the Polis’, in H. Yunis (ed.), Written Texts and the Rise of Literate
Culture in Ancient Greece (Cambridge), 38–58.
——(2003b), ‘ “Hieroi Logoi” and “Hierai Bibloi”: The (Un)Written Margins of the Sacred
in Ancient Greece”, HSCP 101, 207–66.
Henry, A. S. (1977), The Prescripts of Athenian Decrees (Leiden).
——(1991), ‘A Lead Letter from Torone’, Arch. Ephem. 130, 65–70.
——(2001), ‘A Lead Letter’, in A. Cambitoglou, J. Papadopoulos and O.T. Jones (eds),
Torone I. The Excavations of 1975, 1976 and 1978 (Athens), 765–71.
Hermann, P. (1981), ‘Teos und Abdera in 5. Jahrhundert v. Chr.’, Chiron 11, 1–29.
Heskel, J. (1997), The North Aegean Wars, 371–360 BC (Stuttgart).
Hesnard, A. (1999), ‘Une Lettre d’armateur sur plomb’, in A. Hesnard et al. (eds.), Mar-
seilles: 10 ans d’archéologie, 2600 ans d’histoire (Marseilles), 44.
Heubeck, A. (1979), Schrift (Göttingen).
Higbie, C. (2013), ‘Lindian Chronicle (532)’, in BNJ <http://referenceworks.brillonline.
com/entries/brill-s-new-jacoby/lindian-chronicle-532-a532>.
Hinz, V. (2001), Nunc Phalaris doctum protulit ecce caput: Antike Phalarislegende und
Nachleben der Phalarisbriefe (Munich).
Hodkinson, O. (2007a), ‘ “Novels in the Greek Letter”: Inversions of the Written–Oral
Hierarchy in the Briefroman “Themistocles” ’, in V. Rimell (ed.), Seeing Tongues, Hearing
Scripts: Orality and Representation in the Ancient Novel (Groningen), 257–78.
——(2007b), ‘Better than Speech: Some Advantages of the Letter in the Second Sophistic’,
in R. Morello and A. D. Morrison (eds.), Ancient Letters: Classical and Late Antique
Epistolography (Oxford), 283–300.
——(forthcoming), ‘Epistolography’, in T. K. Hubbard (ed.), A Companion to Greek and
Roman Sexualities.
Hofmann, V. (forthcoming), ‘Mimesis vel aemulatio? Die hellenistischen Anfänge
der römischen Epistolographie und ihre machtpolitischen Implikationen’, Zeitschrift
der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung, 131 (2014).
Hohti, P. (1976), The Interrelation of Speech and Action in the Histories of Herodotus,
Commentationes Humanarum Litterarum 57 (Helsinki).
Hollmann, A. (2005), ‘The Manipulation of Signs in Herodotos’ Histories’, TAPhA 135/2, 279–327.
Holzberg, N. (1994) (ed.), Der griechische Briefroman: Gattungstypologie und Briefanalyse
(Tübingen).
Hornblower, S. (1982), Mausolus (Oxford).
——(1991), A Commentary on Thucydides, i: Books 1–3 (Oxford).
——(1994), ‘Narratology and Narrative Techniques in Thucydides’, in S. Hornblower
(ed.), Greek Historiography (Oxford), 131–66.
——(1996), A Commentary on Thucydides, ii: Books 4–5.24 (Oxford).
——(2008), A Commentary on Thucydides, iii: Books 5.25–8.109 (Oxford).
Horsfall, N. (2008), Virgil, Aeneid 2: A Commentary (Leiden).
How, W. W., and Wells, J. (1928), A Commentary on Herodotus, i-ii, 2nd edn (Oxford).
Hoz, M. P. de (1997), ‘Epigrafía griega en Hispania’, Epigraphica 59, 29–96.
Hubbard, T. (2007), ‘Theognis’ Sphregis: Aristocratic Speech and the Paradoxes of Writing’,
in C. Cooper (ed.), Politics of Orality (Leiden), 193–215.
Hunt, H. S., and Edgar, C. C. (1932), Select Papyri, i (London).
Hunter, R. L. (1983), Eubulus: The Fragments (Cambridge).
Hurlet, F. (2010), ‘Pouvoirs et autoreprésentation du Prince à travers la correspondance
impériale d’Auguste à Trajan (27 avant J.-C.–117 après J.-C.)’, in I. Savalli-Lestrade and
I. Cogitore (eds.), Des rois aux prince: Pratiques du pouvoir monarchique dans l’Orient
hellénistique et romain (Grenoble), 123–45.
Iannucci, A. (1998), ‘Elegia e lotta politica. Note di lettura a Crit. fr. 5 W.’, AION (filol) 20,
107–27.
 Bibliography 397
Iannucci, A. (2002), La parola e l’azione: I frammenti simposiali di Crizia (Bologna).
Invernizzi, A. (2003), ‘They Did Not Write on Clay: Non-Cuneiform Documents and
Archives in Seleucid Mesopotamia’, in M. Brosius (ed.), Ancient Archives and Archival
Traditions: Concepts of Record-Keeping in the Ancient World (Oxford), 302–22.
Inwood, B. (2007), ‘The Importance of Form in Seneca’s Philosophical Letters’, in
R. Morello and A. D. Morrison (eds.), Ancient Letters: Classical and Late Antique
Epistolography (Oxford), 133–48.
Jacoby, F. (1913), ‘Herodotos’, in RE, suppl. ii (Stuttgart), 205–520.
Jailenko, V. P. (2001), ‘Über den Umgang mit pontischen Inschriften’, Das Altertum 46, 223–32.
Jebb, R. C. (1888), ‘The Speeches in Thucydides’, in E. Abbott (ed.), Hellenica: A Collection
of Essays on Greek Poetry, Philosophy, History and Religion (London), 266–323.
Jedrkiewicz, S. (1997), Il convitato sullo sgabello: Plutarco, Esopo e i Sette Savi (Pisa).
Jeffery, L. H. (1967), ‘Archaia Grammata: Some Ancient Greek Views’, in W. C. Brice (ed.),
Europa: Festschrift Ernst Grumach (Berlin), 152–66.
——(1990), The Local Scripts of Archaic Greece, rev. edn with a supplement by
A. W. Johnston (Oxford).
Jenkins, T. E. (2005), ‘At Play with Writing: Letters and Readers in Plautus’, TAPhA 135, 359–92.
——(2006), Intercepted Letters: Epistolarity and Narrative in Greek and Roman Literature
(Lanham, MD).
Johnson, W. A., and Holt, P. R. (2008) (eds.), Ancient Literacies (Oxford).
Johnston, A. W. (2005), ‘Kommos: Further Iron Age Pottery’, Hesperia 74/3, 309–94.
——, and Andreiomenou, A. (1989), ‘A Geometric Graffito from Eretria’, ABSA 84,
217–20.
Johnston, S. I. (1999), Restless Dead: Encounters between the Living and the Dead in Ancient
Greece (Berkeley).
Jones, C. P. (2006), Philostratus, iii: Apollonius of Tyana: Letters of Apollonius, Ancient
Testimonia, Eusebius’ Reply to Hierocles (Cambridge, MA).
Jong, I. J. F. de (1987), Narrators and Focalizers: The Presentation of the Story in the Iliad
(Amsterdam).
——(1991), Narrative in Drama: The Art of the Euripidean Messenger-Speech (Leiden).
——(1999), ‘Aspects narratologiques des Histoires d’Hérodote’, Lalies 19, 217–75.
——(2002), ‘Developments in Narrative Technique in the Odyssey’, in M. Reichel and
A. Rengakos (eds.), Epea Pteroenta: Beiträge zur Homerforschung. Festschrift für Wolf-
gang Kullmann zum 75. Geburtstag, 77–91.
Jordan, D. R. (1978), review of M. Lang, Graffiti and dipinti. The Athenian Agora XXI,
Princeton 1976, Archaeological News 7, 92–4.
——(1985), ‘A Survey of Greek Defixiones’, GRBS 26/2, 151–97.
——(1999), ‘Three Curse Tablets’, in D. R. Jordan, H. Montgomery, and E. Thomassen
(eds.), The World of Ancient Magic (Bergen), 115–24.
——(2000), ‘A Personal Letter Found in the Athenian Agora’, Hesperia 69, 91–103.
——(2003a), ‘A Letter from the Banker Pasion’, in D. Jordan and J. Traill (eds.), Lettered
Attica: A Day of Attic Epigraphy (Athens), 23–39.
——(2003b), ‘EΛΛHNIΚΗ ΕΠΙΣΤΟΛΟΓΡΑΦIA. ΠΡΩIMEΣ ΕΠΙΓΡΑΦΙΚΕΣ ΜΑΡΤΥ-
ΡΙΕΣ’, in A.-Ph. Christides (ed.), ΙΣΤΟΡΙΑ ΤΗΣ ΕΛΛΗΝΙΚΗΣ ΓΛΩΣΣΑΣ ΑΠΟ ΤΙΣ
ΑΡΧΗΣ ΕΩΣ ΤΗΝ ΥΣΤΕΡΗ ΑΡΧΑΙΟΤΗΤΑ (Athens, 2001), 1012–21.
——, and Curbera, J. (2008), ‘A Lead Curse Tablet in the National Archaeological
Museum, Athens’, ZPE 166, 135–50.
Jouan, F. (1966), Euripide et les légendes des chants cypriens (Paris).
——, and van Looy, H. (2000), Euripide, viii.2: Fragments (Paris).
Kahil, L. (1990), ‘Iphigeneia’, in LIMC, v (Zurich), 706–19.
——(1994), ‘Proitos’, in LIMC, vii (Zurich), 525–6.
Kallet-Marx, R. (1995), ‘Quintus Fabius Maximus and the Dyme Affair’, CQ 45, 129–53.
398 Bibliography
Katz, J. (1998), ‘How to be a Dragon in Indo-European: Hittite illuyankaš and its Linguistic
and Cultural Congeners in Latin, Greek, and Germanic’, in J. H. Jasanoff, H. C. Melchert,
and L. Oliver (eds.), Mír Curad: Studies in Honor of Calvert Watkins (Innsbruck),
317–34.
Kauffmann-Samaras, A. (1992), ‘Mousaios’, in LIMC, vii.1 (Zurich), 685–7.
Kelly, Th. (1985), ‘The Spartan Scytale’, in J. W. Eadie and J. Ober (eds.), The Craft of the
Ancient Historian: Essays in Honor of Chester G. Starr (Lanham, MD), 141–69.
Kennedy, G. (1973), ‘Introduction’, in P. A. Stadter (ed.), The Speeches in Thucydides
(Chapel Hill, NC), ix–xii.
Kienast, D. (1973a), Philipp II. von Makedonien und das Reich der Achaimeniden, Abhan-
dlungen der Marburger Gelehrten Gesellschaft, Jahrgang 1971, no. 6 (Munich).
——(1973b), ‘Presbeia, griechisches Gesandtchaftswesen’, in RE Suppl., xiii (Munich),
499–628.
King, L. W. (1907), Chronicles concerning Early Babylonian Kings, Including Records of the
Early History of the Kassites and the Country of the Sea, ii (London).
Klauck, H.-J. (2006), Ancient Letters and the New Testament: A Guide to Context and
Exegesis (Waco, TX) (rev. tr. of Die antike Briefliteratur und das Neue Testament: Ein
Lehr- und Arbeitsbuch (Stuttgart, 1998)).
Klein, J. (2000), ‘The So-Called “Spell of Nudimmud” (ELA 134–155): A Reexamination’, in
S. Graziani (ed.), Studi sul Vicino Oriente Antico dedicati alla memoria di Luigi Cagni, ii
(Naples), 563–84.
Kleingünther, A. (1933), —æH  æ : Untersuchungen zur Geschichte einer Fragestel-
lung (Leipzig).
Knox, B. M. W. (1966), ‘Second Thoughts in Greek Tragedy’, GRBS 7, 213–32; repr. in
B. Knox, Word and Action: Essays on the Ancient Theater (Baltimore, 1979), 275–94.
——(1968), ‘Silent Reading in Antiquity’, GRBS 9, 421–35.
——(1972), ‘Euripides’ Iphigenia in Aulide 1–163 (In That Order)’, YCS 22; repr. in
B. Knox, Word and Action: Essays on the Ancient Theater (Baltimore), 1979, 275–94
[to which I refer].
Koenig, J. (2007), ‘Alciphron’s Epistolarity’, in R. Morello and A. D. Morrison (eds.),
Ancient Letters: Classical and Late Antique Epistolography (Oxford), 257–82.
Komoróczy, G. (1975), ‘Ätiologie der Schrifterfindung in Enmerkar-Epos’, Altorientalische
Forschungen 3, 19–26.
Konstantakos, I. (2000), ‘Notes on the Chronology and Career of Antiphanes’, Eikasmos:
Quaderni Bolognesi di filologia classica 11, 173–98.
Koskenniemi, H. (1956), Studien zur Idee und Phraseologie des griechischen Briefes bis 400
nach Chr. (Helsinki).
Kossatz-Deissmann, A. (1990), ‘Iris I’, in LIMC, v.1 (Zurich), 741–60.
Koumanoudis, S. (1986), ‘˜Ø æŁø ØŒ ’, Horos 4, 157–9.
Kovacs, D. (1980), ‘Shame, Pleasure and Honor in Phaedra’s Great Speech (Euripides,
Hippolytus 375–387)’, AJPh 101, 287–303.
——(2003), ‘Towards a Reconstruction of Iphigenia Aulidensis’, JHS 123, 77–103.
Kranz, W. (1933), Stasimon (Berlin).
——(1961), ‘Sphragis, Ichform und Namensiegel als Eingangs- und Schluβmotiv antiker
Dichtung’, Rhein. Mus. 104, 3–46 and 97–124.
Krauskopf, I. (1981), ‘Aliunea’, in LIMC, i.1 (Zurich), 529–31.
——(1994), ‘Palamedes’ (Etruskisch)’, in LIMC, vii.1 (Zurich), 147–9.
Kroll, J. H. (1977), ‘An Archive of the Athenian Cavalry’, Hesperia 46/2, 83–140.
Kurke, L. (1999), Coins, Bodies, Games and Gold: The Politics of Meaning in Archaic Greece
(Princeton).
Kyriakou, P. (2006), A Commentary on Euripides’ Iphigenia in Tauris (Berlin).
Labarre, G. (1996), Les cités de Lesbos aux époques hellénistique et impériale (Paris).
 Bibliography 399
Lallot, J. (1997), Apollonius Dyscole: De la construction, i: Introduction, texte et traduction,
ii: Notes, 2 vols. (Paris).
——(1998), La Grammaire de Denys le Thrace (Paris).
Lane Fox, R. (2006), ‘The Letter to Gadatas’, in G. E. Malouchou and A. P. Matthaiou (eds.),
ØÆŒ ı Ø : Ø  Å W. G. Forrest (Athens), 149–71.
——(2010), ‘Thucydides and Documentary History’, CQ, ns 60, 11–29.
Lang, M. L. (1976), Graffiti and dipinti, Athenian Agora XXI (Princeton).
——(1984), Herodotean Narrative and Discourse (Cambridge, MA).
Langdon, M. K. (2005), ‘A New Greek Abecedarium’, Kadmos 44, 175–82.
Lardinois, A., Blok, J., and van der Poel, M. G. M. (2011) (eds.), Sacred Words: Orality,
Literacy and Religion (Leiden).
Lasserre, F. and Bonnard, A. (1958), Archiloque. Fragments (Paris).
Lateiner, D. (1987), ‘Nonverbal Communication in the Histories of Herodotus’, Arethusa
20, 83–119.
Lattimore, R. (1939), ‘The Wise Advisor in Herodotus’, CPh 34, 24–35.
——(1951), Homer: The Iliad (Chicago).
Laurenzi, L. (1941), ‘Iscrizioni dell’Asclepieo di Coo’, Clara Rhodos 10, 25–38.
Lazzarini, M. L. (1998), ‘Questioni relative all’origine dell’alfabeto greco’, in G. Bagnasco
Gianni and F. Cordano (eds.), Scritture mediterranee fra IX e VIII sec. a.C. (Milan),
53–61.
Le Guen, B. (2001), Les Associations de technites dionysiaques à l’époque hellénistique,
2 vols. (Nancy).
Lee, K. H. (1997), Euripides: Ion (Warminster).
Lefèvre F. (1998), ‘Traité de paix entre Démétrios Poliorcète et la confédération étolienne
(fin 289 ?)’, BCH 122, 109–41.
——, Laroche, D., and Jacquemin, A. (1995), ‘Delphes, le roi Persée et les Romains’, BCH
119, 125–136.
Lejeune, M. (1960), ‘A propos d’un plomb inscrit d’Elne’, REA 62, 62–79.
Lenfant, D. (2004), Ctésias de Cnide: La Perse. L’Inde. Autres fragments (Paris).
Lenger, M.-T. (1980), Corpus des Ordonnances des Ptolémées (C.Ord.Ptol.), 2nd edn
(Brussels).
Létoublon, F. (1987), ‘Le Messager fidèle’, in M. J. Bremer, I. J. F. de Jong, and J. Kalff (eds.),
Homer: Beyond Oral Poetry: Recent Trends in Homeric Interpretation (Amsterdam),
123–44.
——(2003), ‘La Lettre dans le roman grec ou les liaisons dangereuses’, in S. Panayotakis,
M. Zimmerman, and W. Keulen (eds.), The Ancient Novel and Beyond (Leiden),
271–88.
Lewis, S. (1996), News and Society in the Greek Polis (Chapel Hill, NC).
Lewy, H. (1897–1902), ‘Palamedes’, in W. H. Roscher (ed.), Lexicon der griechischen und
römischen Mythologie (Leipzig), 1264–73.
Liddel, P. (2003), ‘The Places of Publication of Athenian State Decrees from the 5th Century
bc to the 3rd Century ad’, ZPE 143, 79–93.
——(2009), ‘The Decree Cultures of the Ancient Megarid’, CQ, ns 59, 411–36.
Liebeschütz, J. H. W. G. (2004), ‘The Collected Letters of Ambrose of Milan: Correspond-
ence with Contemporaries and with the Future’, in L. Ellis and F. L. Kidner (eds.), Travel,
Communication and Geography in Late Antiquity: Sacred and Profane (Aldershot),
95–108.
Linant de Bellefonds, P. (1990), ‘Hippolytos’, in LIMC, v (Zurich), 445–64.
——(1994), ‘Phaidra’, in LIMC, vii (Zurich), 356–9.
Linders, T. (1992), ‘Inscriptions and Orality’, Symbolae Osloenses 67, 27–40.
Linforth, I. M. (1931), ‘Two Notes on the Legend of Orpheus’, TAPhA 62, 5–17.
——(1941), The Arts of Orpheus (Berkeley).
400 Bibliography
Lissarrague, F. (1985), ‘Paroles d’images: remarques sur le fonctionnement de l’écriture
dans l’imagerie attique’, in A. M. Christin (ed.), Ecritures, ii (Paris), 71–95.
——(1992), ‘Graphein: écrire et dessiner’, in C. Bron, and E. Kassapoglou (eds.), L’Image
en jeu (Lausanne), 189–203.
Livingstone, N. (2001), A Commentary on Isocrates’ Busiris (Leiden).
Llewelyn, S. R. (1994), ‘The Conveyance of Letters’, in New Documents Illustrating Early
Christianity 7, 1–57.
Lloyd-Jones, H. (1994–6), Sophocles, i: Ajax, Electra, Oedipus Tyrannus; ii: Antigone, The
Women of Trachis, Philoctetes, Oedipus at Colonus; iii: Fragments (Cambridge, MA).
Lombardi, P. (2010), ‘La “colpa” di Gadatas. Osservazioni lessicali sulla “Lettera di Dario a
Gadatas” ’, Incidenza dell’antico 8, 163–95.
Lombardo, M. (1988), ‘Marchands, transactions économiques, écriture’, in M. Detienne
(ed.), Les Savoirs de l’écriture: En Grèce ancienne (Lille), 159–87.
Longo, O. (1978), ‘Scrivere in Tucidide: comunicazione e ideologia’, in E. Livrea and
G. A. Privitera (eds.), Studi in onore di Anthos Ardizzoni (Rome), 519–54.
——(1981), Tecniche della comunicazione in Grecia antica (Naples).
——(1997), ‘Circolazione dell’informazione’, in S. Settis (ed.), I Greci: Storia, cultura, arte,
società, ii.2: Una storia Greca: Definizione (Turin), 655–79.
López Jimeno, M. del A. (1990), ‘Las cartas de maldición’, Minerva 4, 133–44.
——(1999), Nuevas Tabellae Defixionis Áticas (Amsterdam).
Loraux, N. (1979), ‘La Gloire et la mort d’une femme’, Sorcières 18, 51–7.
——(1981), L’Invention d’Athènes (Paris).
——(1986), ‘Thucydides a écrit la guerre du Péloponnèse’, Mètis: Anthropologie des
mondes grecs anciens 1, 139–61.
Luppe, W. (2011), ‘Die “Palamedes” und die “Polyidos”-Hypothesis. P. Mich. Inv. 3020
(A)’, ZPE 176, 52–5.
Luraghi, N. (2010), ‘The Local Scripts from Nature to Culture’, Classical Antiquity 29/1,
68–91.
Luschnat, O. (1970), ‘Thukydides’, in RE Suppl. xii, 1085–1354.
Ma, J. (1999), Antiochos III and the Cities of Western Asia Minor (Oxford).
——(2000), ‘Seleukids and Speech-Act Theory: Performative Utterances, Legitimacy and
Negotiation in the World of the Maccabees’, SCI 19, 71–112.
——(2009), ‘City as Memory’, in G. Boys-Stones, B. Graziosi and Ph. Vasunia (eds.), The
Oxford Handbook of Hellenic Studies (Oxford), 248–59.
MacDowell, D. M. (1990), Demosthenes: Against Meidias (Oxford).
——(2004), Demosthenes: Speeches 27–38, Austin, TX.
McGing, B. (2010), Polybius’ Histories (Oxford).
McKechnie, P. R., and Kern, S. J. (1988) (eds. and trans.), Hellenica Oxyrhynchia (London).
McLean, B. H. (2002), An Introduction to Greek Epigraphy of the Hellenistic and Roman
Periods from Alexander the Great down to the Reign of Constantine (323 b.c.–a.d. 337)
(Ann Arbor).
Mader, G. (2006), ‘Fighting Philip with Decrees: Demosthenes and the Syndrome of
Symbolic Action’, AJPh 127, 367–86.
Maffi, A. (1988), ‘Écriture et pratique juridique dans la Grèce classique’, in M. Detienne
(ed.), Les Savoirs de l’écriture: En Grèce ancienne (Lille), 188–210.
Magini, D. (1997–8), ‘ “Scrivere nelle tavolette della mente”: la memoria nei tragici greci’,
Rendiconti dell’ Accademia di Archeologia, Lettere e Belle Arti di Napoli, ns 67, 219–46.
——(1998), ‘Parole e grammata nella tragedia greca: Eschilo e Sofocle’, Atti Accademia
Pontaniana 47, 21–47.
Magnetto, A. (1997), Gli arbitrati interstatali greci: Introduzione, testo critico, traduzione,
commento e indici, ii: Dal 337 al 196 a.C. (Pisa).
Maidment, K. J. (1941), Minor Attic Orators: Antiphon, Andocides (Cambridge, MA).
Malherbe, A. J. (1988), Ancient Epistolary Theorists (Atlanta).
 Bibliography 401
Malosse, P.-L. (2004), Lettres pour toutes circonstances: Les Traités épistolaires du Pseudo-
Libanios et du Pseudo-Démétrios de Phalère (Paris).
Manni Piraino, T. (1969), ‘Un plomb inscrit archaïque d’Himère’, REA 71, 301–4.
——(1970), ‘Plomb inscrit d’Himère. Note additionnelle’, REA 72, 385.
——(1972), ‘Ancora sulla lamina plumblea iscritta di Imera’, in Studi e materiali. Istituto di
archeologia dell’Università di Palermo. Quaderno imerese 1, 107–9.
Mari, M. (2006), ‘L’Activité législative du roi et des cités en Macédoine’, in A.-M. Guimier-
Sorbets, M. B. Hatzopoulos, and Y. Morizot (eds.), Rois, cités, nécropoles: Institutions,
rites et monuments en Macédoine (Athens), 209–25.
——(2009), ‘La tradizione delle libere poleis e l’opposizione ai sovrani. L’evoluzione del
linguaggio della politica nella Grecia ellenistica’, in G. Urso (ed.), Ordine e sovversione nel
mondo greco e romano (Pisa), 87–112.
——(forthcoming), ‘Powers in Dialogue: The Letters and Diagrammata of Macedonian
Kings to Local Communities’, in P. Ceccarelli, L. Doering, T. Foegen, and I. Gildenhard
(eds.), Configuring Communities: The Socio-Political Dimensions of Ancient Epistologra-
phy (Oxford).
Marincola, J. (1997), Authority and Tradition in Ancient Historiography (Cambridge).
Martin, R. (2001), ‘Just Like a Woman: Enigmas of the Lyric Voice’, in A. Lardinois and
L. Mc Clure (eds.), Making Silence Speak: Women’s Voices in Greek Literature and Society
(Princeton), 55–74.
Martin, T. W. (2010), ‘Investigating the Pauline Letter Body: Issues, Methods, and
Approaches’, in S. E. Porter and S. A. Adams (eds.), Paul and the Ancient Letter Form
(Leiden), 185–212.
Martin, V., and Budé, G. (1928), Eschine: Discours, ii (Paris).
Mastronarde, D. J. (1994), Euripides: Phoenissae (Cambridge).
Matthews, V. J. (1974), ‘The Hemerodromoi: Ultra-Long Distance Running in Antiquity’,
CW 68, 161–9.
Mauersberger, A., et al. (1956–2004), Polybios-Lexicon (Berlin).
Meineke, A. (1839), Historia critica comicorum graecorum (Berlin).
Meritt, B. D. (1940), Epigraphica Attica (Cambridge, MA).
Merkelbach, R. (1954), Die Quellen des griechischen Alexanderromans (Munich).
Michalowski, P. (1993), Letters from Early Mesopotamia (Atlanta).
——(1999), ‘Commemoration, Writing and Genre in Ancient Mesopotamia’, in C. Kraus
(ed.), The Limits of Historiography: Genre and Narrative in Ancient Historical Texts
(Leiden), 69–90.
Millar, F. (1967), ‘Emperors at Work’, JRS 57, 9–19.
——(1992), The Emperor in the Roman World, 2nd edn (London).
Millender, E. (2001), ‘Spartan Literacy Revisited’, Classical Antiquity 20, 121–64.
Millett, P. (1991), Lending and Borrowing in Ancient Athens (Cambridge).
Milns, R. D. (2000), ‘The Public Speeches of Demosthenes’, in I. Worthington (ed.),
Demosthenes: Statesman and Orator (London), 205–23.
Minon, S. (2007), Les Inscriptions éléennes dialectales (VIe–IIe siècle avant J.-C.), 2 vols. (Geneva).
Mirhady, D. C. (2000), ‘Demosthenes as Advocate. The Private Speeches’, in I. Worthington
(ed.), Demosthenes: Statesman and Orator (London), 181–204.
Missiou, A. (2011), Literacy and Democracy in Fifth-Century Athens (Cambridge).
Mitchell, L. G. (1997), Greeks Bearing Gifts: The Public Use of Private Relationships in the
Greek World, 435–323 BC (Cambridge).
——(2001), ‘Euboean Io’, CQ, ns 51/2, 339–52.
——(2006), ‘Greeks, Barbarians, and Aeschylus’ Suppliants’, Greece and Rome, ns 53,
2052–223.
Moatti, C. (2006), ‘Translation, Migration, and Communication in the Roman Empire:
Three Aspects of Movement in History’, Classical Antiquity 25, 109–140.
Moggi, M. (1972), ‘Autori greci di Persika. I: Dionisio di Mileto’, ASNP, 3rd ser., 2, 433–68.
402 Bibliography
Moles, J. (1999), ‘Anathema kai Ktema: The Inscriptional Inheritance of Ancient Histori-
ography’, Histos 3, 27–59 <http://research.ncl.ac.uk/histos/documents/1999.03Moles-
AnathemakaiKtema2769.pdf>.
Momigliano, A. (1966), ‘Fattori orientali della storiografia ebraica postesilica e della
storiografia greca’, in Terzo contributo alla storia degli studi classici e del mondo antico
(Rome), 807–18.
Monaco, G. (1965) ‘L’epistola nel teatro greco’, Dioniso 39, 334–51.
Montanari, F. (1995), ‘II grammatico Tolomeo Pindarione, i poemi omerici e la scrittura’,
in F. Montanari, Studi di filologia omerica antica, ii (Pisa), 41–58.
Morello, R., and Morrison, A. D. (2007) (eds.), Ancient Letters: Classical and Late Antique
Epistolography (Oxford).
Morgan, K. A. (2000), Myth and Philosophy from the Presocratics to Plato (Cambridge).
Morrison, J. V. (1994), ‘A Key Topos in Thucydides: The Comparison of Cities and
Individuals’, AJPh 115, 525–41.
Mosley, D. J. (1973), Envoys and Diplomacy in Ancient Greece (Wiesbaden).
Most, G. W. (1995), ‘Reflecting Sappho’, BICS 40, 15–38.
Muir, J. (2009), Life and Letters in the Ancient Greek World (London).
Mülke, C. (2007), ‘P.Oxy. LXXI 4807: Sophocles, Epigonoi’, in R. Hatzilambrou,
P. J. Parsons, J. Chapa et al. (eds. and trans.), The Oxyrhynchus Papyri, lxxi (London).
Müller, C. W. (1990), ‘Der Palamedesmythos im Philoktet des Euripides’, RhM 133,
193–209.
Müller, P. (1992), ‘Oiax’, in LIMC, vi.1 (Zurich), 983–4.
Munson, R. V. (2007), ‘The Trouble with the Ionians: Herodotus and the Beginning of the
Ionian Revolt (5.28–38.1)’, in E. Irwin and E. Greenwood (eds.), Reading Herodotus:
A Study of the logoi in Book 5 of Herodotus’ Histories (Cambridge), 146–67.
Murray, O. (2001a), ‘Herodotus and Oral History’, in N. Luraghi (ed.), The Historian’s
Craft in the Age of Herodotus (Oxford), 16–44.
——(2001b), ‘Herodotus and Oral History Reconsidered’, in N. Luraghi (ed.), The Histor-
ian’s Craft in the Age of Herodotus (Oxford), 314–25.
Musti, D. (1986), ‘Democrazia e scrittura’, Scrittura e civiltà 10, 21–48.
——(1997), Demokratia: Origini di un’idea, 2nd edn (Rome) (1st edn, 1995).
—— and Torelli, M. (1991) (eds.), Pausania: Guida della Grecia, iv: La Messenia (Milan).
Mylonas Shear, I. (1998), ‘Bellerophon Tablets from the Mycenaean World? A Tale of
Seven Bronze Hinges’, JHS 118, 187–9.
Nadjo, L. (2004), ‘Plaute épistolier’, in L. Nadjo and E. Gavoille (eds.), Epistulae antiquae, iii
(Louvain), 111–25.
Nafissi, M. (2004), ‘Tucidide, Erodoto e la tradizione su Pausania nel V secolo’, RSA 34,
115–48.
Naiden, F. S. (1999), ‘The Prospective Imperfect in Herodotus’, HSCP 99, 135–49.
Natoli, A. F. (2004), The Letter of Speusippus to Philip II (Stuttgart).
Neer, R. (1995), ‘The Lion’s Eye: Imitation and Uncertainty in Attic Red Figure’, Repre-
sentations 51, 118–53.
——(1998), ‘Imitation, Inscription, Antilogic’, Mètis: Anthropologie des mondes grecs
anciens 13, 17–38.
Nenci, G. (1994), Erodoto: Le storie, v: La rivolta ionica (Milan).
——(1998), Erodoto: Le storie, vi: La battaglia di Maratona (Milan).
Neri, C. (2007), ‘I rimedi dell’oblio (Eur. Palam. fr. 578 K.)’, Eikasmos 18, 167–71.
Nesselrath, H.-G. (1996), ‘Antiphanes’, in DNP, i, 781–2.
Nicolai, R. (1992), La storiografia nell'educazione antica (Pisa).
——(2004), Studi su Isocrate: La comunicazione letteraria nel IV sec. a.C. e i nuovi generi
della prosa (Rome).
Nieddu, G. F. (2004), La scrittura ‘madre delle Muse’: agli esordi di un nuovo modello di
comunicazione culturale (Amsterdam).
Nisoli, A. G. (2007), ‘Parole segrete: le defixiones’, Acme 60/3, 36–45.
 Bibliography 403
Norwood, G. (1930), ‘The Babylonians of Aristophanes’, CPh 25, 1–10.
Nouhaud, M. (1982), L’Utilisation de l’histoire par les orateurs attiques (Paris).
Noussia-Fantuzzi, M. (2010), Solon the Athenian: The Poetic Fragments (Leiden).
Ober, J. (1989), Mass and Elite in Democratic Athens (Princeton).
——(2004), ‘I, Socrates . . . The Performative Audacity of Isocrates’ Antidosis’, in D. Depew
and T. Poulakos (eds.), Isocrates and Civic Education (Austin, TX), 21–43.
Ohlert, K. (1886), Rätsel und Gesellschaftspiele der alten Griechen (Berlin).
Oikonomides, A. N. (1983),‘An “Unpublished” Greek Inscription from Rhode (Roses) in
Spain’, AncW 7, 107–10.
——(1986), ‘Graffiti-Inscriptions from the Excavations of the Athenian Agora at Kera-
meikos’, Horos 4, 43–64.
Olding, G. (2007), ‘Myth and Writing in Aeschines’ Against Timarchus’, in C. Cooper (ed.),
Politics of Orality (Leiden), 155–70.
Olson, R. S. (2010), Tragedy, Authority, and Trickery: The Poetics of Embedded Letters in
Josephus (Washington, DC).
Olson, S. D. (2002), Aristophanes: Acharnians (Oxford).
——(2007), Broken Laughter: Select Fragments of Greek Comedy (Oxford).
Orsi, D. P. (2002), ‘Trattative internazionali nelle Elleniche senofontee. Aspetti del lessico: i
verbi della comunicazione’, in L. R. Cresci, F. Gazzano, and D. P. Orsi (eds.), La retorica
della diplomazia nella Grecia antica e a Bisanzio (Rome), 69–109.
Osborne, R. (1999), ‘Inscribing performance’, in S. Goldhill and R. Osborne (eds.), Per-
formance Culture and Athenian Democracy (Cambridge), 341–58.
——(2009), ‘The Politics of an Epigraphic Habit: The Case of Thasos’, in L. Mitchell and
L. Rubinstein (eds.), Greek History and Epigraphy: Essays in Honour of P. J. Rhodes
(Swansea), 103–14.
——(forthcoming), ‘Letters, Diplomacy, and the Roman Conquest of Greece’, in
P. Ceccarelli, L. Doering, T. Foegen, and I. Gildenhard (eds.), Configuring Communities:
The Socio-Political Dimensions of Ancient Epistolography (Oxford).
O’Sullivan, N. (1996), ‘Written and Spoken in the First Sophistic’, in I. Worthington (ed.),
Voice into Text: Orality and Literacy in Ancient Greece (Leiden) 1996, 115–27.
——(2008), ‘The Authenticity of [Alcidamas’] Odysseus: Two New Linguistic Consider-
ations’, CQ, NS 58, 638–47.
O’Sullivan, P. (2000), ‘Satyr and Image in Aeschylus’ Theoroi’, CQ, ns 50, 353–66.
Palaima, T. (2004), ‘ “Archives” and “Scribes” and Information Hierarchy in Mycenaean
Greek Linear B Records’, in M. Brosius (ed.), Ancient Archives and Archival Traditions:
Concepts of Record-Keeping in the Ancient World (Oxford), 153–94.
Palme-Koufa, A. (1996), ‘Die Graffiti auf der Keramik’, in R. C. S. Felsch (ed.), Kalapodi,
i (Mainz am Rhein), 273–331.
Pantos, A. P. (1996), ‘Porträtsiegel in Kallipolis: Einige methodologische Bemerkungen’, in
M.-F. Boussac and A. Invernizzi (eds.), Archives et sceaux du monde hellénistique /
Archivi e sigilli nel mondo ellenistico, BCH suppl. 29 (Athens), 185–94.
Parker, R. (2005), Polytheism and Society at Athens (Oxford).
Parmeggiani, G. (2012), ‘Sui fondamenti della tesi antica della paternità anassimenea del
Tricarano’, Histos 6, 214–27. <http://research.ncl.ac.uk/histos/documents/2012A10Par-
meggianiTricarano.pdf>.
Parry, A. (1981) [1957], LˇGˇS and ¯Gˇ˝ in Thucydides (Salem, NH).
Parsons, P. J. (1980), ‘Background: The Papyrus Letter’, Didactica Classica Gandensia 20, 3–19.
Paschidis, P. (2008), Between City and King: Prosopographical Studies on the Intermediaries
between the Cities of the Greek Mainland and the Aegean and the Royal Courts in the
Hellenistic Period (322–190 bc) (Athens).
Pastorino, A. (1955), ‘Turpilio fr. 213 R3’, Maia 7, 43–5.
404 Bibliography
Paton, W. R. (2010–12), Polybius: The Histories, 6 vols., rev. edn and trans. by F. W. Walbank
and C. Habicht; Unattributed fragments ed. and trans. by S. D. Olson (Cambridge, MA).
Patrito, P. (2001), ‘Sul ‘Frammento di Dike’ (= Aesch. Frr. 281 a-b Radt)’, Quaderno del
Dipartimento di filologia A. Rostagni, 77–95.
Payton, R. (1991), ‘The Ulu Burun Writing-Board Set’, AnSt 41, 99–106.
Pébarthe, C. (2005), ‘Inscriptions et régime politique: le cas athénien’, in A. Bresson,
A.-M. Cocula, and C. Pébarthe (eds.), L’Écriture publique du pouvoir (Bordeaux), 169–82.
——(2006), Cité, démocratie et écriture: Histoire de l’alphabétisation d’Athènes à l’époque
classique (Paris).
Pechstein, N., and Krumeich, R. (1999), ‘Achaios. Omphale’, in R. Krumeich, N. Pechstein,
and B. Seidensticker (eds.), Das griechische Satyrspiel (Darmstadt), 539–42.
Pédech, P. (1964), La Méthode historique de Polybe (Paris).
Pelling, C. (1996), ‘The Urine and the Vine: Astyages’ Dreams at Herodotus I 107–8’, CQ,
ns 46, 68–77.
Perrin, B. (1914–26), Plutarch’s Lives, With an English Translation (Cambridge, MA).
Perry, B. E. (1952), Aesopica (Urbana, IL).
Pfeiffer, R. (1968), History of Classical Scholarship, i (Oxford).
Piccaluga, G. (1988), ‘Potmos impara a scrivere’, Scrittura e civiltà 12, 5–46.
Piccirilli, L. (1997), in M. G. Angeli Bertinelli, M. Manfredini, L. Piccirilli, and G. Pisani
(eds.), Plutarco: Le vite di Lisandro e di Silla (Milan).
——(2002), L’invenzione della diplomazia nella Grecia antica (Rome).
Porciani, L. (1997), La forma proemiale: Storiografia e pubblico nel mondo antico (Pisa).
——(2003), ‘Logoi, erga, documenti. Il caso della tregua del 423 a.C. fra Atene e Sparta’, in
A. M. Biraschi et al. (eds.), L’uso dei documenti nella storia antica (Naples), 315–29.
Poster, C. (2007), ‘A Conversation Halved: Epistolary Theory in Greco-Roman Antiquity’,
in C. Poster and L. C. Mitchell (eds.), Letter-Writing Manuals and Instruction from
Antiquity to the Present (Columbia, SC), 21–51.
Powell, B. B. (1989), ‘Why Was the Greek Alphabet Invented? The Epigraphic Evidence’,
Classical Antiquity 8, 321–50.
——(1991), Homer and the Origin of the Greek Alphabet (Cambridge).
——(1997), ‘Homer and Writing’, in I. Morris and B. B. Powell (eds.), A New Companion
to Homer (Leiden), 5–32.
——(2002), Writing and the Origins of Greek Literature (Cambridge).
Powell, J. E. (1938), A Lexicon to Herodotus (Cambridge).
Pownall, F. (2013), ‘Duris of Samos (76)’, in BNJ <http://referenceworks.brillonline.com/
entries/brill-s-new-jacoby/duris-of-samos-76-a76>.
Prandi, L. (2003), ‘Riflessioni sull’uso dei documenti in Polibio’, in A. M. Biraschi et al.
(eds.), L’uso dei documenti nella storia antica (Naples), 375–90.
Preisendanz, K. (1972), ‘Fluchtafel (Defixio)’, in Reallexikon für Antike und Christentum,
viii (Stuttgart), 1–29.
Pritchard, J. B. (1969), Ancient Near Eastern Texts relating to the Old Testament, 3rd edn
(Princeton).
Pritchett, W. K. (1974), The Greek State at War, ii (Berkeley).
Pucci, P. (1988), ‘Inscriptions archaïques sur les statues des dieux’, in M. Detienne (ed.), Les
Savoirs de l’écriture en Grèce ancienne (Lille), 480–97.
——(1993), ‘L’apologie d’Apollon dans Hérodote 1, 91’, Mètis: Anthropologie des mondes
grecs anciens 8/1–2, 7–20.
——(1996), Enigma, segreto, oracolo (Pisa).
——(2003) [1977], ‘Euripides: The Monument and the Sacrifice’, in J. Mossmann (ed.),
Oxford Readings in Classical Studies: Euripides (Oxford), 139–69.
Purcell, N. (1995), ‘Roman Urban Society and the Game of Alea’, Past and Present, no. 147, 3–37.
 Bibliography 405
Py, M. (2009), Lattara. Lattes, Hérault. Comptoir gaulois méditerranéen entre Étrusques,
Grecs et Romains (Paris).
Rainey, A. F. (1968), ‘The Scribe at Ugarit. His Position and Influence’, in Proceedings of the
Israel Academy of Sciences and Humanities 3, 126–47.
Randall, D. (2008), ‘Epistolary Rhetoric, the Newspaper, and the Public Sphere’, Past and
Present, no. 198, 3–32.
Raubitschek, A. E. (1943), ‘Inscriptions’, in G. R. Davidson and D. Burr Thompson (ed.),
Small Objects from the Pnyx, i, Hesperia Suppl. 7, 1–11.
Reed, J. T. (1997), ‘The Epistle’, in S. E. Porter (ed.), Handbook of Classical Rhetoric in the
Hellenistic Period (Leiden), 171–93.
Restani, D. (1995), ‘I suoni del telaio. Appunti sull’universo sonoro degli antichi Greci’, in
B. Gentili and F. Perusino (eds.), Metrica, ritmica e musica greca in memoria di Giovanni
Comotti (Pisa), 93–109.
Reynolds, J. (1982), Aphrodisias and Rome (London).
Rhodes, P. J. (1972), The Athenian Boule (Oxford).
——(2003), ‘Nothing to do with Democracy: Athenian Drama and the Polis’, JHS 123,
104–19.
——, with Lewis, D. M. (1997), The Decrees of the Greek States (Oxford).
——, and Osborne, R. (2003), Greek Historical Inscriptions 403–323 bc (Oxford).
Ribbeck, O. (1898), Comicorum Romanorum Fragmenta, 3rd edn (Leipzig).
Richardson, N. J. (1974), The Homeric Hymn to Demeter (Oxford).
Richer, N. (1998), Les Éphores: Études sur l’histoire et sur l’image de Sparte (VIII–III s. av.
J.-C.) (Paris).
Rigsby, K. J. (1996), Asylia: Territorial Inviolability in the Hellenistic World (Berkeley).
——, and Hallof, K. (2001), ‘Aus der Arbeit der “Inscriptiones Graecae”. 10, Decrees of
inviolability for Kos’, Chiron 31, 333–45.
Robb, K. (1978), ‘The Poetic Sources of the Greek Alphabet’, in E. Havelock and
J. Hershbell (eds.), Communication Arts in the Ancient World (New York), 23–36.
Robert, L. (1978), ‘Documents d’Asie Mineure, V-XVII’, BCH 102, 395–543.
Robertson, J. S. (2004), ‘The Possibility and Actuality of Writing’, in S. D. Houston, The
First Writing: Script Invention as History and Process (Cambridge), 16–38.
Robson, E. (2007), ‘Literacy, Numeracy, and the State in Early Mesopotamia’, in Lomas
et al. (eds.), Literacy and the State in the Ancient Mediterranean (London), 39–52.
Rodríguez Somolinos, H. (1998), ‘Inscriptiones Graecae Antiquissimae Iberiae (IGAI)’, in
J. Mangas and D. Plácido, (eds.), Testimonia Hispaniae Antiqua, ii a (Madrid), 333–62.
Roller, D. W. (2013), ‘Theokritos of Chios (760)’, in BNJ <http://referenceworks.brillonline.
com/entries/brill-s-new-jacoby/Theokritos-of-chios-760-a760>.
Rood, T. (1998), Thucydides: Narrative and Explanation (Oxford).
Rosén, H. B. (1987–97), Herodotus: Historiae, 2 vols. (Stuttgart).
Rosen, R. M. (1997), ‘The Gendered Polis in Eupolis’ Cities’, in G. W. Dobrov (ed.), The City
as Comedy: Society and Representation in Athenian Drama (Chapel Hill, NC), 149–76.
——(1999), ‘Comedy and Confusion in Callias’ Letter Tragedy’, CPh 94.2, 147–67.
——(2003), ‘Revisiting Sophocles’ Poimenes: Tragedy or Satyr Play?’, in A. H. Sommerstein
(ed.), Shards from Kolonos: Studies in Sophoclean Fragments (Bari), 373–86.
Rosenmeyer, P. A. (2001), Ancient Epistolary Fictions: The Letter in Greek Literature
(Cambridge, MA).
——(2006), Ancient Greek Literary Letters: Selections in Translation (London).
Rösler, W. (1980), Dichter und Gruppe: Eine Untersuchung zu den Bedingungen und zur
historischen Funktion früher griechischer Lyrik am Beispiel Alkaios (Munich).
——(2002), ‘The Histories and Writing’, in E. J. Bakker, I. J. F. de Jong, and H. van Wees
(eds.), Brill’s Companion to Herodotus (Leiden), 79–94.
Rousset, D. (2002a), ‘Inscriptions hellénistiques d’Amphissa’, BCH 126 (1), 83–95.
——(2002b), Le Territoire de Delphes et la terre d’Apollon (Paris).
406 Bibliography
Rowe, G. O. (1968), ‘Demosthenes’ First Philippic: The Satiric Mode’, TAPhA 99, 361–74.
Rubinstein, L. (1998), ‘The Athenian Political Perception of the Idiotes’, in P. Cartledge,
P. Millett, and S. von Reden (eds.), Kosmos: Essays in Order, Conflict and Community in
Classical Athens (Cambridge), 125–43.
—— (2013), ‘Spoken Words, Written Submissions, and Diplomatic Conventions: The Import-
ance and Impact of Oral Performance in Hellenistic Inter-polis Relations’, in C. Kremmydas
and K. Tempest (eds.), Hellenistic Oratory: Continuity and Change (Oxford), 165–99.
Ruijgh, C. J. (2001), ‘Le Spectacle des lettres, comédie de Callias’, Mnemosyne, 4th ser., 54,
257–335.
Rusjaeva, A. S., and Vinogradov, J. G. (1991), ‘Der “Brief des Priesters” aus Hylaia’, in R. Rolle,
M. Müller-Wille, and K. Schietzel (ed.), Gold der Steppe: Ausstellungskatalog, (Schleswig).
Russell, F. S. (1999), Information-Gathering in Classical Greece (Ann Arbor).
Rusten, J., and Cunningham, I. C. (2002), Theophrastus: Characters; Herodas: Mimes;
Sophron and Other Mime Fragments (Cambridge, MA).
Rychlewska, L. (1971), Turpilius: Fragmenta (Leipzig).
Saïd, S. (1985), Sophiste et tyran, ou: Le Problème du Prométhée enchaîné (Paris).
Sallaberger, W. (1999), ‘Wenn du mein Bruder bist, . . . ’: Interaktion und Textgestaltung in
altbabylonischen Alltagsbriefen (Groningen).
Salzman, M. R. (2004), ‘Travel and Communication in the Letters of Symmachus’, in
L. Ellis and F. L. Kidner (eds.), Travel, Communication and Geography in Late Antiquity:
Sacred and Profane (Aldershot), 81–94.
Samama, E. (2003), Les Médecins dans le monde grec: Sources épigraphiques sur la naissance
d’un corps médical (Geneva).
Sanmarti, E., and Santiago, R. A. (1987), ‘Une lettre grecque sur plomb trouvée à Emporion
(Fouilles 1985)’, ZPE 68, 119–27.
—— —— (1988a), ‘Notes additionnelles sur la lettre sur plomb d’Emporion’, ZPE 72,
100–2.
—— ——(1988b), ‘La Lettre grecque d’Emporion et son context archeologique’, Revue
Archéologique de Narbonnaise 21, 3–17.
—— ——(1989), ‘Une nouvelle plaquette de plomb trouvée à Emporion’, ZPE 77, 36–8.
Santi Amantini, L. (1995), in L. Santi Amantini, C. Carena, and M. Manfredini (eds.),
Plutarco: Le vite di Demetrio e di Antonio (Milan).
——(2005), Dalle parole ai fatti: Relazioni interstatali e comunicazione politica nel mondo
antico (Rome).
Santiago, R. A. (1990a), ‘Encore une fois sur la lettre sur plomb d’Emporion (1985)’, ZPE
80, 79–80.
——(1990b), ‘Notes additionnelles au plomb d’Emporion 1987’, ZPE 82, 176.
——(2003), ‘Las láminas de plomo de Ampurias y Pech Maho revisitadas’, ZPE 144, 167–72.
Santiago Álvarez, R. A., and Gardeñes Santiago, M. (2006), ‘Algunas observaciones a la
“Lettre d’Apatorios à Léanax” ’, ZPE 157, 57–69.
Saprykin, S. Yu., and Fedoseev, N. F. (2008), ‘New Lead Plaques with Greek Inscriptions
from East Crimea’, VDI (3), 72–9 (in Russian).
—— ——(2010a), ‘A Fragment of Economic Letter on Lead from Panticapaeum’, VDI (1),
50–8 (in Russian).
—— ——(2010b), ‘New Lead Plaques with Greek Inscriptions from East Crimea, Bos-
poros’, in R. W. V. Catling and F. Marchand (eds.), Onomatologos: Studies in Greek
Personal Names presented to Elaine Matthews (Oxford), 427–34.
Saussy, H. (2006), ‘Writing in the Odyssey: Eurykleia, Parry, Jousse, and the Opening of a
Letter from Homer’, Arethusa 29, 299–338.
Savalli-Lestrade, I. (2003), ‘L’Élaboration de la décision royale dans l’Orient hellénistique’,
in F. Prost (ed.), L’Orient méditerranéen de la mort d’Alexandre aux campagnes de
Pompée: Cités et royaumes à l’époque hellénistique (Rennes), 17–39.
 Bibliography 407
Savalli-Lestrade, I. (2007), ‘L’Art de recevoir à la court des Lagides’, in J.-P. Caillet and
M. Sot (eds.), L’Audience: Rituels et cadres spatiaux dans l’antiquité et le haut Moyen-Âge
(Paris), 93–111.
——(2010), ‘Intitulés royaux et intitulés civiques dans les inscriptions de cités sujettes de
Carie et de Lycie (Amyzon, Eurômos, Xanthos). Histoire politique et mutations institu-
tionnelles’, Studi ellenistici 24, 127–48.
Scafuro, A. (2004), ‘The Rigmarole of the Parasite’s Contract for a Prostitute in Asinaria:
Legal Documents in Plautus and his Predecessors’, Leeds International Classical Studies
3/4: 1–21.
Scardino, C. (2007), Gestaltung und Funktion der Reden bei Herodot und Thukydides
(Berlin).
Scherf, J. (1999), ‘Kekrops’, in DNP, vi, 381–2.
Scheurer, S., and Kansteiner, S. (1999), ‘Sophokles. Amphiareos’, in R. Krumeich,
N. Pechstein, and B. Seidensticker (eds.), Das griechische Satyrspiel (Darmstadt), 236–42.
Schierl, P. (2006), Die Tragödien des Pacuvius: Ein Kommentar zu den Fragmenten mit
Einleitung, Text und Übersetzung (Berlin).
Schindel, U. (1970), ‘Phrynichos und die Rückberufung des Alkibiades’, RhM 113, 281–97.
Schloemann, J., and Krumeich, R. (1999), ‘Achaios. Iris’, in R. Krumeich, N. Pechstein, and
B. Seidensticker (eds.), Das griechische Satyrspiel (Darmstadt), 524–9.
Schmidt, M. (1972), ‘Der Tod des Orpheus in Vasendarstellungen aus Schweizer Sammlun-
gen’, in H. P. Isler and G. Seiterle (eds.), Zur griechischen Kunst, Antike Kunst, Beiheft 9
(Basle), 95–105.
Schmidt, P. L. (1997), ‘Brief ’, in DNP, ii, 771–5.
Schmitt, H. H. (1964), Untersuchungen zur Geschichte Antiochos des Grossen und seiner
Zeit (Wiesbaden).
Schneider, J. (2004), ‘Les Traditions relatives à l’origine de l’alphabet grec dans les scholies
de la tekhne de Denys le Thrace’, BAGB 1, 120–57.
Schoep, I. (2007), ‘The Social and Political Context of Linear A Writing on Crete’,
in K. Lomas et al. (eds.), Literacy and the State in the Ancient Mediterranean (London),
53–62.
Schultz, W. (1912), Rätsel aus dem hellenischen Kulturkreise (Leipzig).
——(1914), ‘Rätsel’, RE, iA, 62–125.
Scodel, R. (1980), The Trojan Trilogy of Euripides (Göttingen).
——(2011), ‘Euripides, the Derveni Papyrus, and the Smoke of Many Writings’, in
A. Lardinois, J. Blok, and M. G. M. van der Poel (eds.), Sacred Words: Orality, Literacy
and Religion (Leiden), 79–98.
Scullion, S. (2002), ‘Tragic Dates’, CQ, NS 52, 87–101.
Seaford, R. (2000), ‘The Social Function of Attic Tragedy: A Response to Jasper Griffin’, CQ,
ns 50, 30–44.
Segal, Ch. (1985), ‘Tragedy, Orality, Literacy’, in B. Gentili and G. Paioni (eds.), Oralità,
cultura, letteratura, discorso: Atti del convegno internazionale, Urbino 1980 (Rome),
199–227.
——(1986), ‘Writing, Truth, and the Representation of the Self ’, in Ch. Segal, Interpreting
Greek Tragedy: Myth, Poetry, Text (Ithaca, NY), 75–109.
——(1993), Euripides and the Poetics of Sorrow (Durham, NC).
Shapiro, H. A. (1995), Greek Vases in the San Antonio Museum of Art, San Antonio, TX.
Shebalin, N. V. (1968), ‘On the Antiquities of Olbia (Based on Graffiti)’ (in Russian), in
V. F. Gajdukevič (ed.), Antitchnaia istoria i kultura Sredizemnomoria i Pritchernomoria
[Ancient History and Culture of the Mediterranean and the Black Sea] (Leningrad),
296–9.
Sherk, R. K. (1969), Roman Documents from the Greek East: Senatus Consulta and Epistulae
to the Age of Augustus (Baltimore).
408 Bibliography
Sickinger, J. (1994), ‘Inscriptions and Archives in Classical Athens’, Historia 43, 286–96.
——(1999), Public Archives and Records in Classical Athens (Chapel Hill, NC).
——(2009), ‘Nothing to do with Democracy: “Formulae of Disclosure” and the Athenian
Epigraphic Habit’, in L. Mitchell and L. Rubinstein (eds.), Greek History and Epigraphy:
Essays in Honour of P. J. Rhodes (Swansea), 87–102.
Siebert, G. (1990), ‘Hermes’, LIMC, v.1, 285–387.
Slater, N. W. (1996), ‘Literacy and Old Comedy’, in I. Worthington (ed.), Voice into Text:
Orality and Literacy in Ancient Greece (Leiden), 99–112.
——(2002), ‘Dancing the Alphabet: Performative Literacy on the Attic Stage’, in
I. Worthington and J. M. Foley (eds.), Epea & Grammata: Oral and Written Communi-
cation in Ancient Greece (Leiden), 117–29.
Slavova, M. (2009), ‘The Greek Language on the Bulgarian Coast of the Black Sea in
Antiquity (6th Cent. bc–1st Cent. bc)’, in G. Vottéro (ed.), Le Grec du monde colonial
antique, i: Le N. et N.-O. de la Mer Noire (Nancy), 195–220.
Slings, S. R. (1994), ‘Notes on the Lead Letters from Emporion’, ZPE 104, 111–17.
——(2002), ‘Oral Strategies in the Language of Herodotus’, in E. J. Bakker, I. J. F. de Jong,
and H. van Wees (eds.), Brill’s Companion to Herodotus (Leiden), 53–77.
Smith, J. A. (2003), ‘Clearing up Some Confusion in Callias’ Alphabet Tragedy: How to
Read Sophocles’ Oedipus Tyrannus 332–333 et al.’, CPh 98, 313–29.
Solmsen, F. (1944), ‘The Tablets of Zeus’, CQ 38, 27–30.
Solomonik, E. I. (1987), ‘Deux lettres antiques provenant de Crimée’, VDI (3), 114–31 (in
Russian, with summary in French).
Sommerstein, A. H. (1997), ‘The Theatre Audience, the Demos, and the Suppliants of
Aeschylus’, in C. Pelling (ed.), Greek Tragedy and the Historian (Oxford), 63–79.
——(2000), ‘The Prologue of Aeschylus’ Palamedes’, RhM 143, 118–27.
——(2008) (ed. and trans.), Aeschylus, 3 vols. (Cambridge, MA).
Sosin, J. D. (2008), ‘The New Letter from Pasion’, ZPE 165, 105–8.
Sotinel, C. (2004), ‘How Were Bishops Informed? Information Transmission across the
Adriatic Sea in Late Antiquity’, in L. Ellis and F. L. Kidner (eds.), Travel, Communication
and Geography in Late Antiquity: Sacred and Profane (Aldershot), 63–72.
Squillace, G. (2004), BÆغE j æÆ Ø: Filippo II e Alessandro Magno tra opposizione e
consenso (Soveria Mannelli).
Stéfanis, A. D. (1997), Le Messager dans la tragédie grecque (Athens).
Stehle, E. M. (1994), ‘Cold Meats: Timokreon on Themistokles’, AJPh 115, 507–24.
——(1997), Performance and Gender in Ancient Greece (Princeton).
Steiner, D. T. (1994), The Tyrant’s Writ: Myths and Images of Writing in Ancient Greece
(Princeton).
——(2001), Images in Mind: Statues in Archaic and Classical Greek Literature and Thought
(Princeton).
Stieber, M. C. (2011), Euripides and the Language of Craft (Leiden).
Stirewalt, M. L., Jr. (1993), Studies in Ancient Greek Epistolography (Atlanta).
Stockert, W. (1992), Euripides: Iphigenie in Aulis (Vienna).
Stoddart, S. K. F., and Whitley, J. (1988), ‘The Social Context of Literacy in Archaic Greece
and Etruria’, Antiquity 62 (237), 761–72.
Stolba, V. (2005), ‘A Greek Private Letter from the Settlement of Panskoye I’, VDI (4),
76–87 (in Russian, with summary in English).
Storey, I. C. (1995–6), ‘Notes on Unassigned Fragments of Eupolis’, Museum Criticum
30–1, 137–57.
——(2003), Eupolis, Poet of Old Comedy (Oxford).
Stronk, J. P. (2007), ‘Ctesias of Cnidus, a Reappraisal’, Mnemosyne 60, 25–58.
 Bibliography 409
Sullivan, R. G. (2007), ‘Classical Epistolary Theory and the Letters of Isocrates’, in C. Poster,
and L. C. Mitchell (eds.), Letter-Writing Manuals and Instruction from Antiquity to the
Present (Columbia, SC), 7–20.
Svenbro, J. (1984), La parola e il marmo: Alle origini della poetica greca (Turin).
——(1993), Phrasikleia: An Anthropology of Reading in Ancient Greece (Ithaca, NY)
(English translation of the French edition, Paris 1988).
——(2008), ‘Grammata et stoikheia. Les Scholies à la grammaire de Denys le Thrace’,
Kernos 21, 197–210.
Sykutris, J. (1931), ‘Epistolographie’, in RE Suppl., v (Stuttgart), 186–220.
Symington, D. (1991), ‘Late Bronze Age Writing-Boards and their Uses: Textual Evidence
from Anatolia and Syria’, AnSt 41, 111–23.
Taplin, O. (1986), ‘Fifth Century Tragedy and Comedy: A Synkrisis’, JHS 106, 163–74.
——(2007), Pots and Plays: Interactions between Tragedy and Greek Vase-Painting of the
Fourth Century BC (Los Angeles).
——, and Wyles, R. (2010) (eds.), The Pronomos Vase and its Context (Oxford).
Telò, M., and Porciani, L. (2002), ‘Un’alternativa per la datazione dei Demi di Eupoli’,
QUCC, ns 72/3, 23–40.
Thomas, R. (1989), Oral Tradition and Written Record in Classical Athens (Cambridge).
——(1992), Literacy and Orality in Ancient Greece (Cambridge).
——(2000), Herodotus in Context: Ethnography, Science and the Art of Persuasion
(Cambridge).
——(2009), ‘Writing, Reading, Public and Private “Literacies”: Functional Literacy and
Democratic Literacy in Greece’, in W. A. Johnson and P. R. Holt (eds.), Ancient Literacies
(Oxford), 13–46.
——(2010), ‘ “And you, the demos, made an uproar”: Performance, Mass Audiences and
Text in the Athenian Democracy’, in A. Lardinois, J. Blok, and M. G. M. van der Poel
(eds.), Sacred Words: Orality, Literacy and Religion (Leiden), 161–87.
Thompson, H. A. (1948), ‘Excavations in the Athenian Agora: 1948’, Hesperia 17, 211–29.
Thompson, S. (1955–8), Motif-Index of Folk-Literature: A Classification of Narrative
Elements in Folktales, Ballads, Myths, Fables, Mediaeval Romances, Exempla, Fabliaux,
Jest-Books, and Local Legends, rev. edn (Copenhagen).
Thornton, J. (2002), in M. Mari (trans.) and J. Thornton (notes), Polibio: Storie (Libri V–VI)
(Milan).
Thraede, K. (1970), Grundzüge griechisch-römischer Brieftopik (Munich).
Thür, G. (2005), ‘The Role of the Witness in Athenian Law’, in M. Gagarin and D. Cohen
(eds.), The Cambridge Companion to Ancient Greek Law (Cambridge), 146–69.
Timm, T. (2005), Der Diskos von Phaistos—Fremdeinfluss oder kretisches Erbe?
(Norderstedt).
Tite, Ph. L. (2010), ‘How to Begin, and Why? Diverse Functions of the Pauline Prescript
within a Greco-Roman Context’, in S. E. Porter and S. A. Adams (eds.), Paul and the
Ancient Letter Form (Leiden), 57–99.
Todd, S. (1990), ‘The Purpose of Evidence in Athenian Courts’, in P. Cartledge, P. Millett, and
S. Todd (eds.), Nomos: Essays in Athenian Law, Politics and Society (Cambridge), 19–39.
Too, Y. L. (1995), The Rhetoric of Identity in Isocrates (Cambridge).
Torrance, I. (2010), ‘Writing and Self-Conscious Mythopoiesis in Euripides’, Cambridge
Classical Journal 56, 213–58.
Tuplin, C. (2004), ‘Doctoring the Persians: Ctesias of Cnidus, Physician and Historian’, Klio
86, 305–447.
——(2009), ‘The Gadatas Letter’, in L. G. Mitchell and L. Rubinstein (eds.), Greek History
and Epigraphy: Essays in Honour of P. J. Rhodes (Swansea), 155–84.
Traill, A. (2005), ‘Acroteleutium’s Sapphic Infatuation (Miles 1216–83)’, CQ, ns 55, 518–33.
Trapp, M. (2003), Greek and Latin Letters: An Anthology with Translation (Cambridge).
410 Bibliography
Troiani, L. (1979), ‘Il funzionamento dello stato ellenistico e dello stato romano nel V e VI
libro delle Storie di Polibio’, in Ricerche di storiografia greca di età romana (Pisa), 9–19.
Turato, F. (1976), ‘Seduzione della parola e dramma dei segni nell’Ippolito di Euripide’,
Bollettino dell’Istituto di Filologia Greca dell’Università di Padova, iii, 159–83.
Turner, E. G. (1975), ‘I libri nell’Atene del V e IV secolo a.C.’, in G. Cavallo, Libri, editori e
pubblico nel mondo antico (Rome), 5–24 (rev. edn of Athenian Books in the Fifth and
Fourth Century bc (London, 1952)).
Uchitel, A. (2004), ‘Local Differences in Arrangements of Ration Lists on Minoan Crete’, in
M. Brosius (ed.), Ancient Archives and Archival Traditions: Concepts of Record-Keeping
in the Ancient World (Oxford), 139–52.
Untersteiner, M. (1967), Sofisti: testimonianze e frammenti, ii: Gorgia, Licofrone, Prodico,
2nd edn (Florence).
Usener, K. (1994), ‘Palamedes: Bedeutung und Wandel eines Heldenbildes in der antiken
Literatur’, Würzburger Jahrbücher für die Altertumswissenschaft 20, 49–78.
Usher, S. (2009), ‘Oratio Recta and Oratio obliqua in Polybius’, GRBS 49, 487–514.
van den Hout, M. (1949), ‘Studies in Early Greek Letter-Writing I–II’, Mnemosyne, 4th ser.,
2, 19–41 and 138–53.
van Effenterre, H., and Ruzé, F. (1994–5), Nomima: Recueil d’inscriptions politiques et
juridiques de l’archaïsme grec, 2 vols., Rome.
van Leeuwen, J. (1898), Aristophanis Nubes (Leiden).
Vanstiphout, H. (2004), Epics of Sumerian Kings: The Matter of Aratta (Leiden).
Vasunia, Ph. (2001), The Gift of the Nile: Hellenizing Egypt from Aeschylus to Alexander
(Berkeley).
Veldhuis, N. (2006), ‘How did they learn Cuneiform? Tribute/Word List C as an Elemen-
tary Exercise’, in P. Michalowski and N. Veldhuis (eds.), Approaches to Sumerian
Literature: Studies in Honour of Stip (H. L. J. Vanstiphout) (Leiden), 181–200.
Verbrugghe, G. P., and Wickersham, J. M. (1996), Berossos and Manetho, Introduced and
Translated: Native Traditions in Ancient Mesopotamia and Egypt (Ann Arbor).
Versnel, H. S. (1991), ‘Beyond Cursing: The Appeal to Justice in Judicial Prayers’, in
C. A. Faraone and D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion
(Oxford), 60–106.
——(2002), ‘Writing Mortals and Reading Gods: Appeal to the Gods as a Dual Strategy for
Social Control’, in D. Cohen, (ed.), Demokratie, Recht und soziale Kontrolle im klas-
sischen Athen (Munich), 37–76.
——(2009), ‘Prayers for Justice, East and West: New Finds and Publications since 1990’, in
R. Gordon and F. M. Simón (eds.), Magical Practice in the Latin West: Papers from the
International Conference held at the University of Zaragoza, 30 Sept.–1 Oct. 2005
(Leiden), 275–335.
——(2011), Coping with the Gods: Wayward Readings in Greek Theology (Leiden).
Viereck, P. (1888), Sermo graecus quo senatus populusque romanus magistratusque populi
romani usque ad Tiberii Caesari aetatem in scriptis publicis usi sunt (Göttingen).
Vinogradov, Y. G. (1971), ‘Drevnejsee greceskoe pis’mo s ostrova Berezan’ [‘The Most
Ancient Greek Letter from the Island of Berezan’], VDI (4), 74–100.
——(1981), Olbia: Geschichte einer altgriechischen Stadt am Schwarzen Meer, Xenia 1
(Constance).
——(1994), ‘Greek Epigraphy of the North Black Sea Coast, the Caucasus and Central
Asia’, ACSS 1, 63–74.
——(1997a), Pontische Studien: Kleine Schriften zur Geschichte und Epigraphik des
Schwarzmeerraumes (Mainz).
——(1997b), ‘A letter from Gorgippean rural estates’, ACSS 4, 232–44.
 Bibliography 411
Vinogradov, Y. G. (1998), ‘The Greek Colonisation of the Black Sea Region in the Light
of Private Lead Letters’, in G. R. Tsetskhladze (ed.), The Greek Colonisation of the
Black Sea Area (Stuttgart), 153–78.
——, and Golovačeva, I. V. (1990), ‘Novyj istočnik o pokhode Zopiriona’, in V. L. Yanin,
Numizmatičeskie issledovanija po istorii jugo-vostočnoj Evropy (Kishinev), 15–30.
Virgilio, B. (2010), ‘La Correspondance du roi hellénistique’, in I. Savalli-Lestrade and
I. Cogitore (eds.), Des rois au prince: Pratiques du pouvoir monarchique dans l’orient
hellénistique et romain (Grenoble), 101–22.
——(2011), Le Roi écrit: La Correspondance du souverain hellénistique, suivie de deux
letters d’Antiochos III à partir de Louis Robert et d’Adolf Wilhelm (Pisa).
Voelke, P. (2001), Un théatre de la marge: Aspects figuratifs et configurationnels du drame
satyrique dans l’Athènes classique (Bari).
Volkmann, H. (1954), ‘Die Inschriften im Geschichtswerk des Herodot’, in Convivium:
Beiträge zur Altertumswissenschaft Konrat Ziegler zum siebzigsten Geburtstag (Stuttgart),
41–65.
Voutiras, E. (1998), Dˇ˝SˇFO˝ˇS G`
ˇ: Marital Life and Magic in Fourth
Century Pella (Amsterdam).
——(1999), ‘Euphemistic Names for the Powers of the Nether World’, in D. R. Jordan,
H. Montgomery, and E. Thomassen (eds.), The World of Ancient Magic (Bergen), 73–82.
Wachter, R. (1996), ‘Alphabet’, in DNP, i, 536–47.
——(1998), ‘Griechisch åÆEæ: Vorgeschichte eines Grusswortes’, MH 55, 65–75.
Wade-Gery, H. T. (1952), The Poet of the Iliad (Cambridge).
Walbank, F. W. (1957), A Historical Commentary on Polybius, i (Oxford; repr. with
corrections, 1970)
——(1967), A Historical Commentary on Polybius, ii (Oxford).
——(1972), Polybius, Berkeley.
——(1979), A Historical Commentary on Polybius, iii (Oxford).
——(1984), ‘Monarchies and Monarchic Ideas’, in F. W. Walbank, A. E. Astin,
M. W. Frederiksen, and R. M. Ogilvie (eds.), Cambridge Ancient History, vii.1: The
Hellenistic World (Cambridge), 62–100.
——(1985a [1965]), ‘Speeches in Greek Historians’, in F. W. Walbank, Selected Papers:
Studies in Greek History and Historiography (Cambridge), 242–61.
——(1985b [1977]), ‘Polybius’ Last Ten Books’, in F. W. Walbank, Selected Papers: Studies
in Greek History and Historiography (Cambridge), 325–43.
Wallace, S. (2013), ‘Adeimantos of Lampsakos and the Development of the Early Hellenis-
tic Philos’, in V. A. Troncoso and E. M. Anson (eds.), After Alexander: The Time of the
Diadochi (323–281 b.c.) (Oxford), 142–57.
Wankel, H. (1991a), ‘ “The Hypereides Principle?” Bemerkungen zur Korruption in Athen’,
ZPE 85, 34–6.
——(1991b), ‘Demosthenes 51, 12 und die partielle Atimie’, ZPE 85, 37–9.
Waters, K. H. (1966), ‘The Purpose of Dramatisation in Herodotus’, Historia 15, 157–71.
Wehrli, F. (1969), Herakleides Pontikos, vii: Die Schule des Aristoteles, 2nd edn (Basle).
Weißbach, F. H. (1920), ‘Sardanapal’, in RE ia, 2436–75.
Welles, C. B. (1934), Royal Correspondence in the Hellenistic Period (New Haven).
Wessels, A. (1999), ‘Aischylos. Dike-Drama’, in R. Krumeich, N. Pechstein, and
B. Seidensticker (eds.), Das griechische Satyrspiel (Darmstadt), 98–106.
——, and Krumeich, R. (1999), ‘Isthmiastai oder Theoroi’, in R. Krumeich, N. Pechstein,
and B. Seidensticker (eds.), Das griechische Satyrspiel (Darmstadt), 131–48.
West, M. L. (1985), The Hesiodic Catalogue of Women (Oxford).
——(1990), Studies in Aeschylus (Stuttgart).
412 Bibliography
West, M. L. (1997), The East Side of the Helicon: West Asiatic Elements in Greek Poetry and
Myth (Oxford).
——(2013), ‘The Writing Tablets and Papyrus from Tomb II in Daphni’, Greek and
Roman Musical Studies 1, 73–92.
West, S. (1985), ‘Herodotus’ Epigraphical Interests’, CQ, ns 35, 278–305.
——(1988), ‘Archilochus’ Message-Stick’, CQ, ns 38, 42–8.
——(1994), ‘Prometheus orientalized’, MH 51, 129–49.
West, W. C., III (1973), ‘The Speeches in Thucydides: A Description and Listing’, in Ph.
Stadter (ed.), Speeches in Thucydides (Chapel Hill, NC), 3–15.
Westlake, H. D. (1970–1), ‘Diplomacy in Thucydides’, Bulletin of the John Rylands Library
53, 227–46.
——(1977), ‘Thucydides on Pausanias and Themistocles—A Written Source?’, CQ, ns 27,
95–110; repr. in H. D. Westlake, Studies in Thucydides and Greek History, Bristol 1989,
1–18.
——(1983), ‘Friends and Successors of Dion’, Historia 32, 161–72.
White, J. L. (1981–2), ‘The Greek Documentary Letter Tradition Third Century bce to
Third Century ce’, in Studies in Ancient Greek Letter Writing, Semeia 22, 89–106.
Whitley, J. (1997), ‘Cretan Laws and Cretan Literacy’, AJA 101, 635–61.
——(2001), The Archaeology of Ancient Greece (Cambridge).
Wilamowitz-Moellendorff, U. von (1899), ‘Lesefrüchte 30’, Hermes 34, 219–22 (repr. in
U. von Wilamowitz-Moellendorff, Kleine Schriften, iv (Berlin, 1962), 60–3).
Wilhelm, A. (1904a), ‘Der älteste griechische Brief ’, JOEAI 7, 94–105.
——(1904b), ‘Über die Zeit einiger attische Fluchtafeln’, JOEAI 7, 105–26.
——(1909), ‘Der Brief des Artikon’, JOEAI 12, 118–26.
——(1984), Abhandlungen und Beiträge zur griechischen Inschriftenkunde, i (Leizpig).
Wilkins, J. (2000), The Boastful Chef: The Discourse of Food in Ancient Greek Comedy
(Oxford).
Willi, A. (2005), ‘˚   I ŁÅŒ. Zur Vermittlung der Alphabetschrift nach Griechen-
land’, MH 62, 162–71.
Williamson, M. (1995), Sappho’s Immortal Daughters (Cambridge, MA).
Wilson, J.-P. (1997–8), ‘The “Illiterate Trader”?’ BICS 42, 29–53.
——(2009), ‘Literacy’, in K. A. Raaflaub and H. van Wees (eds.), A Companion to Archaic
Greece (Oxford), 542–63.
Wilson, P. (2009), ‘Tragic Honours and Democracy: Neglected Evidence for the Politics of
the Athenian Dionysia’, CQ, ns 59, 8–29.
Wirbelauer, E. (2004), ‘Eine Frage von Telekommunikation? Die Griechen und ihre Schrift
im 9.–7. Jahrhundert v. Chr.’, in R. Rollinger and C. Ulf (eds.), Griechische Archaik:
Interne Entwicklungen—Externe Impulse (Berlin), 187–206.
Wise, J. (1998), Dionysos Writes: The Invention of Theatre in Ancient Greece (Ithaca, NY).
Wohl, V. (1998), ‘Plato avant la lettre: Authenticity in Plato’s Epistles’, Ramus 21/1, 60–93.
Woodard, R. D. (2010), ‘Phoinikeia Grammata: An Alphabet for the Greek Language’, in
E. J. Bakker (ed.), A Companion to the Ancient Greek Language (Chichester), 25–46.
Woodford, S. (1994a), ‘Palamedes’, in LIMC, vii.1 (Zurich), 145–9.
——(1994b), ‘Palamedes Seeks Revenge’, JHS 144, 164–9.
Worman, N. (2002), The Cast of Character: Style in Greek Literature (Austin, TX).
Wörrle, M. (1978), ‘Epigraphische Forschungen zur Geschichte Lykiens II’, Chiron 8,
241–6.
Worthington, I. (1999), Greek Orators, ii: Dinarchus and Hyperides (Warminster).
——(2006), Demosthenes: Speeches 60 and 61, Prologues, Letters (Austin, TX).
Wright, M. (2005), Euripides’ Escape-Tragedies: A Study of Helen, Andromeda and Iphigenia
among the Taurians (Oxford).
Yatromanolakis, D. (2007), Sappho in the Making: The Early Reception (Cambridge, MA).
Young, D. C. (1983), ‘Pindar’s Pythians 2 and 3’, HSCP 87, 31–48.
 Bibliography 413
Zecchini, G. (2003), ‘Le lettere come documenti in Polibio’, in A. M. Biraschi et al. (eds.),
L’uso dei documenti nella storia antica (Naples), 415–22.
Zeitlin, F. I. (1982), Under the Sign of the Shield: Semiotics and Aeschylus’ Seven against
Thebes (Rome).
——(1994), ‘The Artful Eye: Vision, Ecphrasis and Spectacle in Euripidean Theatre’, in
S. Goldhill and R. Osborne (eds.), Art and Text in Ancient Greek Culture (Cambridge),
138–96.
Zieliński, T. (1931), ‘Die Märchenkomödie in Athen’, Iresione, i: Dissertationes ad comoe-
diam et tragoediam spectantes continens (Leopoli), 8–75.
 Index Locorum

Abbreviations for authors and works tend to follow the conventions of the Oxford Classical
Dictionary—for all others, see under Abbreviations above.

Greek and Latin authors and texts Prometheus Vinctus

Achaeus (TrGF i, 20 ed. Snell) 3–4: 199 n. 51
Omphale F 33: 239 449– 50: 188 n. 15
Aelianus 454–61: 66–7
Varia Historia 459–60: 187–8, 358
1. 25: 56 n. 127, 165 n. 180 506: 188 n. 16
12. 51: 325 535: 259 n. 59
705–6: 188–9
Aeneas Tacticus 788–9: 188–9
22. 3: 11 and n. 39 789: 187 n. 13, 188
22. 22: 11 n. 39 Septem
Aeschines 72: 192 n. 30
On the Embassy (2) 427–31: 193
31: 280 and n. 54 432–4: 193 and n. 33
90: 280 and n. 55 434: 187 n. 13, 193
124–5: 281 468: 187 n. 13, 193
128–9: 281 and n. 57 518: 192 n. 30
130: 11 and n. 38 646: 187 n. 13
134–5: 282 646–8: 193–4, 197
Against Ctesiphon (3) 660: 187 n. 13, 194
132: 282 and n. 60 Supplices
162: 282 179: 187 n. 13, 194
164: 283 180: 192 n. 30
225: 285 438–440: 196
238: 282 and n. 60 455: 196
250–1: 285–6 457–8: 196
463: 15, 196
Aeschylus 708–9: 187 n. 13, 194, 197
Agamemnon 760–1: 196–7
242–3: 187 n. 13, 189, 190 n. 22 774–5: 197
538: 98 938: 195
650–1: 87 942–9: 194–7
801: 187 n. 13 945: 196
801–2: 189 and n. 21 947: 15 n. 54
803: 190 and n. 22 948: 263
1329: 187 n. 13, 190, 198, 204 991–2: 187 n. 13, 194
Choephoroe 1012: 199 n. 51
207 and 209: 190 Fragments (TrGF iii, ed. Radt)
232: 190 Theoroi
450: 187 n. 13, 190 F 78: 191–2
699: 187 n. 13, 190, 207 n. 78 Palamedes
Eumenides F 181: 79
49: 190 F 181a: 187–8 and n. 15
50: 190 F 182: 79–80
273: 82 n. 80 F 182a: 66 n. 25, 187
275: 187 n. 13, 190 Phryges
306: 47 n. 92 F 266: 97 n. 136
322: 47 n. 92 inc. fab.
Persae F 281a, 20–22: 15 and n. 55, 184 n. 5,
254: 184 and n. 5 197–8
294–5: 184 n. 5 F 293: 199 n. 51, 241 n. 192
783: 199 n. 51 F 331: 199; F 358: 199, 260
416 Index Locorum
Agathon (TrGF i, 39 ed. Snell) 14. 24: 19 n. 71
F 4: 237 14. 45: 19 and n. 70
Alcaeus 14. 60: 19 and n. 71
F 401B V. (428a Campbell): 32–3 Anticlides (FGrH 140)
F 130b V (130b Campbell): 32–3 F 11ab: 64 n. 19, 358, 360
Alcidamas Antiphanes (PCG ii, ed. Kassel-Austin)
On the Sophists (1) Knoithideus
12: 144 F 122: 249 n. 222
15: 255 Neottis
22–3: 144 F 167: 256 n. 253
25: 255 Poiesis
32: 144–5 and n. 127 F 189: 247
33: 144 and n. 125 Problema
34: 144 and n. 126 F 192: 249
Odysseus (2) Sappho
75 and n. 71, 88 n. 103 F 194: 244–7, 253 and nn. 242, 254–7
24: 67–8, 75 F 195: 249
Alexis (PCG ii, ed. Kassel–Austin) Tritagonistes
Epistole F 207: 247
F 81: 243 Antiphon
Cleobulina 1. 30: 270–1
F 101: 249 5. 53–6: 269–70
inc. fab.
F 272: 240 Apollodorus (FGrH 244)
F 165: 358
Ampelius
Liber memorialis [Apollodorus]
8. 5: 88 n. 103 Bibl.
2. 1. 5: 74 and n. 56
Anaxandrides (PCG ii, ed. Kassel-Austin) 2. 1. 5. 14: 83 n. 85
Gerontomania 3. 2. 2. 1: 83 n. 85
F 10: 69 and n. 37, 252 Epit.
Anaxilas (PCG ii, ed. Kassel-Austin) 5. 19: 87
Neottis
F 22: 249 Apollonius Dyscolus
De syntaxi seu constructione
Anaximander (FGrH 9) 3. 63–77: 92 n. 115
F 3: 64 n. 19, 358
Apollonius Rhodius
Anaximenes of Lampsacus (FGrH 72) 1. 133–8: 73–4
T 6: 166 and n. 183
F 11 a and b: 166 and n. 185 Apollonius of Tyana
F 13: 212 n. 92 Letters
F 41: 166–7 62. 1–2 and 63: 322 and n. 60
Andocides Archilochus
On the Mysteries F 185 W.: 32
83–6: 212 n. 92 F 89 W.: 32
Andriskos of Naxos (FGrH 500) Aristomenes (PCG ii, ed. Kassel-Austin)
F 1: 43 n. 1. Goetes
F 9: 242 n. 199
Andron of Ephesus (FGrHCont iv, 1005)
F 5/6: 70 n. 38, 359 Ariston F 14 Wehrli (Rusten 2002 col. 17): 294
and n. 97.
Andron of Halicarnassus (FGrH/BNJ 10)
F 9: 70 and n. 38, 359 Aristophanes
Anonymus De Mulieribus Acharnenses
2: 159 n. 158 676–81 and 684: 253
7: 88 681: 253 n. 242
685–6: 254
Anthologia Palatina 705: 254
6. 269: 251 n. 234 716: 254
7. 193. 4: 251 n. 234 Aves
9. 684: 359 879: 94 n. 122
 Index Locorum 417
960–92: 242 Rhetorica
974: 15 n. 54 1368b5–25: 201 n. 58
982: 203 n. 63 1407b34: 228
1035–57: 242 [Athenaion Politeia]
1050–4: 260 n. 262 43. 6: 267
1288: 263 n. 270 [De mundo]
Ecclesiazusae 398a30–5: 12 and n. 43
1050: 242 n. 197, 260 398b4–7: 128 n. 84
Equites [De Mirabilibus Auscultationibus]
1037: 251 and n. 232 57, 834b: 198 n. 48
1058–9: 251 and n. 232 Fragments (ed. Rose)
Lysistrata F 501: 358, 360
513: 260 n. 262 Arrian
1074–5: 94 Anabasis
Nubes 1. 10. 4: 266
584: 95 2. 14. 1–3: 166 n. 185
607–11: 94–5 2. 14. 5–9: 166 n. 185
608–9: 95 n. 127
770–2: 259 n. 259 Asclepiades of Tragilus (FGrH 12)
931: 255 F 6b: 67 n. 30
1320: 254 n. 247 Astydamas (TrGF i, 60 ed. Snell)
Pax Palamedes
582: 98 F 5a: 76
523: 98
Athenaeus
635–7: 254 and n. 243
1. 3 A: 32
Plutus
1. 19 A: 80
322–3: 96 n. 134, 98
1. 28 bc: 65
567–70: 254
6. 222 A: 247
Ranae
6. 223 D–224 B: 256 n. 253
91: 255
6. 230 EF : 161 n. 168, 162 n. 172
917: 255
6. 244 A: 15
943: 262
6. 248 E: 165 and n. 179
948–54: 255
6. 250 AD: 165 and n. 179
1407–10: 262–3
7. 276 A: 235 n. 16–6
1069: 255
7. 289 DE: 325
1492: 255
10. 426 E: 76 n. 64
Thesmophoriazusae
10.432 D: 259 n. 259
431–2: 299
10. 451 CD: 238–9
768–84: 82–4, 241–2
10. 453 C–454 A: 235–6
848: 84 and n. 91
10. 454 BC: 236–7
Vespae
10. 454 D: 237
583–7: 82 n. 82
10. 454 DE: 237
Fragments (PCG iii.2, ed. Kassel-Austin):
10. 454 F: 238
Babylonii
11. 466 DE: 239–40
F 71: 127 n. 80, 240 n. 90
11. 466 EF: 239
Gerytades
11. 481 EF: 239–40
F 163: 259 n. 259
12. 539 B–540 A: 165 and n. 181
Tagenistae
13. 586 C: 161 n. 168, 162 n. 171
F 506: 262 n. 268, 263 n. 270
13. 595 AC: 161 n. 168, 162 n. 171
inc. fab.
13. 599 D: 248 n. 216
F 634: 240
14. 614 C: 69 n. 37
Aristoteles Babrius
Poetica 75: 198 n. 49
1454b30: 206 127: 198 n. 49
1454b30–5: 230
1455a16–19: 230 Bacchylides
1458a26: 251 n. 234 19: 74 n. 53
Politica Berossus (FGrH/BNJ 680)
1270b28º31: 330 and n. 89 F 1: 23–4
1272a35–9: 330 and n. 89 F 4: 65 n. 24
418 Index Locorum
Callias (PCG iv, ed. Kassel-Austin) Ctesias (FGrH 688)
T 7 K.-A.: 235–6, 249 n. 223 T 7d: 157–8
inc. fab. T 8: 155 n. 152
F 28: 249 n. 223 F 1b: 13 n. 44, 156, 158
F 5: 156
Callimachus F 7: 159 n. 158
frr. 429–53 Pf.: 15 F 8a: 159
Hymn 6 (Demeter) F 8b: 159F 15: 158
56: 198 n. 49 F 27: 157
Chares of Mytilene (FGrH 125) F 32: 157–8
F 10: 56 n. 127, 165
Curtius
Cicero: 4. 1. 7–14: 166 n. 185
familiares
Cypria (PEG, ed. Bernabé)
2. 4. 1: 2
argumentum
2. 9. 2: 3 n. 8
p. 40, 30–33 and p. 43, 66:
15. 16. 1: 3 n. 8
74 and n. 55
15. 21. 4: 6.
F 30: 74 and n. 55
Philippicae
2. 7: 3 Demetrius
Ad Quintum fratrem De elocutione,
1. 1. 37: 2 n. 2 213: 158–9
Clemens Alexandrinus 223: 3
Stromateis 225: 3 and n. 9
1, 16, 75. 1: 64 n. 19 226: 3 and n. 10
227: 4
Comica adespota (PCG viii, ed. 228: 142
Kassel-Austin) 234: 4, 6 and n. 18, 267 n. 9
F 456: 242 n. 195, 321 and n. 65
[Demetrius]
F 1084. 29–33: 240 n. 190, 243
Typoi epistolikoi (ed. Weichert),
F 1096. 25, 44 and 48: 243
proemium, 5 and n. 16
F 1139. 4: 243
1: 5–6, 8
Cornelius Scipio Africanus, P., (FGrH 232)
F 1: 176 and n. 215 Demosthenes
First Philippic (4)
Crates (PCG iv, ed. Kassel-Austin) 19: 276
Rhetores F 30: 20: 277
253 and n. 241 30: 277
Cratinus (PCG iv, ed. Kassel-Austin) 37–8: 277
Cleobulinae On Halonnesus (7)
F 92–101: 248 1: 277
F 94: 251 and n. 232 33: 277 and n. 45
F 96: 251 n. 234 Third Philippic (9)
F 101: 249 n. 222 27: 279
Nomoi Reply to Philip (11):
F 128: 259 n. 259 166–7 and nn. 185–6
Pytine 17: 279 and n. 52
F 208 and 209: 242 n. 196 Philip (12):
Trophonius 166–7 and n. 186, 276
F 243: 241 2: 275 n. 37
Cheirones 22: 275 n. 37
F 267: 263 n. 270 On the Crown (18)
inc. fab. 127: 293 n. 93
F 316: 17, 241 164–7: 279
191 and 198: 254 n. 247
Cratinus iunior (PCG iv, ed. Kassel-Austin)
209: 282 n. 59, 293 n. 93
inc. fab.
218–21: 279–80
F 12: 215 n. 100
261: 293 n. 93
Critias (88 D.-K.) On the Corrupt Embassy (19)
B 2. 10: 65 36–8: 278
B 5: 120 n. 59 40–1: 278 and n. 47
B 6: 259 n. 259 44–5: 279
B 22: 253 n. 240 51–2: 279
 Index Locorum 419
174: 275 and n. 33 14. 80. 7: 163 n. 177
187: 279 14. 101. 2: 12 n. 44
200: 282 n. 59, 293 n. 93 17. 39. 1–2: 166 n. 185
237: 293 n. 93 18. 8. 4: 266
314: 293 n. 93 18. 48. 2: 284 n. 65
Against Aristocrates (23) 19. 1. 4: 13 n. 44
115: 175–6 19. 13. 5: 12 n. 44
151–2: 275 and n. 34 19. 13. 7: 12 n. 44
153–6: 275 19. 14. 4: 12 n. 44
178: 276 19. 57. 5: 12 and n. 44
Against Nausimachus and Xenopeithes (28) 19. 85. 5: 12 n. 44
12: 273 19. 100. 3: 12 n. 44
Againt Phormio (34) 20. 18. 1: 13 n. 44
8: 273–4 Diogenes Laertius
28: 241 n. 193, 274 1. 89: 248 n. 218
28–9: 274 2. 44: 84 n. 90
32: 274 n. 31 3. 61: 92 n. 114
Against Timotheus (49) 8. 86: 15
13: 265 and n. 3 9. 6: 32
Against Polycles (50) 9. 54: 84 n. 90
18: 273
On the trierarchic crown (51) Dionysius of Halicarnassus
12: 254 and n. 246 De Thucydide
Against Nicostratus (53) 42: 142
5–6: 273 Dionysius of Miletus (FGrH 687)
46: 273 F 1: 64 n. 19, 358
Against Dionysodorus (56) Dionysius Scytobrachion (FGrH/BNJ 32)
8: 274 n. 32 F 8: 64 n. 19, 69, 361–2
10: 274
Dionysius Thrax
Dictys Techne
6. 2: 87 n. 102 9: 14 and n. 51
Dinarchus Diphilus (PCG v, ed. Kassel-Austin)
Against Demosthenes (1) inc. fab.
27: 284 F 101: 256 n. 252
35–6: 283–4 Dosiadas (FGrH/BNJ 458)
Dio Chrysostomus F 6: 358, 360
Orationes Duris (FGrH/BNJ 76)
13. 21: 84 n. 90 F 6: 69 and n. 34, 359
59, 1–11: 76 n. 65 F 51: 165
Diodorus Siculus F 66: 127 n. 80
1. 49. 2: 258 Empedocles (31 D.-K.)
1. 70. 4: 166 and n. 182 B 23: 261 n. 264
1. 95: 122–3 and n. 68 Ephorus (FGrH 70)
2. 13. 2: 156 F 105a, b, c: 64 n. 19,
2. 18. 1–2: 158 74 n. 53, 358–61
2. 22. 5: 156
2. 23. 3: 156 Eubulus (PCG v, ed. Kassel-Austin)
2. 26. 8: 13 n. 44, 15 and n. 54 Neottis
2. 32. 4: 156 F 69: 240
2. 34. 1–5: 159 n. 158 Sphingokarion
3. 67. 1: 64 n. 19 F 106: 251
3. 67. 2: 67 n. 30 F 107: 248 n. 220
4. 25. 1: 67 n. 30 Eupolis (PCG v, ed. Kassel-Austin)
5. 57: 360–61 Aureum genus
5. 58. 3: 64 n. 19 F 316: 96
5. 74.1: 361 Demoi
11. 21. 4: 12 n. 44 F 96: 254 n. 245
11. 28. 5: 12 n. 44 F 99. 35: 97
11. 45. 2: 12 n. 44 F 102: 254 n. 245
13. 54. 3: 12 n. 44 F 116: 255
420 Index Locorum
Eupolis (PCG v, ed. Kassel-Austin) (cont.) 128–9: 232 n. 159
inc. fab. 153–6: 232
F 327: 242 n. 195, 263 n. 270 198: 74 n. 56
F 331: 95–6 320–6: 233
Euripides 331: 233
Alcestis 340: 233
965–72: 217 342–5: 233
966–9: 68 n. 32 363: 234
1002–5: 213–4 743: 234
Bacchae 794–800: 234
442: 17 798: 215
Electra 891: 234–5
956: 207 n. 78 Iphigenia in Tauris
1073: 207 n. 78 31–3: 224 n. 132
Hecuba 220: 224 n. 132
291–9: 211 222–3: 206 n. 76
736–7: 209 n. 86 582–94: 17, 225 and n. 136
798–805: 210–11 589: 17
807–8: 209 727–8: 228–9
864–7: 86 n. 98, 211 731–3: 17
Heracles 735: 16
1118: 208 745: 16
Hippolytus 755–8: 226
113: 91, 221 759–65: 226
386–7: 216 760–1: 231
451–4: 215–6 763: 252
487: 216 n. 103 765: 224 n. 133
612: 221 n. 121 769: 226
646: 220 770–1: 226–7
670–71: 219 n. 116 774: 227
856–9: 220 778: 227
864–5: 220 786–7: 227
877: 252 791–2: 227
877–80: 220 793–4: 227
925–6: 220 808: 227
928–31: 220 814: 227 and n. 142
935: 184n. 5 821–2: 228
952–5: 216–7 1092: 251 n. 233
954: 68 n. 32 1446: 17
1004–5: 208 1446–7: 229
1051–6: 221 1464–5: 227–8 n. 142
1057–9: 221 Medea
1078–9: 221 n. 121 663–4: 98 n. 141
1252–4: 204 n. 65, 221 Orestes
1310–11: 222 432: 73 n. 49, 74, 87 n. 102
1320–3: 221 Phoenissae
1428: 217 n. 106 129–30: 208
Ion 574–6: 214–5
265: 208 1135: 208
271: 208 1506: 251 n. 233
442–3: 209–10 Supplices
1146: 208 430–2: 211
Iphigenia in Aulis 433–4: 211
35–7: 230 and n. 154 1201–4: 212–3
94: 230 Troades
98–9: 230 686–7: 207–8 and n. 80
108: 230 1188–91: 213
111–3: 231 1244–5: 214
117–18: 231 Fragments (TrGF v, ed. Kannicht):
115–23: 231 Antigone
F 169: 207 n. 78
 Index Locorum 421
Erechtheus Hermesianax
F 369: 215 F 7. 47–56 Powell: 248 n. 216
Theseus Hermippus (FGrH (Continued) 1026)
F 382: 236–7 F 51: 284 and n. 64
Melanippe:
F 506: 210 Herodotus
Meleager 1. 37. 2: 120 n. 61
F 528a: 206 n. 76 1. 47–48. 1: 114
Palamedes 1. 51. 3: 130
F 578: 81–2, 251–2, 358 1. 53: 106 n. 16, 108, 110
F 588: 84 n. 90 1. 60. 5: 108 n. 24, 111 n. 33
F 588a: 82–3, 84 n. 91 1. 69. 1–2: 109
Peleus 1. 90. 3: 121 n. 62
F 618: 207 and n. 79 1. 94: 80 n. 73
Pleisthenes 1. 99–100: 89, 233
F 627: 16 n. 59, 217–8 1. 108. 4–5: 119
Phrixus b 1. 123.3–124: 113, 127–8
F 819: 69 n. 33 1. 124: 16, 114 and n. 40, 118 n. 53
inc. fab. 1. 125: 113, 115, 118 n. 53, 124, 125,
F 923: 215 127–8, 184 n. 5
F 1041: 207 n. 79 1. 129. 1–3: 119–21
F 1059, 5–6: 208 and n. 81 1. 206. 1: 107 and n. 21, 109 n. 27
Palamedes’ Hypothesis (Luppe 2011): 76–7, 1. 212: 107 n. 21, 109 n. 27
83, 84–5 2. 18. 1: 120 n. 61
Stheneboea’s Hypothesis: 222 2. 114. 1: 107 n. 19, 124 n. 71
2. 114. 2–3: 107 n. 20
Eustathius 2. 125. 6: 118 n. 53
Commentarii in Homeri Iliadem (ed. 3. 14. 8–9: 109 n. 26, 111 n. 35
M. van der Valk) 3. 14. 10: 107 n. 21
in Il. 1. 24, 29. 15: 18 n. 65 3. 20–2: 114
in Il. 2. 308, 228. 2–4: 74 n. 55 3. 21. 1: 108 n. 23, 110
in Il. 2. 308, 228. 8: 73 n. 49 3. 21. 3: 110
in Il. 2. 641–3, 312. 6: 18 n. 65 3. 40: 15, 113, 114 and n. 40, 121–2,
in Il. 6.169, 632. 10: 245 n. 205 127–8
Euthycles (PCG v, ed. Kassel-Austin) 3. 41. 1: 118 n. 53, 122
T 1: 243 3. 42. 4–43. 1: 113, 115, 122–3
Gorgias 3. 43: 15, 118 n. 53
Palamedes 3. 68. 4–5: 107 n. 20
6–7: 85 3. 69. 1: 124 n. 71
30: 86 3. 69. 2: 107 n. 21
3. 119. 3: 107 n. 21, 109, 110 and n. 28,
Gregorius of Nazianzus 115
or. 4, 107 (Against Julian 1. 644): 72 3. 119. 4: 107 n. 20, 108 n. 23
Hecataeus (FGrH 1) 3. 119. 5: 107 n. 21, 109, 110 and n. 28,
F 20: 64 n. 19, 74 n. 53, 75, 358 115
Hecataeus of Abdera (FGrH 264) 3. 119. 6: 107 n. 21
F 6: 74 n. 53 3. 121. 1–2: 123 n. 70
F 25: 166 and n. 182 3. 122. 1–4: 123–4
3. 122. 3: 107 and nn. 19, 22, 114, 116
Hellanicus (FGrH 4) 3. 123. 1–2: 126
F 178/178ab: 88–9 3. 126. 1: 124 n. 73
Hellenica Oxyrhynchia 3. 126. 2: 124
F C v 2 (Flor.): 162–3 and n. 176 3. 127. 3: 125
F B xiii 7–8 (Lond.): 162–3 3. 128: 15, 113, 114 and n. 41, 118 n. 53,
and n. 177 125, 127–8
Heraclides Ponticus 3. 128. 5: 124 n. 73
F 149 (ed. Wehrli): 94 n. 121 3. 142–3: 126 n. 76
F 156: 93–4 3. 150–60: 172 n. 202
4. 36. 2: 120 n. 60
Heraclitus (22 D.-K.) 4. 80. 2: 107 nn. 19–20
B 101a: 262 4. 126: 107 n. 21
422 Index Locorum
Herodotus (cont.) 9. 98. 2: 107 n. 19
4. 127: 107 n. 21 9. 98. 3: 82 n. 82, 107 n. 21, 116–17
4. 131. 4: 114 Hesiod Aegimius
5. 14: 16, 113, 115, 118 n. 53, 124 n. 71, F 297 M.-W.: 74
127–8
5. 24: 107 n. 21, 114, 116, 120 n. 58 Hesychius (Hesychii Lexicon, ed. K. Latte,
5. 35: 113, 115 and n. 43, 127–8 P. A. Hansen and I. C. Cunningham)
5. 49: 15 Æ 2842 IºØæØ: 28
5. 52–3: 113, 127  1775  Łıæ ªæÆ Æ Ø: 243 n. 199
5. 57–61: 64 n. 19, 65, 74 n. 54  1992 ØçŁæƺØç : 28
5. 58. 3: 15, 16 and n. 59, 28, 358–9  2401 æ ŒBæı : 11 n. 38
5. 94. 2: 120 n. 61  1799 KŒçØ ÆØ: 64 n. 20
5. 95: 32  5255 KØºÆ : 18 n. 65
5. 98: 108 n. 24, 110–11 and n. 32  658 ÆæƪªfiÅ: 206
6. 4: 113, 115, 128  1190 ŒıºÆØ: 198 n. 49
6. 61: 67 n. 27 ç 688 çØØŒ Ø ªæ ÆØ: 206
6. 67. 2: 129 Homer
6. 97. 2: 107 n. 21 Iliad
6. 98. 3: 114 n. 38 6. 168–70: 48 n. 95, 59–62, 222–3, 245
6. 105–6: 11, 93, 108 n. 24 n. 205
7. 35. 2: 107 and nn. 19 and 21, 109, 110 6. 169: 15 and n. 55, 197, 202, 359
and n. 28, 115 6.176 and 178: 60
7. 100. 1: 125 n. 75 7. 89–90: 213 n. 96
7. 133–7: 128 7. 175–89: 50
7. 135–136.2: 108 n. 24 24. 112–26: 104–5
7. 140–2: 114 24. 133–7: 105
7. 150. 1–2: 107 n. 22, 108 n. 26, 111 Odyssey
and n. 34 12. 67: 15
7. 157: 108 n. 23, 110 and n. 30 14. 141: 15
7. 158: 108 n. 24
7. 172: 108 n. 24, 110 n. 31 Homeric Hymn to Demeter
7. 239: 15, 113, 115 and n. 44, 118 n. 53, 315–17: 105–6
121, 127–8 321–3: 106
8. 8. 1–2: 12 and n. 43 Hyginus
8. 22: 114–15 and n. 42, 116–17, 127–8 Fab. 105: 78
8. 24. 2: 108 n. 24 Fab. 117: 83 n. 85, 87 n. 102
8. 28. 1: 107 n. 19
8. 29. 1: 107 n. 21 Ion of Chios (FGrH 392)
8. 61. 2: 120 n. 61 F 2: 76 n. 64
8. 67. 2–68. 1: 128 Isocrates
8. 75. 2–3: 108 n. 23, 110, 117 n. 50, 141, Panegyricus (4)
148 n. 137 180: 152 n. 148
8. 85. 3: 139 n. 108 To Philip (5)
8. 90. 4: 126 n. 75 81: 288 and n. 74
8. 98: 113, 127 Trapeziticus (7)
8. 110. 2–3: 108 n. 23, 110, 117 n. 50, 141 81: 287 and n. 73
8. 114: 106 n. 16, 108 n. 24 Busiris (11)
8. 128: 113, 115, 118 n. 53, 127 2: 291
8. 135–6: 114 Panathenaicus (12)
8. 140: 109 n. 26 107: 152 n. 148
8. 136. 1: 118 n. 53 Antidosis (15)
8. 140–41: 112–13, 124 n. 71 1: 291 n. 89
8. 142: 108 n. 23, 110 and n. 30 Trapeziticus (17)
8. 221. 2: 82 n. 82 52: 287 and n. 72
9. 7: 108 n. 23, 110 and n. 30 57–8: 287
9. 12: 11, 108 n. 24 Aegineticus (19)
9. 18. 3: 107 nn. 19–21 12–14: 287
9. 21. 1–2: 109 n. 26 Ep. 1. 1–3: 290 and nn. 82–5
9. 48: 111 Ep. 2. 13: 289
9. 49: 111 Ep. 3. 1: 289
9. 60: 107 n. 21 Ep. 3. 4: 290
 Index Locorum 423
Jerome Menander (BNJ 783)
Ep. 8.1: 3 n. 8, 250 F 5: 359
Josephus Moeris
Antiquitates Judaicae Lexicon
12. 148–9: 306 n. 30 å 37: 96 and n. 133
12. 166–70: 322 and n. 69 Nicolaus of Damascus (FGrH 90)
12. 225–8: 322 and n. 69 F 1: 158
14. 241–4: 311 n. 46 F 5: 158–9
Justin Nicostratus (PCG vii, ed. Kassel-Austin)
1. 2. 1 –11: 88 n. 104 Rhetor(?)
Libanius F 24: 253 and n. 241
Decl. 2 (De Socrati silentio) Nonnus
28: 83 n. 86 Dionysiaca
Lindian Chronicle (FGrH/BNJ 532) 1. 481: 198 n. 49
F 1 (A) 6–7: 268 n. 11 4. 259–64: 64 n. 19
F 2 (B) 5–9: 268 and n. 11 Nepos
F 2 (B) 15–17: 64 n. 20 Eumenes (1): 266
Livy Miltiades (4) 3: 93 n. 119
23. 27. 10: 171 n. 198 F 59 Marshall: 243 n. 202
24. 31. 6: 171 n. 196
40. 23–4: 178 n. 221 Nostoi (PEG, ed. Bernabé)
42. 49: 171 n. 199 F 1: 74, 87
Lucian Orus (ed. Alpers)
Consonantium lis F A 42: 17 n. 65
5. 3: 362 Pacuvius
De mercede conductis Doulorestes
12: 198 n. 49 113–15, 136, 137–8 Ribbeck3: 83 n. 85
De Saltatione
69: 72 n. 45 Paeon of Amathus (FGrH/BNJ 757)
Pro lapsu inter salutandum F 2: 162
2: 90–1 and n. 11 Parthenius
3: 91 Amatoriae narrationes
4: 92 n. 114 9. 5: 43 n. 78
6: 92 n. 114, 93–4 Pausanias
10: 4 and n. 12 1. 22. 6: 83 n. 85, 87 n. 102
Lyrica Adespota 1. 28. 4: 93 n. 119
1018 Campbell: 206 n. 76 2. 20. 3: 73 n. 49
Lycurgus 4. 22. 5–6: 322 and n. 71
F xv 7 Conomis (31 Malcovati): 285 6. 18. 2: 166
6. 18. 5: 166, 167
Lysias (ed. Carey): 8. 54. 6: 93 n. 119
On the property of Aristophanes (19) 9. 31. 4: 16
23: 271–2 and n. 25 10. 31. 2: 74 n. 54
For Polystratus (20) 10. 31. 1–3: 76
27: 271 and n. 24
In Thrasybulum Pherecrates (PCG vii, ed. Kassel-Austin)
F 158: 241 Graes
Macho (PCG vii, ed. Kassel-Austin) F 38 and 39: 94 n. 122
F 2: 243 Pherecydes (FGrH 3)
Mnaseas F 4: 74 n. 56, 83 n. 85
F 52 Cappelletto: 358, 360 F 21: 69 n. 33
Menander Philemo (PCG vii, ed. Kassel-Austin)
Leucadia Apollo(?)
11–16: 248 F 10: 71 n. 43, 257–8
Sicyonii, Palamedes(?)
130–41 and 141–4: 243 and n. 200 F 60: 258 n. 257
248: 243 n. 200 inc. fab.
Fragments (PCG vi.2, ed. Kassel-Austin) F 132: 243 and n. 201,
F 238: 243 258 n. 257
424 Index Locorum
Philochorus (FGrH 328) Curculio
F 55 a and b: 167 n. 186 426–36: 244 n. 204
F 89: 161 and n. 146 Miles
F 91: 161 and n. 164 1216–83: 246 n. 207
Philonides (PCG vii, ed. Kassel-Austin) Persa
inc. fab. 500–27: 244 n. 204
F 7: 201 n. 57 Pseudolus
35–6, 63–4: 3 n. 8
Philostratus (ed. Kayser) 41–73: 244 and n. 204
Vitae Sophistarum II 998–1014: 244 n. 204
33. 3 (628): vii Rudens
De gymnastica 13–15: 198 n. 49
4: 12 21: 198 n. 49
Photius 478: 239 n. 185
Bibl. 1156–65: 240 n. 190
72. 35b 35: 155 n. 152 Plinius
72. 43a 17–28: 158 NH
72. 44b 20–45a 4: 157 7. 84: 93 n. 119
7. 192–3: 363–4
Phylarchus (FGrH 81) 7. 193: 64 n. 19
F 41: 165 and n. 181 13. 88: 223 n. 127
F 55: 13 and n. 46
F 59: 165 n. 181 Plutarch
Agesilaus
Pindar 21. 5: 325
Olympian Odes 23. 6: 325
2. 85: 251 n. 233 Alcibiades
10. 1–3: 259 and nn. 258, 261 33. 1: 120 n. 59
F 260 S.-M.: 75 Alexander
Plato comicus (PCG vii, ed. Kassel-Austin) 34. 1: 266
Phaon Artaxerxes
F 189: 250 and n. 227 11: 149
inc. fab. 21. 4: 157–8
F 202: 254 Demetrius
24. 6–10: 295 and n. 99–100
Plato Demosthenes
Epistulae 9. 5: 256 n. 253
2. 312d: 15 n. 56 20. 4–5: 284
Gorgias 23. 2: 284
451c: 298 n. 4 29.3–30.1: 284
Leges 31. 3: 284
3, 680a: 212 n. 92 Dion
Phaedrus 14. 4–7: 163
257e1–258a9: 298 n. 4 26. 7–10: 163–4
274c: 64 n. 19 31. 2–32. 1: 164
275e: 290 n. 84 Eumenes
Philebus 1. 2–3: 266
18b: 64 n. 19 Lycurgus
Protagoras 19. 4: 323
335e: 11 25. 4: 132 n. 94
Respublica Lysander
2. 364e3–365a3: 68 n. 32 7. 3: 321
7. 522d: 80 n. 75 14. 4–5: 323 and n. 73
Theaetetus 19.4: 323–4
191c–192c: 259 and n. 259 19. 7–12: 321 n. 66
Plautus 20. 1–4: 323–4
Bacchides 23: 324
728–47: 244 n. 204 25–6: 324
810–11: 223 n. 127 28. 2–5: 324
996–1035: 244 n. 204 Nicias
1000: 294 n. 97 5: 146 n. 130, 233
 Index Locorum 425
Pericles 4. 67. 8: 170
26. 4: 127 n. 80 4. 87. 8: 170
Phocion 5. 17. 9: 170
21. 1: 266 5. 20: 170
30. 9–10: 284 n. 65 5. 26. 2: 170
Theseus 5. 28: 173
20: 162 5. 28. 3: 170
Moralia 5. 28. 4: 168 n. 188, 170
(Rect. rat. aud.) 5. 37–8: 173
47 E: 204 n. 66 5. 38: 170
(Apophth. Lac.) 5. 42. 7–8: 173, 177 n. 220
211 BC: 324 n. 75 5. 43. 1: 173
212 E: 325 5. 50. 11–13: 170, 173–4
213 DE: 325 5. 57–8: 170
218 EF: 325 5. 61. 3: 170
219 A: 235 5. 101. 6–7: 171
219 D: 323 5. 102. 2: 171
222 AB: 325 5. 102. 8: 171
225 CD: 325 7. 14b: 171 and n. 196
228 D: 323 8. 15. 9: 171
229 B: 323 n. 73 8. 16. 9: 171
(Apophth. Lacaen.) 8. 17: 171, 177 n. 230
241 E: 325 n. 76 9. 5. 1: 171
(De mul. vir.) 9. 5. 6: 171
254 D: 43 9. 20. 3: 172 n. 200
(De glor. Ath.) 10. 9. 3: 176
347 C: 93–4 10. 43–7: 179
(An seni) 11. 5. 23: 172 n. 200
790 AB: 2 13. 3. 1: 172
Polemo (FHG iii ed. Müller) 13. 4. 7: 168 n. 188
F 32: 74 n. 55 13. 9. 4: 172
F 83: 198 n. 48 15. 1. 4: 171
15. 16. 8: 172
Pollux 15. 25. 4: 174 n. 207, 177 n. 220
1. 96: 321 15. 29. 6: 168 n. 190
3. 148. 16: 93 n. 119 16. 9. 1: 176 n. 216
7. 211 16. 15. 8: 176
10. 57: 243 n. 199 16. 20: 161 and n. 165
Polyaenus, 16. 36. 3–4: 175–6
Stratagemata 16. 36–7: 175–6
1 prooem. 12: 75 and nn. 59–60 18. 33. 1–3: 178
4. 2. 8: 127 n. 79 21. 8. 2: 174 n. 209
4. 3. 19: 127 n. 79 21. 11: 177
4. 11. 2–3: 127 n. 79 21. 26. 77: 174 n. 209
5. 17: 172 n. 202 21. 44. 3: 174 n. 209
5. 26: 11 n. 38 22. 3. 1: 174 n. 209
7. 16: 163 n. 177 22. 10. 10–12: 175 n. 210
22. 12. 7: 175 n. 210
Polybius 24. 2. 1: 174 n. 209
1. 3. 3–4: 168 24. 2. 4: 174 n. 209
1. 79. 9–10: 169, 172–3 27. 4. 1: 171
2. 6. 4: 169 27. 7. 1–10: 175
2. 50. 3: 169 and n. 193 28. 13. 6–8: 174 n. 209
2. 50. 10: 169–70 n. 193 29. 25. 7: 171–2, 175
2. 61. 1–5: 13 and n. 46, 170 30. 8: 176 and n. 214
and n. 194 36. 10. 5: 176
4. 2: 173 41. 13. 8–14.1: 172 and n. 201
4. 9: 170
Sappho
4. 22. 2: 15 n. 54
F 31 V.: 33, 246 and n. 207
4. 24. 5: 172 n. 200
4. 25. 4: 172 n. 200 Schol. Aesch.
4. 26. 5–6: 170 PV 458: 66 n. 25, 187
426 Index Locorum
Schol. Ar. Oedipus Tyrannus
Nub. 411: 201
609a: 95 and n. 125 Philoctetes
609b: 97 and n. 135 1325: 200
Plut. Trachiniae
322c: 90 and n. 108, 96–7 46–8: 202
Thesm. 156–8: 202
770a/b: 83–4 493: 200, 203
Schol. Dion. Thr. (Grammatici Graeci i. 3, ed. 575–6: 203
Hilgard) 614–5: 200 n. 54
32. 9–13: 357 623: 204
32. 12–13: 69 n. 34 682–3: 203–4, 207
182. 15–183. 23: 357–8 1165–7: 202–3 and n. 63
183. 1: 64 n. 19 1168: 203
183. 11: 64 n. 19 Fragments (TrGF iv, ed. Radt):
183. 14: 63 Amphiaraus
183. 20–1: 69 n. 35 F 121: 238
184. 20: 64 n. 19, 70 n. 38 Andromeda
184. 20–186. 4: 358–60 F 128: 199–200
184. 23: 64 n. 21 Achaion Syllogos
184. 27–9: 69 n. 34 F 144: 205, 260
184. 29–185.2: 69 n. 35 Helenes apaitesis
185. 9: 69 n. 34 F 177: 205 n. 71
185. 24–186. 2: 65 n. 24 Nauplius navigans
190. 19–20: 64 n. 19 F 429: 80, 205 n. 71
190. 19–35: 360 F 432: 80, 87 n. 99
192. 7: 69 n. 35 Palamedes
197. 17–23: 14 n. 52 F 479: 79 n. 70, 80
479: 249 n. 221 Poimenes
F 514: 64 n. 19, 205–6
Schol. Eur. F 520: 206
Orestes Tereus
432: 72 and n. 47, 77–8, 83 n. 85 F 586: 206
Rhesus F 595: 206 and n. 76
895: 67 n. 30 Triptolemos
Schol. Hom. F 597: 200–1
Il. 6.168–9a: 60 inc. fab.
Schol. Pind. F 784: 200 and n. 55, 218 n. 110
Ol. 6. 190–1: 321 n. 66 F 811: 201
F 890: 206 n. 76
Schol. Soph.
Aj. 781: 18 n. 65 schol. Steph. Byz. s.v. —ƺ،: 198 n. 48
OC 1601: 18 n. 65 Stesichorus PMG
Servius in F 213: 62–3, 74–5, 358, 360
Aen. 2. 81: 76 n. 65, 78–9 Strabo
Skamon of Mytilene (FGrH/BNJ 476) 5. 4. 13: 13 and n. 47
F 3: 69–71 13. 1. 8: 32
Simonides Suetonius,
PMG F 178: 15 Divus Augustus 49. 3: 13 n. 47.
Sophocles Suda (Suidae Lexicon, ed. Adler)
Ajax Æ 4076 (IæåÆØ æÆ B ØçŁæÆ ): 28
748: 199 3664 (s æØ): 92 n. 114
Antigone Å 305 ( ææ  ): 12 n. 42
448: 201 Œ 1547 (ŒÅæŒØ ªæÆç): 200 n. 55
450–5: 201–2  44 (—ÆºÆ Å ): 73
Electra ç 787 (ØØŒØÆ ªæ ÆÆ): 68
639: 184 n. 5 å 156 (åÆ æØ): 92
Oedipus Coloneus å 162 (åÆ æØ): 96 n. 133
1601: 199 å 164 (åÆ æØ): 97 n. 135
 Index Locorum 427
Tacitus 3. 29: 132
Annales 11. 14: 70 and n. 39, 73 3. 104. 5: 142 n. 117
Tatianus 3. 113. 3: 137
Oratio ad Graecos 1: 88–9 4. 22. 1–3: 136
4. 40. 2: 137
Theocritus of Chios (FGrH /BNJ 760) 4. 50. 2: 137, 141–2, 149
T 1: 162 4. 117. 3: 94 n. 122
Theodectas of Phaselis (TrGF i, 72 ed. Snell) 4. 16. 3–17. 1: 133
inc. fab. 4. 19. 1: 133
F 6: 237 4. 20. 3–4: 133–4
Theophrastus 4. 104. 4: 142 n. 117
Characters 5. 9. 7: 137 n. 102
23. 4: 293–4 and n. 95 5. 17. 2: 142 n. 117
24. 13: 294 5. 27. 2: 136
5. 36: 136
Theognetus (PCG vii, ed. Kassel-Austin)
5. 37: 136
F 1: 262 n. 268 5. 43. 3: 136
Theognis (ed. West) 5. 45: 132
19–23: 33–4 and n. 48 5. 61. 2–3: 134 and n. 98
39–40: 256 n. 254 5. 77. 1: 142 n. 117
46 and 50: 256 n. 254 5. 77. 5: 152 n. 149
1081–2: 256 n. 254 5. 78: 142 n. 117
Theopompus (FGrH 115) 6. 81: 132 and n. 93
T 20a: 161 n. 168, 162 n. 171 7. 7–8: 145
T 48: 161 and n. 167 7. 8. 2: 143 and n. 121, 146 n. 133
F 251: 161 n. 168 7. 10: 142, 146 n. 133
F 252: 161 and n. 168, 162 n. 172 7. 11–15: 142–6, 149, 275
F 253: 161 and n. 168, 162 and n. 171 7. 11. 1: 143 and nn. 120, 122
F 254: 161 and n. 168, 162 and n. 171 7. 16: 142
7. 48. 2: 148 n. 136
Thucydides
8. 33. 3: 146 and n. 133, 325
1. 1: 130–1
8. 38. 4: 146, 325
1. 22. 1: 132, 145 and n. 128
8. 39. 2: 146
1. 36. 4: 132 n. 92
8. 45. 1: 146–7
1. 44: 132 n. 92
8. 50–51: 147–8
1. 53. 1–3: 135
8. 50. 2: 137
1. 67: 132 n. 92
1. 72–9: 131–2 nn. 91–2 Timaeus of Tauromenium (FGrH/BNJ 566)
1. 90. 1–2: 132 n. 92 F 32: 165 and n. 179
1. 91: 131 n. 91 F 113: 163–4
1. 119–24: 132 n. 92 Timocles (PCG vii, ed. Kassel-Austin)
1. 125. 1: 152 n. 149 Epistulae frr. 9–10: 243
1. 128. 3: 140 Timon of Phlius,
1. 128. 6: 142 n. 117 Suppl. Hell. 835: 206 n. 74
1. 128. 6–129. 1: 138–9
1. 128. 7: 16, 137, 142 n. 117 Timonides (FGrH/BNJ 561)
1. 129. 1: 16, 18 F 1: 163–4
1. 129. 3: 138–9, 142 n. 117 Turpilius
1. 130. 1–2: 140, 233 F 195 Rychlewska: 3 n. 8, 250
1. 131. 1: 140 Tzetzes
1. 132. 2: 215 n. 99 Allegoriae Iliadis (ed. Boissonade)
1. 132. 5: 138, 140, 231 n. 155 Prol. 402–5: 88 n. 103
1. 134. 1: 140
1. 137. 3: 140 Virgil
1. 137. 4: 137, 141, 142 n. 117 Aeneid 2. 82: 73, 78
1. 138. 1: 141 Xenarchos (PCG vii, ed. Kassel-Austin)
1. 139. 3: 132–3, 137, 142 n. 117 Pentathlos F 6: 201 n. 57
2. 2. 4: 135 Xenophon
2. 5. 5: 135 Agesilaus
2. 71. 1–72.2: 132 n. 92 8. 3: 18 n. 68, 154–5, 325
2. 72. 2: 132 n. 92 Anabasis
2. 73. 2–74. 1: 134–5 1. 2. 20: 126 n. 75
428 Index Locorum
Xenophon (cont.) Inscriptions and papyri
1. 6. 3: 18 n. 68, 153 NB: The Roman letters listed in Appendix 3
1. 8. 25: 149 do not figure in the index. The Greek letters
3. 1. 4–5: 154 listed in Appendices 1 and 3 are included;
5. 3. 13: 150 n. 143 references to discussion in the text are to
7. 2. 8: 154 page numbers, followed by their appendix
Apologia Socratis number. The Lindian Chronicle and the Helle-
26: 84 n. 90 nica Oxyrhynchia are listed in the index of
Cyropaedia: Greek and Latin passages.
2. 2. 9: 18 n. 68 Archaiologike Ephemeris 1917 10 n. 304:
2. 9–10: 153 app. 3 no. 21
2. 10: 18 n. 68
4. 5. 26: 18 n. 68 Bernabé – Jiménez San Cristóbal 2008
4. 5. 26–34: 153 L 7a and b (Graf and Johnston 2007 no.
4. 5. 34: 18 n. 68 26a and b, lamellae from Pelinna):
4. 7. 12–13: 153–4 54–5
5. 5. 4: 18 n. 68, 153 L 8 (G.-J. no. 3, lamella from Thurii): 55
7. 3. 15: 150 n. 143 n. 124
8. 2. 16–17: 18 n. 68, 153 L 10 a and b (G.-J. nos. 6–7, lamellae
8. 6. 17–18: 12 n. 43, 154 from Thurii): 55
Hellenica L 13a (G.-J. no. 28, lamella from Pherae):
1. 1. 23: 150, 152 n. 147, 275, 321 55
1. 4. 3: 18 n. 68, 151, 152 L 14 (G.-J. no. 17, lamella from
1. 7. 4: 18 n. 68, 151 and n. 145, 275 Rethymnon): 54
1. 7. 17: 151 and n. 146, 275 L 15 (G.-J. no. 15, lamella from
2. 1. 7: 321 Eleutherna): 54
3. 3. 8: 150, 275 L 15a (lamella from Heraclea): 54
3. 10: 150–1 L 16b (G.-J. no. 31, lamella from Pella):
4. 8. 11: 321 53–4
5. 1. 5: 321 L 16k (G.-J. no. 37, lamella from
5. 1. 6: 321 Vergina): 54
5. 1. 30: 151–2 L 16n (G.-J. no. 30, lamella from
5. 1. 31: 152–3, 300 and n. 13 Amphipolis): 55
5. 1. 32: 151 BGU
6. 2. 25: 321 1232: 13 n. 45
6. 3. 9: 153 CEG i
6. 13. 12: 153 11, 12 and 13: 213 n. 96
7. 1. 36–9: 153 179, 256, 351 and 421: 214–5 n. 99
7. 1. 37: 153 432 and 453: 27
7. 1. 39: 18 n. 68 454: 27–8, 31
Memorabilia CID 4
3. 10. 1: 261 n. 265 11, 1–5: 328
115: 326–7, app. 3 no. 42
Zonaras 120: 316, 326, app. 3 no. 45
Lexicon (ed. J.A.H. Tittmann) 121 and 122: app. 3 no. 51–52
804 (Kغ): 17 and n. 65 123–5: app. 3 no. 56
CIRB
Hebrew, Sumerian and Akkadian texts 1, 2 and 3: 302
Chronicle of Early Kings (Grayson 1975) DT
Chron. 20. 7: 61–2 43: 48 and n. 96, 52 n. 108
Enmerkar and the Lord of Aratta (ETCSL 44: 48 and n. 96, 52 n. 108
c.1.8.2.3): 23, 63 52: 52
Genesis 39:1–20: 60 n. 6 DTA
1 Kings 21: 9–10: 62 n. 15 55: 48 n. 94
2 Kings 10:1–10: 62 n. 15 67: 51–2
1 Maccabees 99: 51
12:6–8: 322 and n. 68 102: 48 and n. 96, 49–50
12:19–23: 322 and n. 68 103: 48 and n. 96, 49–50
14:16–23: 322 and n. 68 107: 50 and n. 100, 51 and n. 102
2 Samuel 11: 60–1 158: 48 n. 94
Sargon and Ur-Zababa (ETCSL c.2.1.4): 60–81 Fouilles de Delphes
and n. 8 III 1.487b: 321, app. 3 no. 64
 Index Locorum 429
III 1.488: app. 3 no. 65 2 21: app. 3 no. 60
III 1.489: app. 3 no. 66 2 367: app. 3 nos. 40–41
III 1.490: app. 3 no. 67 IG ix
III 1.495: app. 3 no. 68 12 3, 750: app. 3 no. 22
III 1.578: app. 3 nos. 51–52 12 4, 796B: app. 3 no. 30
III 2.48: 326 n. 78, app. 3 no. 61 2, 11: app. 3 no. 25
III 2.94: 326, app. 3 no. 36 2, 518: 321, app. 3 no. 49
III 2.120: 314–15, app. 3 no. 32 2, 1106: app. 3 no. 48
III 4.33: 326, app. 3 no. 55 IG x
III 4.38, 3: app. 3 no. 56 2, 1 1028: 328, app. 3 no. 3
III 4.59: 312 n. 49 IG xi
III 4.69: 312 n. 49 4, 1053: 328, app. 3 no. 3
III 4.153: 309 IG xii
IC 2, 96 and 97: 64 n. 20
1 v 52: 313, 315–6, 318, app. 3 no. 71 4, 1 142: 319 and n. 60
1 v 53: 313, 315–16, 318, app. 3 no. 72 4, 1 169: 313–14, app. 3 no. 29
1 viii 7: 317, 319, app. 3 no. 5 4, 1 214: 317–20, 328, app. 3 no. 1
1 xxiv 1: 313, 315–6, 318, app. 3 no. 74 4, 1 247: 317, 319, app. 3 no. 5
2 I 2 B: 317–18 and n. 57, app. 3 no. 24 4, 1 248: 317, 319, app. 3 no. 6
2 iii 2: 317–19, app. 3 no. 26 5, 868A: 317–19, app. 3 no. 4
2 v 17: 313, 315–6, 318, app. 3 no. 70 9, 5: app. 3 no. 47
2 v 19: 315, 317, 319–20, app. 3 no. 16 I. Gonnoi
2 x 2: 316, 317, app. 3 no. 13 93, 23–7: 320 n. 63
2 xii 11: 64 n. 20 I. Iasos
2 xv 2: 313, 315–16, 318, app. 3 no. 73 4: 310 and n. 43
2 xvi 2: 317–19, app. 3 no. 4 606: app. 3 no. 17
2 xxiii 1: 318–9, app. 3 no. 15 I. Labraunda
2 xxiii 3: 315–19, app. 3 no. 12 i, 11: 302–3
2 xxvi 1: 317, app. 3 no. 11 ii, 40: 300–2
3 iii 2: 317, app. 3 no. 14 I. Lampsakos
3 iv 9: 315, 317, 320, app. 3 nos. 37–38 4: 169 and n. 191,
4 168: 317, 319, app. 3 no. 6 267, 295 n. 98
I. Cos 7: app. 3 no. 57
ED 56: 307–8 I. Magnesia
ICS 40: 328 n. 84, app. 3 no. 7
no. 143: 28 67: 315, 317, 319, app. 3 no. 8
no. 217: 28 91: app. 3 no. 18–19
I. Délos 105: 315, 317, 320, app. 3 nos. 37–38
iv, 1510: 326, app. 3 no. 27 115: 36 and n. 58, 139 n. 108
iv, 1534, 1535, 1543 and 1549: 321–2 I.v.Olympia
n. 67 7: 30 n. 29
IG ii2 52, 29–40: 320 n. 63, app. 3 no. 39
387: 266 n. 7 IPArk
486: 266 n. 7 6: app. 3 no. 60
1096: 315, 326, app. 3 nos. 62–63 19: app. 3 nos. 40–41
1132: 326–7, app. 3 no. 42 I. Thasos (Dunant-Pouilloux)
1134: 316, 326, app. 3 no. 45 172: app. 3 no. 58
1136: 326, app. 3 no. 54
1456.42–3: 64 n. 20 Lang, Agora 1976
2783: 235 n. 166 B1 (letter from Athens): 44, app. 1 no. 35
2784: 235 n. 166 B2 (letter from Athens): 44, app. 1 no. 36
IG iv B7 (letter from Athens): 44, app. 1 no. 37
179: 34 and n. 52 B9 (letter from Athens): 44, app. 1 no. 38
IG v LSAG 464 n. A ( = Dunst 1969, letter from
1 8: 320, app. 3 no. 50 Roses): 43, 57, app. 1 no. 25
1 9: 320, app. 3 no. 59 Meiggs-Lewis
1 10: 320, app. 3 no. 69 12: 36 n. 58, 139 n. 108
1 28: 320, app. 3 no. 23 30, 35–41: 64 and n. 20
1 30: 320, app. 3 no. 44
1 1524: 312 n. 47 P. Bon.: 5 n. 15
1 1566: 321, app. 3 no. 64 P. Cair. Zen. 3. 59426 (SB 3. 6804): 1
430 Index Locorum
P. Hal. 7. 6: 15 n. 54 43, 434 (curse from Pella): 52
P. Hib. 110: 13 n. 45 43, 488 (letter to Tegeas from Torone):
P. Lond. Lit. 63: 19 n. 71 43–4, app. 1 no. 33
P. Mich. Inv. 3020(A): 73 n. 49, 83 and n. 84 43, 707. 12–15: 311 n. 46, 327
P. Oxy. iv 710, 1–4: 13 n. 45 44, 753 (text from Himera): app. 1 no. 28
P. Paris.: 5 n. 15 46, 943 (letter from Olbia): 41, app. 1
Rhodes and Osborne (2004) no. 15
no. 2, 61–2: 185 n. 41 47, 1175 (letter from Gorgippia): 40 n. 70,
no. 33: 266 n. 7 41, app. 1 no. 13
no. 55: 300–2 47, 1745: 306 n. 31
no. 56: 301 48, 988 (letter from Berezan): 39, 41,
no. 64: 266 n. 7 app. 1 no. 2
no. 84: 266 48, 1011 (letter from Olbia): 41, app. 1
no. 86: 266 no. 3
48, 1012 (letter of Apatorios from Olbia):
Saprykin and Fedoseev VDI 2010 (letter from 28, 39, 41–2, app. 1, no. 5
Panticapaeum): 41, app. 1 no. 9 48, 1024 (letter from Phanagoria): 39–40,
SEG 41, app. 1, no. 4
2, 330: 315, app. 3 no. 43 48, 1029 (letter from Zhivakov Hill): 41,
11, 1025: app. 3 no. 53 app. 1 no. 7
23, 489: app. 3 no. 31 49, 325 (letter from the Pnyx of Athens):
26, 117: 326, app. 3 no. 46 44–5, app. 1 no. 40
26, 845 (letter of Achillodoros from 49, 1116: 313–14, app. 3 no. 29
Berezan): 15–16, 38–9, 41–2, 50, 276 (letter of Lesis from Athens):
app. 1 no. 1 45–6, app. 1 no. 41
26, 1123: 321–2 n. 67 50, 704 (letter of Hermaios from
27, 631 (contract of Spensithios): Panticapaeum): 40–41, 57, app. 1
64 n. 20 no. 10
28, 1224: 308 and n. 38, 311 and n. 45, 50, 1356: 312 n. 47
328 51, 1056: 317–20, 328, app. 3 no. 1
30, 85: 315, 326, app. 3 nos. 62–63 52, 541: 321, app. 3 no. 28
31, 985 (public imprecations from Teos): 52, 994 (1) (Decourt, IGF 130, letter from
64 n. 20 Agde): app. 1 no. 30
33, 841 (letter? from Roses): app. 1 no. 27 53, 256 (letter of Pasion from Athens):
35, 1071 (letter from Emporion): 43, 46, 57–8, app. 1 no. 42
app. 1 no. 26 54, 691 (graffito from Nikonion): 355
36, 949 (letter from Olbia in the Var): 54, 983 (letter of Megistes from
app. 1 no. 31 Marseilles): 43, app. 1 no. 29
36, 1218: 311 n. 46 55, 859 (letter from Panskoje I): 41,
37, 665 (letter to Neomenios from app. 1 no. 17
Kerkinitis): 40, 41, app. 1 no. 8 58, 775 (letter from Acra): 43 n. 78,
37, 673 (curse from Olbia): 51 n. 105 355–6
37, 709: app. 3 no. 20 SGD
37, 838 (letter from Emporion): 42–3, 57, 18: 48 n. 94
app. 1 no. 23 48: 48 n. 94
38, 13 l. 4: 16 n. 58 109: 48 n. 96, 51 and n. 104
38, 85: 315, 326, app. 3 nos. 62–63
38, 619: 302 and n. 18 Syll.3
38, 1036 (lead tablet from Pech-Maho): 283 (letter of Alexander to the Chians):
355 266
38, 1476: 315, app. 3 nos. 9–10 332: 302 and n. 18
39, 1088 (letter from Emporion): 43, 543: 309 and n. 39
app. 1 no. 24 598: app. 3 nos. 18–19
39, 1290. iii: 303 n. 21 622B: 315, 317, 319–20, app. 3 no. 16
39, 1426: 311 n. 46 683: 320 n. 63, app. 3 no. 39
40, 301: 34 and n. 52 1259 (letter of Mnesiergos from Athens):
42, 661: 302 45, 56, app. 1 no. 39
42, 710 (‘priest’s letter’ from Olbia): 41, 1260 (letter of Artikon from Olbia): 41,
app. 1 no. 6 app. 1 no. 12
42, 711 (letter from Kozyrka): 40, 41, Welles,
app. 1 no. 11 RC 1: 304
42, 722 (graffito from Panskoje): 354–5 RC 3. 79: 306 n. 31
 Index Locorum 431
RC 4: 305 RC 61: 305–6
RC 14: 308 and n. 36 Ziebarth
RC 44: 306–7 Neue Verfluchungstafeln
RC 45: 307 20: 50–1
RC 52: 305 26: 50–1
RC 53: 305
 Thematic Index (Names, Places, Topics)

Argos Deioces 89, 121, 140

and epigraphical letters 328, app. 3 nos. 7 Dike (and writing) 15 n. 55, 49, 187 n. 13,
and 33–35 192–4, 197–8, 210, 259
and the traditions on the invention of ο (forms of)
writing 63, 68, 73–4 and n. 54 in city decrees 295 n. 99, 298–300, 309,
Atossa (and communication) 24, 88–9, 99, app. 3 nos. 2, 4, 6, 8, 12, 22, 28
140 n. 109 in city letters 315–16, app. 3 nos. 1, 7, 11,
Attica, Athens 24, 26, 27, 36, 70, 72, 74
and curses 44 n. 82, 46 n. 88, 48–52, 354 in royal documents 301–2
and early letters 44–7, 351–4 in royal letters (absence of) 300,
and official letter writing 265–295, 326–7, 303–6, 315–6
331–2
and the traditions on the invention of enactment and resolution formulae 298–9
writing 69–71, 358–61 and n. 4, 301–2, 314, 315–7, 368
and writing 30–1, 69–71, 238–9 Enmerkar and the invention of writing
23, 63
beacons/fires (and long-distance K ººÆØ 108, 110 n. 30, 117–18
communication) 12, 80, 86–8 envelope 23, 61
Bellerophontes 48, 56, 59–62, 197, 200 n. 56, K Ø
222–3, 245 n. 205, 359 in city letters 316–17
books 15, 28, 31, 39 n. 69, 68 and n. 32, 185 n. in decrees 299
8, 210, 215–19, 234–5, 241–2 and nn. in royal letters (rarity of) 306–8
191, 195 and 197, 250, 262–3, 293 n. 93 epigram (and communication) 10 n. 33, 36
brevity (epistolary) 4, 146, 150, 167, 329 n. n. 54, 56 n. 125, 57 n. 128, 58, 90 n. 110,
87, 270, 289–90, 321 n. 65, 322–5 and n. 72 213–5, 260
epistolary past 38–9, 42–3, 44–5 and n. 83,
ªæÆ / ªæÆÆ (semantics of) 14–19, 65, 108, 351
67–8, 81, 86, 114 n. 40, 116, 140, 141 epistolary style 3–6, 96–7 and nn. 133–5, 290
(showing), 151–3 K Ø ººø vii, 18, 38 and n. 65, 44–6, 56–8, 91
ªæÆç as painting 15, 16 n. 61, 186–7, n. 113, 94, 118, 146, 227, 265 n. 3, 266 n.
189–92, 193, 207–9, 216, 237, 261–2 6, 271, 277, 284, 289 and n. 78, 335, 336,
344, 352, 353, 355
Cadmus 64–5, 68–71, 74 and nn. 53–4, 258, K Ø º
358, 360, 362–4 in curses 48–50
Crete as marker of epistolarity 289
and official letter writing 313, 315–20, as oral command (‘instructions’) 17–19,
327, 330, app. 3 nos. 4, 5, 6, 8, 11, 12, 94, 120, 146, 199–200 and nn. 51, 53,
13, 14, 15, 16, 24, 26, 37–38, 70–74 203–4, 225, 229, 232, 241
and pre-alphabetic scripts 24–5 semantics of 17–19, 94, 120, 147, 199–200
and the traditions on the invention of and nn. 51, 53, 203–4, 229, 232, 241, 225
writing 68–9, 358–60 and n. 136, 227, 229
and writing 30–1, 53–4, 69 ergon
curses (and letters) 36, 38 n. 64, 46 n. 88, 47–53, As opposed to logos/onoma 119–21, 224
56–8, 64, 159, 220, 227, 340, 348, 354 n. 133, 231–2 n. 159, 226
Writing as an ergon 120, 172 n. 200
Danaus (and writing) 64, 73–5, 78, 194–7,
199 n. 51, 358, 360 Findspots of letters, c. 550 BC to c. 350 BC xx
deceitfulness folds (of tablet) 15 n. 54, 18–19, 25, 34, 52, 54,
of letters 59–62, 75, 76 n. 64, 116–18, 126–7, 60–1, 183, 184 and n. 5, 194–5 and n. 41,
129, 141, 147 n. 135, 150–1, 179, 331 205, 210, 215, 221 n. 121, 224 n. 134,
of oral communication 75–6, 79, 82, 226–9, 230–1, 233 n. 161, 243, app. 1
116–18, 123, 130, 148 n. 137, 163, 204, 331 passim (folds of real lead tablets)
 Thematic Index (Names, Places, Topics) 433
formula valetudinis viii, 1–2, 10, 43, 45, 300, Linus (and invention of writing) 67, 69,
345, 349, 352 359–60
Funerary inscriptions (address in) 29–30, lists and catalogues 24–5, 30, 153–4, 198–9,
57–8, 90 and n. 110, 213–15 205, 207
of discoveries 65, 66–7, 79–81, 86, 89,
Gorgo 88, 113, 325 187–8
greetings of names (in curses) 48, 50–2
epistolary (development of ) 6, 10, 35–6,
39–45, 55, 56–7, 90–2, 99 manipulation of communication
in official letters, with åÆ æ Ø 91, 94–7 oral 82 and n. 82, 116–18, 195 n. 38
in oral official communications 90–2, written 82 and n. 82, 116–18, 127, 156,
93–5, 97–8 274, 276
types of 4 n. 12, 91–2 and n. 114, 93 messengers (carrying an oral or written
message)
justice aggeloi 11–13, 62, 103, 107, 111 n. 35,
administered through letters 88–9 178–9, 192, 241
in curses 48–9, 53 and n. 111, app. 1 no. 38 dromokerykes 11–2
unwritten 201–2, 211–12 and n. 91, 260 grammatophoroi 13, 163–4, 168, 170–1
and writing 86, 152, 197–8, 201, 209–12 and n. 193, 179
hemerodromoi 11–13, 91–3, 97 n. 137
Killing of messengers 78, 124–5, 140, 223 keryx, kerykes (heralds) 11–2, 32, 62, 103,
107 and n. 19, 108 n. 24, 109 n. 26, 111
ºÆºø 3 n. 9, 19, 245, 251, 254–8 and nn. 33–4, 117, 122–3, 131, 134–6, 137,
language and signs 14–5, 27, 30, 59–60 and n. 140, 142 n. 117, 178–9, 192, 194, 196, 200,
2, 78 and n. 68, 80, 87–8, 94 n. 122, 186–7 203–4, 230, 241–2, 267, 282, 303
and n. 10, 197, 202, 206, 220 and n. 117, presbeis (ambassadors) 12, 103, 131–4, 145,
236, 246 170, 176, 259, 267, 274, 278–9, 285,
layout 28, 31 n. 38, 299 n. 7 308–9, 313, 321
letter to Gadatas 36 and n. 58, 139 n. 108. mirror 3–4 and n. 8, 6, 144–5 and n. 127
letters motivation formula with K Ø or K
bringing death to their bearer 59–62, 140, 223 in decrees 266 n. 7, 299, 308, 309, 313–4,
and curses 35, 38 n. 64, 44 and n. 82, 46 app. 3 nos. 22, 29
n. 88, 47, 49–51, 53, 57–8, 159 in letters of cities 316–17, app. 3 nos. 2, 5,
deceitful 173, 220–1, 282–3, 323–4 6, 13, 43, 70, 71, 72, 74
and dialogue 3–6, 9, 89, 143–5, 256, 299 in royal documents 301–2
false 75, 113, 127, 222 in royal letters 306–7
forged 75, 82, 85, 114–5, 125–6, 127, 138 Musaeus (and invention of writing) 66–7, 71,
n. 106, 162, 168, 170, 171 n. 196, 172–4, 358, 360
175, 177–8, 269–70, 285
of Greek cities 175–6, 177–8, 267–8, 274–5, Oeax 73–4, 76–7 and n. 67, 82–5, 87
281–2, 311–7, 317–27, 327–30 official communication 5–7, 11–13, 32, 37 n.
of Hellenistic kings 13, 167, 169–70, 173–4, 60, 43 and n. 78, 44, 56–7, 91–2, 94–9,
177–8, 266–7, 274–80, 281–2, 300, 111–13, 114 and n. 40, 143 and n. 119,
302–11, 328–30 146, 149, 150–3, 155, 169, 170–2, 176,
intercepted (actually or potentially) 12, 177, 200, 226, 231, 260, 266–8, 275–7,
113, 119, 138, 141–2, 150, 153, 163, 170, 293, 297–333
173, 176, 232–3, 235, 306, 322, 324 oracular tablets/ messages 101 n. 3, 104 n. 13,
lost 163–4, 226, 250 108, 114, 118 n. 52, 130, 189, 202–5,
by Persians 112–3, 115, 118, 125, 127–9, 217–18, 241, 242, 248–9, 250–51, 260,
136–9, 141–2, 149–50, 151–3, 153–4, 163 263, 324, 326, 355
n. 167, 178–9, 206, 266–7 n. 8, 282, 284, oral messages
286, 287–8, 293, 332 through intermediaries, creating a
private form of communication 18, 62 distance 109–10, 113, 128
letters, representation in vase–paintings oral and written communication, relationship
Iobates reading letter handed by of 3–4, 12, 17–18, 23, 33 and n. 45, 35
Bellerophontes 233 and n. 126 and n. 54, 39–40, 45 n. 83, 47, 50 n. 99,
Iphigenia handing letter to Pylades 229–30 56, 62, 79, 87 n. 101, 90, 92, 101–2,
and nn. 149–50 103–4, 114–16, 118, 122–3, 125 n. 74,
Phaedra’s tablet 219 n. 114 127–30, 136–7, 145– 6, 186 and n. 9, 188,
Proetus handing letter to 193–4, 195, 201–2, 205, 217, 219, 229,
Bellerophontes 222 and nn. 122–3 232, 234, 259, 268–9, 271, 278, 288
434 Thematic Index (Names, Places, Topics)
Orpheus Persian 16 n. 61, 118 n. 73, 125–6
and books 216–17 of tyrant 126
iconography 68 n. 32 secrecy
inventing writing 66–8, 358, 360, 362 marking oral exchanges 146 n. 32, 306
‘Orphic’ lamellae (and letters) 53–6 marking written documents 191 and n. 24,
146, 195–7, 203–5, 232–4 , 239, 273,
Palamedes 24, 26, 63, 66–9, 72–88, 145, 283–4
187–8, 205, 218, 223, 241–2, 251–2, 256, Sicily
258, 263–4, 270, 358–61, 363–4 curses 47–8, 51
in Etruria 76 n. 63 letters 36, 37 n. 59, 198 n. 48, 350
iconography 76 and nn. 62–3 Sparta
Pausanias the regent 12 n. 44, 137–40, 233, and oral communication 110, 129,
270, 323 132–3, 157
performative effect and letters 113, 115, 121, 136, 137, 138–40,
through monumentalization 7–8 and n. 24, 146–9, 150–51, 154–5, 170, 178–9,
267, 297–8, 317 320–26, 327, 330–2, app. 3 nos. 23, 28,
Persia 44, 49, 59, 64, 69
and letters 141–2, 149, 153–4 skytale 32, 140, 238–9, 321 and n. 66,
Pheidippides: 11, 93, 97 n. 137 323, 332
Phoenician letters, phoinikeia, and writing 31 n. 36, 151–3, 211–2, 238–9
poinikazein 64–5, 68–70, 74 n. 54, 78, Sphragis, seal 15 n. 54, 18, 25, 33–35, 37,
206, 357–61 40, 61, 82 n. 81, 125, 140, 151, 152–3,
Phoenicians (and invention of writing) 26, 172 n. 201, 175, 178, 194–7, 200, 204,
28–9, 64–5, 68–9, 70–1, 74 n. 53, 77–8, 80 220, 223, 230–5, 260, 295 n. 98, 312
n. 74, 205–6, 357–63 n. 49, 313, 320 n. 63, 322 n. 70,
Phoenix son of Agenor 64, 68–9 324, 332
Phoenix the praeceptor of Achilles 69, 357–9 stratagems (involving letters) 75–6 n. 61,
prescript 82–4, 113, 116–17, 121, 127, 140, 150–1,
of decrees 299, 309, 314 172 n. 202, 175–6, 219–20, 324
of letters 1, 4 n. 12, 10 n. 34, 35–6, 38–45, support of (letter) writing 15–6, 18, 25–6,
50, 52, 56–7, 89–92, 98, 109–13, 115, 118 27 n. 17, 30 n. 31, 32, 37, 47 n. 92, 60 n. 3,
n. 54, 122, 124, 129, 159, 167, 289, 300, 115 n. 43, 178–9, 199, 259, 260, 298,
313–4, 316, 318, 322 323 n. 73
Proetus 59–63, 222–3
Connection with Palamedes 73 theatre
Femininity of 62 and n. 14, 264 and writing 183–6, 198 n. 48
Prometheus 66–71, 79, 187–9, 358–60 Thebes
As place of transmission of writing 64–5,
reading (mainly of letters) 3, 29–30, 35 and 68, 70–71, 74 n. 53
n. 54, 48, 52, 118, 158, 198 n. 47, 199, 205 Writing official letters 318–9
and n. 69, 215, 226–7, 231–2, 237, Themistocles 8 n. 24, 113–17, 127–8, 137,
239–40, 242, 243, 244, 258, 271, 286, 324 139–41, 148 n. 137, 176 n. 212
privately 17 n. 62, 114, 122, 145, 163–4, trade (and long–distance
220, 245–6, 258 communication) 28–9, 38–9, 44 n. 79,
publicly 17 n. 62, 106, 113, 125, 165, 152–3, 42–4, 46–7, 273–4, 292, app. 1 nos. 1, 5,
163–4, 165, 170, 171 n. 196, 175, 192–4, 8–10, 12, 14, 20–2, 25B, 1–3
268, 270, 275–80, 284, 287, 308, 315–6, trust/mistrust in written documents 81–2
332
rhetores 144–5, 167, 244–6, 251–6, 264, voice (of letters) 192 and n. 30
271–295, 297
riddles 19, 23, 226–7, 235–9, 244–6, 248–52, women
254–6 As inventors 88–9
And (letter) writing 3 n. 9, 61, 62, 88–9,
Sappho 33, 88–9, 238, 244–58, 294 162, 164, 187, 194–7, 198 and n. 48,
Sargon of Akkad 60–62 200–1, 205, 206, 219 n. 115, 235, 245–6,
scribe 16 n. 61, 26, 38, 41–2, 45–6, 64 n. 20, 255–7, 325
198 n. 47 will, testament 81–2, 168 n. 90, 177 n. 220,
secretary (ªæÆÆØ  vel sim.) 202, 243, 270–1, 292
at Athens 195 n. 41, 293 n. 93, 299 writing
of Greek poleis 320–1 creating a distance 9–10, 58, 72, 88–9, 99,
of Hellenistic kings 170, 266, 321 n. 67 128, 140, 233
 Thematic Index (Names, Places, Topics) 435
enabling an exchange 122, 144, 200 194–5, 197–8, 200–201, 203–5, 210 and
n. 454, 225 nn. 88–9, 235, 239, 259 and nn. 258–61,
enabling poetic invention 19, 24, 66–68, 259–60
186, 189, 213–4 and n. 95, 215, 217 metaphorically speaking 38, 81, 115 and
n. 106, 234 n. 45, 116, 130, 138
enabling exact memory 19 n. 71, 65, (showing), 189, 102, 193, 206–7 and n.
66–7, 71, 82 and nn. 80–81, 86, 76, 218, 220, 245–6, 254–5
143–5, 187–8, 212–3, 224, 235,
259 n. 259, 260 åÆ æ Ø
for lists 24–5, 30, 48, 50–51, 184, 188, 190, for oral greeting 90–95
197–9, 205, 207, 213 n. 96, in gold lamellae 53–6
metaphorical (for firm in letters 1, 6, 10, 35–6, 39–45, 55, 56–7,
remembrance) 184–5, 186–9, 190–1, 90–2, 99, 294, 300, 309, 313, 315